=== 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: fb438e6dbebd23a036a1436a9c376208ffa38d38 ################# 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 2021.11.002, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 5.2.3, 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 (15.09.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2021.11.002 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 5.2.3 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := b4336484fb65b0e73211a8f920ae4361c7c353fd USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/01 job id: 41477302 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/02 job id: 41477305 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/03 job id: 41477306 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/04 job id: 41477308 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/05 job id: 41477309 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/06 job id: 41477310 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/07 job id: 41477313 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/08 job id: 41477314 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/09 job id: 41477316 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/10 job id: 41477317 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/11 job id: 41477319 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/12 job id: 41477320 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/13 job id: 41477321 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/14 job id: 41477324 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/15 job id: 41477325 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/16 job id: 41477326 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/17 job id: 41477327 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/18 job id: 41477328 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/19 job id: 41477330 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/20 job id: 41477331 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/21 job id: 41477332 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/22 job id: 41477334 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/23 job id: 41477335 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/24 job id: 41477338 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/25 job id: 41477339 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/26 job id: 41477340 --- 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/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 0.028 0.036 134.131 134.131 farming_run 1 2.0 133.641 133.643 134.097 134.101 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.449902E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.026 116.815 116.816 qs_energies 1 2.0 0.000 0.000 116.604 116.607 mp2_main 1 3.0 0.000 0.000 114.718 114.721 mp2_gpw_main 1 4.0 0.020 0.025 113.885 113.889 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 94.805 94.888 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 56.195 56.278 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.162 42.477 47.409 get_2c_integrals 1 6.0 0.000 0.000 37.566 38.438 integrate_v_rspace 273 8.0 0.439 0.452 25.147 29.724 pw_transfer 6555 10.6 0.374 0.391 27.522 28.039 fft_wrap_pw1pw2 5465 11.4 0.045 0.046 26.158 26.468 grid_integrate_task_list 273 9.0 20.902 25.981 20.902 25.981 fft_wrap_pw1pw2_100 2178 12.4 1.206 1.422 23.724 24.056 compute_2c_integrals 1 7.0 0.002 0.002 20.334 20.335 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 19.640 20.043 mp2_eri_2c_integrate_gpw 1 9.0 2.383 2.436 19.637 20.040 rpa_ri_compute_en 1 5.0 0.000 0.000 18.971 19.042 cp_fm_cholesky_decompose 12 8.2 17.215 18.139 17.215 18.139 cholesky_decomp 1 7.0 0.000 0.000 16.078 16.974 fft3d_s 5443 13.4 16.166 16.426 16.187 16.448 ao_to_mo_and_store_B_mult_1 272 7.0 10.872 15.588 10.872 15.588 calculate_wavefunction 272 8.0 5.414 5.499 12.546 13.248 rpa_num_int 1 6.0 0.000 0.000 10.588 10.588 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.518 10.546 calc_potential_gpw 544 9.5 0.005 0.006 9.684 10.044 calc_mat_Q 8 8.0 0.000 0.000 9.332 9.417 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.943 9.205 contract_S_to_Q 8 9.0 0.000 0.000 8.752 8.837 potential_pw2rs 545 10.0 0.106 0.108 7.744 8.442 parallel_gemm_fm 14 9.1 0.000 0.000 8.332 8.422 parallel_gemm_fm_cosma 14 10.1 8.332 8.422 8.332 8.422 create_integ_mat 1 6.0 0.023 0.029 8.206 8.206 collocate_single_gaussian 272 10.0 0.040 0.042 7.491 7.729 array2fm 1 7.0 0.000 0.000 7.162 7.685 pw_scatter_s 2720 13.7 4.447 4.599 4.447 4.599 pw_gather_s 2722 13.2 3.978 4.451 3.978 4.451 pw_poisson_solve 545 10.5 2.488 2.529 3.558 3.735 array2fm_buffer_send 1 8.0 3.206 3.477 3.206 3.477 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.888594, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2721.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 0.038 397.025 397.027 farming_run 1 2.0 396.360 396.364 396.987 396.989 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.227268E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.047 210.485 210.486 qs_energies 1 2.0 0.000 0.002 210.202 210.235 scf_env_do_scf 1 3.0 0.000 0.000 106.170 106.170 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.300 105.307 rebuild_ks_matrix 4 6.0 0.000 0.000 105.298 105.306 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.062 105.298 105.306 hfx_ks_matrix 4 8.0 0.001 0.001 104.913 104.916 integrate_four_center 4 9.0 0.144 0.457 104.913 104.916 mp2_main 1 3.0 0.003 0.026 103.737 103.768 mp2_gpw_main 1 4.0 0.035 0.080 102.837 102.847 integrate_four_center_main 4 10.0 0.104 0.490 96.918 99.336 integrate_four_center_bin 263 11.0 96.813 98.978 96.813 98.978 init_scf_loop 1 4.0 0.000 0.000 92.066 92.066 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 75.885 76.968 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.946 56.029 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.161 42.668 47.753 integrate_v_rspace 95 8.0 0.399 0.569 28.618 33.513 pw_transfer 2240 10.6 0.146 0.160 30.167 30.599 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 29.140 29.610 grid_integrate_task_list 95 9.0 23.889 28.963 23.889 28.963 ao_to_mo_and_store_B_mult_1 91 7.0 10.586 28.903 10.586 28.903 mp2_ri_gpw_compute_en 1 5.0 0.057 0.083 26.802 28.519 fft_wrap_pw1pw2_100 730 12.4 1.286 1.456 26.876 27.295 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.846 1.904 25.063 25.074 get_2c_integrals 1 6.0 0.001 0.008 20.830 20.876 compute_2c_integrals 1 7.0 0.003 0.004 19.804 19.810 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 19.343 19.671 mp2_eri_2c_integrate_gpw 1 9.0 1.731 1.859 19.342 19.671 fft3d_s 1823 13.4 18.527 18.870 18.540 18.885 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.102 14.103 calculate_wavefunction 91 8.0 2.022 2.055 9.796 10.092 potential_pw2rs 186 10.0 0.033 0.034 8.709 9.448 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.595 8.775 9.323 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.661 9.071 calc_potential_gpw 182 9.5 0.002 0.002 8.654 8.952 local_gemm 172 8.0 8.219 8.749 8.219 8.749 mp2_ri_gpw_compute_en_comm 22 7.0 0.504 0.519 8.055 8.404 collocate_single_gaussian 91 10.0 0.017 0.022 7.943 8.230 mp2_ri_gpw_compute_en_ener 172 7.0 6.341 6.420 6.341 6.420 mp_sendrecv_dm3 2068 8.0 6.074 6.404 6.074 6.404 mp_sync 38 10.4 2.669 6.109 2.669 6.109 pw_gather_s 912 13.2 4.989 5.477 4.989 5.477 pw_scatter_s 910 13.7 3.979 4.219 3.979 4.219 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.830795, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1514.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 450.322432E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.177 52.583 52.584 qs_mol_dyn_low 1 2.0 0.005 0.020 52.085 52.098 qs_forces 11 3.9 0.005 0.031 51.992 51.994 qs_energies 11 4.9 0.010 0.069 50.484 50.492 scf_env_do_scf 11 5.9 0.000 0.001 44.138 44.138 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 42.054 42.054 dbcsr_multiply_generic 2286 12.5 0.093 0.098 33.395 33.903 qs_scf_new_mos 108 7.5 0.000 0.001 31.822 32.108 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.822 32.108 ot_scf_mini 108 9.5 0.002 0.002 30.169 30.327 multiply_cannon 2286 13.5 0.180 0.189 25.991 27.789 multiply_cannon_loop 2286 14.5 1.475 1.552 25.282 27.022 velocity_verlet 10 3.0 0.009 0.018 25.365 25.368 ot_mini 108 10.5 0.001 0.002 19.331 19.569 qs_ot_get_derivative 108 11.5 0.001 0.002 16.407 16.617 mp_waitall_1 267858 16.1 8.746 15.197 8.746 15.197 multiply_cannon_metrocomm3 54864 15.5 0.066 0.071 5.975 13.455 multiply_cannon_multrec 54864 15.5 4.238 6.609 7.716 11.154 rebuild_ks_matrix 119 8.3 0.000 0.000 8.020 8.160 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.015 8.020 8.160 multiply_cannon_sync_h2d 54864 15.5 5.891 7.385 5.891 7.385 mp_sum_l 7207 12.9 5.461 7.263 5.461 7.263 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.087 7.214 qs_ot_get_p 119 10.4 0.001 0.003 6.257 6.592 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.455 5.903 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.584 5.678 init_scf_run 11 5.9 0.000 0.001 4.809 4.809 scf_env_initial_rho_setup 11 6.9 0.000 0.002 4.809 4.809 dbcsr_mm_accdrv_process 76910 16.1 1.150 1.814 3.402 4.731 sum_up_and_integrate 119 10.3 0.008 0.012 4.684 4.692 integrate_v_rspace 119 11.3 0.002 0.003 4.676 4.685 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.168 4.274 calculate_rho_elec 119 8.7 0.011 0.016 4.168 4.274 qs_ot_p2m_diag 50 11.0 0.004 0.007 3.449 3.522 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 1.798 3.371 rs_pw_transfer 974 11.9 0.012 0.013 3.179 3.314 calculate_dm_sparse 119 9.5 0.000 0.000 2.974 3.123 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.884 3.074 apply_single 119 13.6 0.000 0.000 2.884 3.074 jit_kernel_multiply 13 15.8 2.192 2.860 2.192 2.860 calculate_first_density_matrix 1 7.0 0.001 0.003 2.696 2.699 ot_diis_step 108 11.5 0.006 0.009 2.676 2.676 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.610 2.611 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.386 2.466 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.377 2.377 density_rs2pw 119 9.7 0.004 0.004 2.285 2.374 cp_fm_redistribute_end 50 14.0 2.159 2.355 2.164 2.357 cp_fm_diag_elpa_base 50 14.0 0.191 2.310 0.192 2.323 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.274 2.276 acc_transpose_blocks 54864 15.5 0.222 0.241 1.740 2.189 grid_integrate_task_list 119 12.3 2.028 2.118 2.028 2.118 potential_pw2rs 119 12.3 0.004 0.004 2.069 2.092 init_scf_loop 11 6.9 0.000 0.001 2.066 2.066 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.019 2.058 wfi_extrapolate 11 7.9 0.001 0.002 2.057 2.057 mp_sum_d 4129 12.0 1.395 2.029 1.395 2.029 make_m2s 4572 13.5 0.053 0.055 1.728 1.771 pw_transfer 1439 11.6 0.051 0.055 1.626 1.701 make_images 4572 14.5 0.133 0.138 1.647 1.689 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.552 1.629 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.485 1.497 mp_alltoall_d11v 2130 13.8 1.294 1.475 1.294 1.475 mp_waitany 12084 13.8 1.203 1.416 1.203 1.416 fft3d_ps 1201 14.6 0.356 0.458 1.326 1.396 grid_collocate_task_list 119 9.7 1.288 1.356 1.288 1.356 fft_wrap_pw1pw2_140 487 13.2 0.083 0.096 1.182 1.259 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.762 1.159 acc_transpose_blocks_kernels 54864 16.5 0.246 0.380 0.780 1.064 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.584000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=429.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.202816E+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 2017496. MP_ISendRecv 33558 37093. MP_Wait 40318 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.079 0.137 38.248 38.250 qs_mol_dyn_low 1 2.0 0.006 0.016 37.876 37.883 qs_forces 11 3.9 0.002 0.008 37.752 37.755 qs_energies 11 4.9 0.010 0.043 36.054 36.060 scf_env_do_scf 11 5.9 0.000 0.001 30.647 30.648 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 28.207 28.209 dbcsr_multiply_generic 2286 12.5 0.100 0.103 21.114 21.480 qs_scf_new_mos 108 7.5 0.001 0.001 19.406 19.650 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.405 19.649 ot_scf_mini 108 9.5 0.003 0.004 18.519 18.693 velocity_verlet 10 3.0 0.012 0.042 17.889 17.890 multiply_cannon 2286 13.5 0.207 0.216 16.226 17.772 multiply_cannon_loop 2286 14.5 0.898 0.973 15.159 16.515 ot_mini 108 10.5 0.001 0.002 11.610 11.844 mp_waitall_1 217478 16.2 5.793 10.949 5.793 10.949 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.112 9.442 qs_ot_get_derivative 108 11.5 0.001 0.001 9.167 9.342 multiply_cannon_multrec 27432 15.5 1.966 4.339 5.894 8.723 rebuild_ks_matrix 119 8.3 0.000 0.000 7.074 7.217 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.021 7.074 7.217 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.236 6.367 dbcsr_mm_accdrv_process 47894 16.0 3.006 5.090 3.859 5.710 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.534 4.387 qs_ot_get_p 119 10.4 0.001 0.002 4.067 4.299 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.037 4.114 apply_single 119 13.6 0.000 0.000 3.037 4.114 sum_up_and_integrate 119 10.3 0.015 0.017 4.099 4.106 integrate_v_rspace 119 11.3 0.002 0.005 4.083 4.090 mp_sum_l 7207 12.9 2.076 3.963 2.076 3.963 init_scf_run 11 5.9 0.000 0.001 3.866 3.867 scf_env_initial_rho_setup 11 6.9 0.000 0.002 3.866 3.867 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.513 3.544 calculate_rho_elec 119 8.7 0.021 0.024 3.513 3.543 multiply_cannon_sync_h2d 27432 15.5 2.199 2.873 2.199 2.873 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.647 2.666 make_m2s 4572 13.5 0.052 0.054 2.415 2.630 rs_pw_transfer 974 11.9 0.010 0.012 2.513 2.601 make_images 4572 14.5 0.200 0.237 2.328 2.542 init_scf_loop 11 6.9 0.001 0.004 2.421 2.429 calculate_first_density_matrix 1 7.0 0.000 0.002 2.396 2.399 ot_diis_step 108 11.5 0.011 0.013 2.394 2.395 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.218 2.219 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.082 2.172 calculate_dm_sparse 119 9.5 0.000 0.000 2.077 2.157 density_rs2pw 119 9.7 0.004 0.004 1.882 1.992 grid_integrate_task_list 119 12.3 1.826 1.937 1.826 1.937 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.907 1.908 cp_fm_redistribute_end 50 14.0 1.575 1.886 1.578 1.887 potential_pw2rs 119 12.3 0.006 0.006 1.862 1.870 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.859 1.861 cp_fm_diag_elpa_base 50 14.0 0.300 1.830 0.307 1.860 jit_kernel_multiply 9 16.1 0.802 1.780 0.802 1.780 pw_transfer 1439 11.6 0.063 0.066 1.746 1.778 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.670 1.713 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.658 1.690 make_images_data 4572 15.5 0.045 0.052 1.125 1.532 prepare_preconditioner 11 7.9 0.000 0.000 1.490 1.518 make_preconditioner 11 8.9 0.000 0.001 1.490 1.517 acc_transpose_blocks 27432 15.5 0.107 0.111 1.185 1.487 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.393 1.450 hybrid_alltoall_any 4725 16.4 0.051 0.112 0.973 1.440 wfi_extrapolate 11 7.9 0.001 0.001 1.421 1.422 fft3d_ps 1201 14.6 0.500 0.555 1.354 1.390 fft_wrap_pw1pw2_140 487 13.2 0.087 0.107 1.323 1.355 mp_alltoall_d11v 2130 13.8 1.151 1.351 1.151 1.351 grid_collocate_task_list 119 9.7 1.241 1.323 1.241 1.323 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.295 1.303 mp_allgather_i34 2286 14.5 0.512 1.253 0.512 1.253 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.152 1.203 mp_sum_d 4129 12.0 0.604 1.059 0.604 1.059 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.031 1.035 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.894 0.907 acc_transpose_blocks_kernels 27432 16.5 0.180 0.269 0.658 0.874 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.703 0.769 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.250000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.181818, yerr=1.898238 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 520.286208E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.027 31.664 31.665 qs_mol_dyn_low 1 2.0 0.003 0.003 30.992 31.473 qs_forces 11 3.9 0.002 0.002 30.920 30.921 qs_energies 11 4.9 0.001 0.002 29.361 29.363 scf_env_do_scf 11 5.9 0.000 0.001 24.714 24.715 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.231 22.231 dbcsr_multiply_generic 2286 12.5 0.093 0.094 16.082 16.188 velocity_verlet 10 3.0 0.002 0.002 14.921 14.922 qs_scf_new_mos 108 7.5 0.001 0.001 14.391 14.411 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.390 14.411 multiply_cannon 2286 13.5 0.193 0.201 12.949 13.745 ot_scf_mini 108 9.5 0.002 0.002 13.674 13.690 multiply_cannon_loop 2286 14.5 0.633 0.667 12.182 13.003 ot_mini 108 10.5 0.001 0.001 8.562 8.582 qs_ot_get_derivative 108 11.5 0.001 0.001 7.076 7.094 multiply_cannon_multrec 18288 15.5 1.934 2.940 6.726 7.015 rebuild_ks_matrix 119 8.3 0.000 0.000 6.249 6.269 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 6.248 6.268 dbcsr_mm_accdrv_process 38222 16.0 4.059 5.474 4.707 5.578 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.518 5.537 mp_waitall_1 169478 16.3 2.709 3.895 2.709 3.895 sum_up_and_integrate 119 10.3 0.018 0.019 3.889 3.894 integrate_v_rspace 119 11.3 0.002 0.002 3.870 3.877 init_scf_run 11 5.9 0.000 0.001 3.380 3.380 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.380 3.380 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.637 3.255 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.171 3.176 calculate_rho_elec 119 8.7 0.030 0.031 3.170 3.176 qs_ot_get_p 119 10.4 0.001 0.001 2.972 2.988 init_scf_loop 11 6.9 0.000 0.000 2.463 2.463 multiply_cannon_metrocomm3 18288 15.5 0.045 0.047 1.338 2.456 rs_pw_transfer 974 11.9 0.009 0.010 2.233 2.372 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.019 2.341 apply_single 119 13.6 0.000 0.000 2.019 2.341 calculate_first_density_matrix 1 7.0 0.000 0.000 2.206 2.207 jit_kernel_multiply 11 16.3 0.597 2.028 0.597 2.028 qs_ot_p2m_diag 50 11.0 0.012 0.012 1.960 1.966 grid_integrate_task_list 119 12.3 1.812 1.923 1.812 1.923 density_rs2pw 119 9.7 0.003 0.004 1.769 1.901 make_m2s 4572 13.5 0.044 0.045 1.710 1.852 calculate_dm_sparse 119 9.5 0.000 0.000 1.798 1.808 make_images 4572 14.5 0.190 0.203 1.625 1.765 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.704 1.704 prepare_preconditioner 11 7.9 0.000 0.000 1.697 1.699 make_preconditioner 11 8.9 0.000 0.000 1.697 1.699 potential_pw2rs 119 12.3 0.007 0.008 1.686 1.690 pw_transfer 1439 11.6 0.063 0.066 1.668 1.677 multiply_cannon_sync_h2d 18288 15.5 1.428 1.659 1.428 1.659 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.653 1.654 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.550 1.635 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.578 1.589 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.527 1.536 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.519 1.525 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.486 1.487 cp_fm_redistribute_end 50 14.0 1.104 1.468 1.105 1.468 ot_diis_step 108 11.5 0.011 0.011 1.465 1.466 cp_fm_diag_elpa_base 50 14.0 0.347 1.409 0.361 1.446 mp_sum_l 7207 12.9 1.035 1.401 1.035 1.401 grid_collocate_task_list 119 9.7 1.204 1.285 1.204 1.285 fft3d_ps 1201 14.6 0.506 0.525 1.264 1.274 fft_wrap_pw1pw2_140 487 13.2 0.088 0.091 1.258 1.268 acc_transpose_blocks 18288 15.5 0.074 0.076 1.235 1.258 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.173 1.177 wfi_extrapolate 11 7.9 0.001 0.001 1.131 1.131 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.428 1.055 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.955 0.956 make_images_data 4572 15.5 0.045 0.050 0.768 0.928 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.858 0.879 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.652 0.843 mp_alltoall_d11v 2130 13.8 0.645 0.819 0.645 0.819 acc_transpose_blocks_kernels 18288 16.5 0.207 0.216 0.800 0.811 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.799 0.801 cp_fm_cholesky_invert 11 10.9 0.715 0.719 0.715 0.719 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.656 0.718 rs_pw_transfer_RS2PW_140 130 11.5 0.176 0.180 0.572 0.707 mp_alltoall_z22v 1201 16.6 0.608 0.670 0.608 0.670 mp_waitany 9880 13.7 0.520 0.634 0.520 0.634 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.665000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=494.818182, yerr=2.166614 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 545.370112E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.029 35.337 35.338 qs_mol_dyn_low 1 2.0 0.003 0.003 35.122 35.129 qs_forces 11 3.9 0.001 0.002 35.061 35.062 qs_energies 11 4.9 0.001 0.001 33.312 33.317 scf_env_do_scf 11 5.9 0.000 0.001 28.051 28.052 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.578 24.578 dbcsr_multiply_generic 2286 12.5 0.099 0.102 18.525 18.635 velocity_verlet 10 3.0 0.005 0.008 18.016 18.018 qs_scf_new_mos 108 7.5 0.001 0.001 16.284 16.338 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.283 16.337 ot_scf_mini 108 9.5 0.002 0.003 15.345 15.400 multiply_cannon 2286 13.5 0.231 0.265 14.814 15.203 multiply_cannon_loop 2286 14.5 0.934 0.964 13.848 14.175 ot_mini 108 10.5 0.001 0.001 9.495 9.563 multiply_cannon_multrec 27432 15.5 2.381 3.053 8.895 9.230 dbcsr_mm_accdrv_process 47916 15.9 5.383 7.554 6.420 7.740 qs_ot_get_derivative 108 11.5 0.001 0.001 7.675 7.733 rebuild_ks_matrix 119 8.3 0.000 0.000 6.553 6.611 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 6.553 6.610 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.808 5.860 init_scf_run 11 5.9 0.000 0.001 3.883 3.884 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.883 3.884 sum_up_and_integrate 119 10.3 0.021 0.024 3.805 3.813 integrate_v_rspace 119 11.3 0.002 0.002 3.783 3.792 init_scf_loop 11 6.9 0.000 0.000 3.448 3.448 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.404 3.439 calculate_rho_elec 119 8.7 0.040 0.046 3.403 3.439 qs_ot_get_p 119 10.4 0.001 0.001 3.221 3.300 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.848 3.235 mp_waitall_1 145218 16.4 2.289 2.778 2.289 2.778 prepare_preconditioner 11 7.9 0.000 0.000 2.609 2.617 make_preconditioner 11 8.9 0.000 0.000 2.609 2.617 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.210 2.540 calculate_first_density_matrix 1 7.0 0.000 0.000 2.480 2.482 make_m2s 4572 13.5 0.054 0.056 2.273 2.366 make_images 4572 14.5 0.273 0.330 2.167 2.258 rs_pw_transfer 974 11.9 0.009 0.009 2.145 2.214 calculate_dm_sparse 119 9.5 0.000 0.000 2.142 2.192 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.040 2.163 apply_single 119 13.6 0.000 0.000 2.040 2.163 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.949 1.979 density_rs2pw 119 9.7 0.003 0.004 1.895 1.952 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.912 1.924 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.920 1.921 grid_integrate_task_list 119 12.3 1.819 1.906 1.819 1.906 jit_kernel_multiply 9 16.0 0.978 1.871 0.978 1.871 pw_transfer 1439 11.6 0.063 0.066 1.824 1.857 ot_diis_step 108 11.5 0.012 0.012 1.779 1.779 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.735 1.772 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.729 1.742 potential_pw2rs 119 12.3 0.009 0.009 1.599 1.606 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.593 1.594 acc_transpose_blocks 27432 15.5 0.111 0.114 1.482 1.513 fft3d_ps 1201 14.6 0.534 0.586 1.410 1.438 fft_wrap_pw1pw2_140 487 13.2 0.090 0.106 1.394 1.431 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.827 1.414 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.347 1.348 wfi_extrapolate 11 7.9 0.001 0.001 1.339 1.339 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.328 1.337 cp_fm_redistribute_end 50 14.0 0.889 1.324 0.890 1.324 mp_sum_l 7207 12.9 1.010 1.309 1.010 1.309 cp_fm_diag_elpa_base 50 14.0 0.412 1.259 0.432 1.298 grid_collocate_task_list 119 9.7 1.220 1.292 1.220 1.292 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.118 1.134 cp_fm_upper_to_full 72 13.5 0.808 1.121 0.808 1.121 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.116 1.116 dbcsr_complete_redistribute 329 12.2 0.130 0.152 0.825 1.099 multiply_cannon_sync_h2d 27432 15.5 0.995 1.060 0.995 1.060 make_images_data 4572 15.5 0.045 0.049 0.900 1.009 hybrid_alltoall_any 4725 16.4 0.062 0.152 0.782 0.944 mp_alltoall_d11v 2130 13.8 0.785 0.922 0.785 0.922 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.802 0.881 acc_transpose_blocks_kernels 27432 16.5 0.266 0.276 0.861 0.881 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.593 0.862 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.836 0.840 mp_alltoall_z22v 1201 16.6 0.758 0.786 0.758 0.786 cp_fm_cholesky_invert 11 10.9 0.779 0.782 0.779 0.782 mp_alltoall_i22 627 13.8 0.438 0.741 0.438 0.741 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.338000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=517.818182, yerr=3.638636 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 602.222592E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.027 28.371 28.371 qs_mol_dyn_low 1 2.0 0.003 0.003 28.208 28.215 qs_forces 11 3.9 0.002 0.002 28.097 28.099 qs_energies 11 4.9 0.001 0.001 26.361 26.364 scf_env_do_scf 11 5.9 0.000 0.001 21.514 21.514 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.973 18.973 velocity_verlet 10 3.0 0.021 0.025 14.548 14.553 dbcsr_multiply_generic 2286 12.5 0.092 0.095 12.485 12.615 qs_scf_new_mos 108 7.5 0.001 0.001 11.263 11.290 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.262 11.289 ot_scf_mini 108 9.5 0.002 0.002 10.572 10.599 multiply_cannon 2286 13.5 0.239 0.249 9.808 10.067 multiply_cannon_loop 2286 14.5 0.331 0.340 8.877 9.130 multiply_cannon_multrec 9144 15.5 1.596 1.816 5.773 6.002 ot_mini 108 10.5 0.001 0.001 5.964 5.996 rebuild_ks_matrix 119 8.3 0.000 0.000 5.901 5.922 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 5.901 5.921 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.245 5.264 qs_ot_get_derivative 108 11.5 0.001 0.001 4.655 4.683 dbcsr_mm_accdrv_process 12550 15.8 3.143 3.929 4.075 4.153 sum_up_and_integrate 119 10.3 0.023 0.024 3.610 3.614 integrate_v_rspace 119 11.3 0.002 0.003 3.587 3.591 init_scf_run 11 5.9 0.000 0.001 3.405 3.405 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.405 3.405 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.308 3.317 calculate_rho_elec 119 8.7 0.059 0.061 3.307 3.316 qs_ot_get_p 119 10.4 0.001 0.001 2.774 2.806 mp_waitall_1 121218 16.5 2.035 2.690 2.035 2.690 init_scf_loop 11 6.9 0.000 0.000 2.521 2.522 calculate_first_density_matrix 1 7.0 0.000 0.000 2.254 2.254 make_m2s 4572 13.5 0.034 0.035 1.829 1.974 grid_integrate_task_list 119 12.3 1.856 1.946 1.856 1.946 make_images 4572 14.5 0.267 0.303 1.740 1.884 rs_pw_transfer 974 11.9 0.008 0.008 1.787 1.865 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.791 1.793 density_rs2pw 119 9.7 0.003 0.004 1.701 1.791 prepare_preconditioner 11 7.9 0.000 0.000 1.781 1.787 make_preconditioner 11 8.9 0.000 0.000 1.781 1.787 calculate_dm_sparse 119 9.5 0.000 0.000 1.749 1.766 pw_transfer 1439 11.6 0.063 0.065 1.732 1.739 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.661 1.685 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.643 1.650 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.607 1.608 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.597 1.598 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.521 1.540 jit_kernel_multiply 8 15.6 0.895 1.466 0.895 1.466 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.390 1.396 potential_pw2rs 119 12.3 0.010 0.011 1.389 1.392 grid_collocate_task_list 119 9.7 1.274 1.360 1.274 1.360 fft_wrap_pw1pw2_140 487 13.2 0.085 0.089 1.345 1.355 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.321 1.321 fft3d_ps 1201 14.6 0.539 0.553 1.306 1.312 cp_fm_redistribute_end 50 14.0 0.656 1.298 0.656 1.298 ot_diis_step 108 11.5 0.012 0.013 1.296 1.296 cp_fm_diag_elpa_base 50 14.0 0.598 1.229 0.640 1.284 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.227 1.228 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.183 1.201 apply_single 119 13.6 0.000 0.000 1.183 1.201 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.175 1.188 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.149 1.155 wfi_extrapolate 11 7.9 0.001 0.001 1.107 1.107 hybrid_alltoall_any 4725 16.4 0.063 0.177 0.853 1.081 make_images_data 4572 15.5 0.039 0.043 0.897 1.075 mp_alltoall_d11v 2130 13.8 0.832 0.941 0.832 0.941 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.878 0.936 acc_transpose_blocks 9144 15.5 0.039 0.040 0.899 0.910 cp_fm_cholesky_invert 11 10.9 0.886 0.888 0.886 0.888 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.474 0.810 multiply_cannon_sync_h2d 9144 15.5 0.715 0.792 0.715 0.792 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.777 0.779 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.418 0.770 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.736 0.744 mp_allgather_i34 2286 14.5 0.255 0.733 0.255 0.733 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.673 0.725 mp_alltoall_z22v 1201 16.6 0.641 0.676 0.641 0.676 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 0.659 0.666 mp_sum_l 7207 12.9 0.416 0.643 0.416 0.643 dbcsr_complete_redistribute 329 12.2 0.203 0.222 0.599 0.630 qs_create_task_list 11 7.9 0.000 0.000 0.543 0.568 generate_qs_task_list 11 8.9 0.189 0.212 0.542 0.567 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.371000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=569.181818, yerr=6.779892 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 742.199296E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.034 40.297 40.297 qs_mol_dyn_low 1 2.0 0.003 0.004 40.071 40.078 qs_forces 11 3.9 0.002 0.002 40.009 40.010 qs_energies 11 4.9 0.001 0.001 38.048 38.051 scf_env_do_scf 11 5.9 0.001 0.001 32.526 32.526 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.648 24.649 velocity_verlet 10 3.0 0.002 0.002 22.890 22.896 dbcsr_multiply_generic 2286 12.5 0.101 0.102 17.272 17.520 qs_scf_new_mos 108 7.5 0.001 0.001 15.808 15.907 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.807 15.906 ot_scf_mini 108 9.5 0.002 0.002 14.735 14.836 multiply_cannon 2286 13.5 0.298 0.303 13.354 14.375 multiply_cannon_loop 2286 14.5 0.343 0.350 12.107 13.109 ot_mini 108 10.5 0.001 0.001 8.732 8.851 multiply_cannon_multrec 9144 15.5 3.370 4.670 8.474 8.580 init_scf_loop 11 6.9 0.000 0.000 7.840 7.841 rebuild_ks_matrix 119 8.3 0.000 0.000 6.846 6.993 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.846 6.993 prepare_preconditioner 11 7.9 0.000 0.000 6.904 6.918 make_preconditioner 11 8.9 0.000 0.000 6.904 6.918 qs_ot_get_derivative 108 11.5 0.001 0.001 6.693 6.793 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.443 6.791 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.195 6.328 dbcsr_mm_accdrv_process 12550 15.8 4.130 5.566 4.979 6.319 cp_fm_upper_to_full 72 14.2 3.174 4.574 3.174 4.574 sum_up_and_integrate 119 10.3 0.038 0.038 3.658 3.665 integrate_v_rspace 119 11.3 0.003 0.003 3.620 3.627 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.605 3.610 calculate_rho_elec 119 8.7 0.118 0.121 3.605 3.609 init_scf_run 11 5.9 0.000 0.001 3.438 3.439 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.438 3.439 mp_waitall_1 97218 16.6 2.481 3.436 2.481 3.436 qs_ot_get_p 119 10.4 0.001 0.001 3.263 3.405 dbcsr_complete_redistribute 329 12.2 0.295 0.300 2.021 2.865 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.414 2.861 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.701 2.543 make_m2s 4572 13.5 0.037 0.038 2.305 2.470 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.221 2.454 apply_single 119 13.6 0.000 0.000 2.221 2.454 make_images 4572 14.5 0.351 0.383 2.185 2.351 mp_alltoall_i22 627 13.8 1.437 2.323 1.437 2.323 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.456 2.291 calculate_dm_sparse 119 9.5 0.000 0.000 2.132 2.149 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.201 2.128 calculate_first_density_matrix 1 7.0 0.000 0.000 2.051 2.052 pw_transfer 1439 11.6 0.067 0.068 2.037 2.042 grid_integrate_task_list 119 12.3 2.007 2.012 2.007 2.012 ot_diis_step 108 11.5 0.014 0.014 2.008 2.008 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.987 1.988 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 1.942 1.946 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.819 1.819 mp_sum_l 7207 12.9 1.071 1.792 1.071 1.792 density_rs2pw 119 9.7 0.003 0.003 1.773 1.786 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.772 1.773 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.665 1.716 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.713 1.713 fft_wrap_pw1pw2_140 487 13.2 0.101 0.122 1.627 1.631 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.537 1.585 fft3d_ps 1201 14.6 0.566 0.576 1.561 1.577 cp_fm_cholesky_invert 11 10.9 1.506 1.509 1.506 1.509 rs_pw_transfer 974 11.9 0.009 0.009 1.461 1.494 grid_collocate_task_list 119 9.7 1.446 1.466 1.446 1.466 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.440 1.442 cp_fm_diag_elpa_base 50 14.0 1.292 1.347 1.438 1.440 hybrid_alltoall_any 4725 16.4 0.088 0.149 1.120 1.375 make_images_data 4572 15.5 0.043 0.046 1.125 1.335 wfi_extrapolate 11 7.9 0.001 0.001 1.334 1.334 potential_pw2rs 119 12.3 0.014 0.014 1.235 1.237 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.211 1.218 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.146 1.165 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.142 1.164 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.993 1.054 multiply_cannon_sync_h2d 9144 15.5 1.041 1.045 1.041 1.045 mp_alltoall_d11v 2130 13.8 1.008 1.021 1.008 1.021 jit_kernel_multiply 6 15.6 0.822 1.019 0.822 1.019 qs_create_task_list 11 7.9 0.000 0.000 0.942 0.955 generate_qs_task_list 11 8.9 0.372 0.392 0.942 0.955 acc_transpose_blocks 9144 15.5 0.038 0.039 0.899 0.901 mp_alltoall_z22v 1201 16.6 0.859 0.896 0.859 0.896 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.861 0.875 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.297000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=701.090909, yerr=10.273130 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 500.084736E+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 1926918. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.028 82.133 82.135 qs_mol_dyn_low 1 2.0 0.004 0.014 81.839 81.849 qs_forces 11 3.9 0.002 0.002 81.732 81.740 qs_energies 11 4.9 0.001 0.002 78.882 78.900 scf_env_do_scf 11 5.9 0.000 0.001 70.034 70.037 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.479 64.479 dbcsr_multiply_generic 2055 12.4 0.105 0.107 51.628 51.891 qs_scf_new_mos 99 7.5 0.000 0.001 47.457 47.581 qs_scf_loop_do_ot 99 8.5 0.000 0.001 47.457 47.581 ot_scf_mini 99 9.5 0.002 0.003 45.014 45.125 multiply_cannon 2055 13.4 0.175 0.184 42.198 43.218 multiply_cannon_loop 2055 14.4 1.549 1.593 41.239 42.310 velocity_verlet 10 3.0 0.002 0.006 41.671 41.672 ot_mini 99 10.5 0.001 0.006 27.284 27.392 qs_ot_get_derivative 99 11.5 0.001 0.001 20.453 20.548 multiply_cannon_multrec 49320 15.4 12.127 12.892 17.139 17.801 rebuild_ks_matrix 110 8.3 0.000 0.000 14.548 14.633 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.014 14.548 14.633 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.771 12.850 mp_waitall_1 241148 16.1 11.379 12.778 11.379 12.778 multiply_cannon_sync_h2d 49320 15.4 9.970 10.636 9.970 10.636 qs_ot_get_p 110 10.4 0.001 0.001 9.300 9.428 multiply_cannon_metrocomm3 49320 15.4 0.075 0.080 6.760 8.164 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.142 7.592 apply_single 110 13.6 0.000 0.001 7.142 7.591 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.830 7.394 sum_up_and_integrate 110 10.3 0.023 0.029 7.134 7.147 integrate_v_rspace 110 11.3 0.002 0.003 7.111 7.128 init_scf_run 11 5.9 0.000 0.001 6.824 6.824 scf_env_initial_rho_setup 11 6.9 0.000 0.004 6.824 6.824 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.640 6.728 ot_diis_step 99 11.5 0.011 0.051 6.547 6.548 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.204 6.312 calculate_rho_elec 110 8.6 0.020 0.024 6.203 6.312 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.036 6.063 mp_sum_l 6514 12.8 5.088 6.002 5.088 6.002 init_scf_loop 11 6.9 0.000 0.000 5.528 5.528 dbcsr_mm_accdrv_process 87628 16.1 2.075 2.206 4.892 5.236 cp_dbcsr_syevd 48 12.0 0.002 0.002 5.032 5.033 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.560 4.561 cp_fm_redistribute_end 48 14.0 3.973 4.535 3.977 4.536 cp_fm_diag_elpa_base 48 14.0 0.553 4.441 0.557 4.465 rs_pw_transfer 902 11.9 0.012 0.014 4.077 4.299 make_m2s 4110 13.4 0.060 0.064 3.887 4.002 wfi_extrapolate 11 7.9 0.001 0.003 3.995 3.995 make_images 4110 14.4 0.178 0.192 3.792 3.911 calculate_dm_sparse 110 9.5 0.000 0.001 3.810 3.892 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.626 3.630 multiply_cannon_metrocomm1 49320 15.4 0.060 0.065 2.556 3.630 density_rs2pw 110 9.6 0.004 0.004 3.317 3.477 grid_integrate_task_list 110 12.3 3.255 3.421 3.255 3.421 prepare_preconditioner 11 7.9 0.000 0.000 3.369 3.387 make_preconditioner 11 8.9 0.000 0.000 3.369 3.387 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.266 3.331 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.190 3.251 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.148 3.194 pw_transfer 1331 11.6 0.054 0.061 2.761 2.823 potential_pw2rs 110 12.3 0.006 0.006 2.755 2.774 calculate_first_density_matrix 1 7.0 0.000 0.000 2.745 2.748 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.670 2.734 jit_kernel_multiply 13 15.9 2.534 2.547 2.534 2.547 mp_alltoall_d11v 2046 13.8 2.063 2.480 2.063 2.480 fft_wrap_pw1pw2_140 451 13.1 0.170 0.189 2.254 2.326 acc_transpose_blocks 49320 15.4 0.214 0.226 2.196 2.286 fft3d_ps 1111 14.6 0.754 0.845 2.207 2.259 grid_collocate_task_list 110 9.6 2.089 2.189 2.089 2.189 mp_sum_d 3883 11.9 1.489 2.019 1.489 2.019 mp_waitany 14300 13.8 1.740 1.943 1.740 1.943 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.917 1.939 make_images_data 4110 15.4 0.043 0.046 1.786 1.927 hybrid_alltoall_any 4261 16.3 0.082 0.482 1.560 1.874 cp_fm_cholesky_invert 11 10.9 1.833 1.837 1.833 1.837 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.640 1.683 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.135000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=474.636364, yerr=2.267266 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 583.249920E+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 1414451. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.040 69.069 69.070 qs_mol_dyn_low 1 2.0 0.004 0.005 68.631 68.641 qs_forces 11 3.9 0.005 0.014 68.561 68.561 qs_energies 11 4.9 0.002 0.004 65.204 65.208 scf_env_do_scf 11 5.9 0.001 0.001 56.571 56.574 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.862 48.862 dbcsr_multiply_generic 2055 12.4 0.113 0.116 37.653 37.787 velocity_verlet 10 3.0 0.002 0.002 36.311 36.312 qs_scf_new_mos 99 7.5 0.001 0.001 32.764 32.891 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.763 32.891 multiply_cannon 2055 13.4 0.224 0.253 30.920 32.068 ot_scf_mini 99 9.5 0.003 0.004 31.105 31.212 multiply_cannon_loop 2055 14.4 0.929 0.952 29.621 30.630 ot_mini 99 10.5 0.001 0.002 18.224 18.346 multiply_cannon_multrec 24660 15.4 7.633 9.373 13.815 15.579 rebuild_ks_matrix 110 8.3 0.000 0.000 13.755 13.855 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.018 13.755 13.854 qs_ot_get_derivative 99 11.5 0.001 0.002 12.427 12.537 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.129 12.207 mp_waitall_1 186928 16.3 7.832 10.154 7.832 10.154 multiply_cannon_sync_h2d 24660 15.4 7.023 7.897 7.023 7.897 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.111 7.840 init_scf_loop 11 6.9 0.000 0.000 7.668 7.669 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.495 7.166 apply_single 110 13.6 0.000 0.001 6.495 7.166 sum_up_and_integrate 110 10.3 0.030 0.035 6.641 6.651 integrate_v_rspace 110 11.3 0.002 0.003 6.611 6.624 qs_ot_get_p 110 10.4 0.001 0.001 6.224 6.357 dbcsr_mm_accdrv_process 52282 16.1 4.681 5.418 6.021 6.272 init_scf_run 11 5.9 0.000 0.001 6.117 6.118 scf_env_initial_rho_setup 11 6.9 0.000 0.002 6.117 6.118 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.924 5.936 calculate_rho_elec 110 8.6 0.039 0.047 5.924 5.936 ot_diis_step 99 11.5 0.011 0.013 5.745 5.746 prepare_preconditioner 11 7.9 0.000 0.000 5.633 5.651 make_preconditioner 11 8.9 0.000 0.001 5.633 5.651 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.670 5.375 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.193 5.351 make_m2s 4110 13.4 0.056 0.060 4.287 4.746 make_images 4110 14.4 0.399 0.446 4.179 4.634 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.276 4.296 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.848 3.849 wfi_extrapolate 11 7.9 0.003 0.008 3.538 3.538 pw_transfer 1331 11.6 0.067 0.070 3.334 3.474 rs_pw_transfer 902 11.9 0.012 0.014 3.260 3.468 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.223 3.360 density_rs2pw 110 9.6 0.004 0.004 3.204 3.357 grid_integrate_task_list 110 12.3 3.151 3.295 3.151 3.295 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.285 3.287 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.220 3.284 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.274 3.277 cp_fm_redistribute_end 48 14.0 2.445 3.252 2.447 3.252 cp_fm_diag_elpa_base 48 14.0 0.770 3.124 0.802 3.214 calculate_dm_sparse 110 9.5 0.001 0.001 2.958 2.983 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.924 2.963 make_images_data 4110 15.4 0.048 0.052 2.384 2.864 hybrid_alltoall_any 4261 16.3 0.102 0.446 2.088 2.849 fft_wrap_pw1pw2_140 451 13.1 0.202 0.219 2.684 2.820 fft3d_ps 1111 14.6 1.071 1.266 2.578 2.706 cp_fm_cholesky_invert 11 10.9 2.611 2.619 2.611 2.619 potential_pw2rs 110 12.3 0.008 0.009 2.525 2.540 calculate_first_density_matrix 1 7.0 0.000 0.001 2.491 2.495 mp_sum_l 6514 12.8 1.839 2.472 1.839 2.472 grid_collocate_task_list 110 9.6 2.123 2.271 2.123 2.271 mp_alltoall_d11v 2046 13.8 1.827 2.000 1.827 2.000 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.964 1.981 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 1.948 1.949 jit_kernel_multiply 10 16.4 0.988 1.930 0.988 1.930 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.792 1.809 mp_allgather_i34 2055 14.4 0.625 1.672 0.625 1.672 multiply_cannon_metrocomm4 22605 15.4 0.073 0.078 0.789 1.575 acc_transpose_blocks 24660 15.4 0.106 0.109 1.528 1.572 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.535 1.547 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.386 1.499 mp_irecv_dv 57340 16.2 0.666 1.457 0.666 1.457 dbcsr_complete_redistribute 325 12.2 0.244 0.305 1.179 1.449 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.070000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=551.909091, yerr=5.899993 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 657.674240E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.055 60.252 60.254 qs_mol_dyn_low 1 2.0 0.003 0.004 59.875 59.884 qs_forces 11 3.9 0.002 0.002 59.731 59.732 qs_energies 11 4.9 0.002 0.006 56.523 56.533 scf_env_do_scf 11 5.9 0.001 0.001 48.567 48.567 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.858 39.859 velocity_verlet 10 3.0 0.001 0.002 32.757 32.763 dbcsr_multiply_generic 2055 12.4 0.108 0.114 28.551 28.835 qs_scf_new_mos 99 7.5 0.001 0.001 25.208 25.292 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.208 25.292 ot_scf_mini 99 9.5 0.002 0.003 23.980 24.086 multiply_cannon 2055 13.4 0.211 0.220 22.110 23.241 multiply_cannon_loop 2055 14.4 0.618 0.634 20.884 22.167 ot_mini 99 10.5 0.001 0.002 13.646 13.753 rebuild_ks_matrix 110 8.3 0.000 0.000 12.284 12.423 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.283 12.423 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.838 10.964 multiply_cannon_multrec 16440 15.4 3.976 5.010 9.719 10.697 mp_waitall_1 146766 16.3 7.205 10.208 7.205 10.208 qs_ot_get_derivative 99 11.5 0.001 0.001 9.156 9.259 init_scf_loop 11 6.9 0.000 0.000 8.670 8.671 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.256 7.169 prepare_preconditioner 11 7.9 0.000 0.000 6.894 6.909 make_preconditioner 11 8.9 0.000 0.001 6.894 6.909 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.253 6.597 sum_up_and_integrate 110 10.3 0.034 0.034 6.484 6.500 integrate_v_rspace 110 11.3 0.002 0.003 6.450 6.466 dbcsr_mm_accdrv_process 34862 16.1 4.835 5.391 5.598 5.776 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.598 5.608 calculate_rho_elec 110 8.6 0.058 0.058 5.597 5.608 qs_ot_get_p 110 10.4 0.001 0.001 5.482 5.600 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.999 5.463 init_scf_run 11 5.9 0.000 0.001 5.462 5.463 apply_single 110 13.6 0.000 0.000 4.999 5.463 scf_env_initial_rho_setup 11 6.9 0.000 0.002 5.462 5.462 make_m2s 4110 13.4 0.049 0.051 4.274 4.566 ot_diis_step 99 11.5 0.011 0.013 4.456 4.456 make_images 4110 14.4 0.393 0.508 4.159 4.450 multiply_cannon_sync_h2d 16440 15.4 3.716 4.273 3.716 4.273 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.811 3.816 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.085 3.715 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.468 3.469 grid_integrate_task_list 110 12.3 3.190 3.395 3.190 3.395 pw_transfer 1331 11.6 0.066 0.073 3.146 3.153 rs_pw_transfer 902 11.9 0.010 0.012 2.953 3.149 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.034 3.044 density_rs2pw 110 9.6 0.004 0.004 2.889 3.040 wfi_extrapolate 11 7.9 0.001 0.001 2.969 2.969 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.923 2.925 cp_fm_redistribute_end 48 14.0 1.815 2.890 1.817 2.891 make_images_data 4110 15.4 0.044 0.048 2.494 2.875 cp_fm_diag_elpa_base 48 14.0 1.011 2.751 1.069 2.856 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.812 2.814 hybrid_alltoall_any 4261 16.3 0.106 0.374 2.280 2.778 cp_fm_cholesky_invert 11 10.9 2.762 2.769 2.762 2.769 fft_wrap_pw1pw2_140 451 13.1 0.214 0.217 2.576 2.587 calculate_dm_sparse 110 9.5 0.001 0.001 2.499 2.539 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.477 2.523 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.393 2.453 calculate_first_density_matrix 1 7.0 0.000 0.000 2.410 2.411 potential_pw2rs 110 12.3 0.011 0.011 2.351 2.363 fft3d_ps 1111 14.6 1.062 1.073 2.345 2.354 grid_collocate_task_list 110 9.6 2.177 2.353 2.177 2.353 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.860 2.336 mp_irecv_dv 48980 15.7 0.791 2.214 0.791 2.214 mp_alltoall_d11v 2046 13.8 1.779 2.168 1.779 2.168 mp_sum_l 6514 12.8 1.552 2.137 1.552 2.137 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 2.036 2.036 dbcsr_complete_redistribute 325 12.2 0.324 0.368 1.416 1.879 cp_fm_upper_to_full 70 13.6 1.384 1.853 1.384 1.853 mp_allgather_i34 2055 14.4 0.549 1.706 0.549 1.706 cp_fm_cholesky_decompose 22 10.9 1.658 1.679 1.658 1.679 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.646 1.660 jit_kernel_multiply 8 16.7 0.371 1.516 0.371 1.516 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.376 1.494 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.455 1.467 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.963 1.417 mp_waitany 17072 13.8 1.147 1.336 1.147 1.336 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.280 1.288 acc_transpose_blocks 16440 15.4 0.072 0.074 1.224 1.234 rs_gather_matrices 110 12.3 0.137 0.150 0.857 1.214 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.254000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.181818, yerr=8.451123 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 734.380032E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.042 65.013 65.015 qs_mol_dyn_low 1 2.0 0.003 0.003 64.730 64.739 qs_forces 11 3.9 0.002 0.004 64.507 64.508 qs_energies 11 4.9 0.003 0.018 61.050 61.053 scf_env_do_scf 11 5.9 0.001 0.001 52.709 52.712 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.812 40.813 velocity_verlet 10 3.0 0.002 0.002 37.151 37.160 dbcsr_multiply_generic 2055 12.4 0.115 0.117 29.433 29.624 qs_scf_new_mos 99 7.5 0.001 0.001 26.256 26.368 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.255 26.367 ot_scf_mini 99 9.5 0.003 0.003 24.631 24.722 multiply_cannon 2055 13.4 0.242 0.260 22.472 23.499 multiply_cannon_loop 2055 14.4 0.881 0.901 20.941 21.546 ot_mini 99 10.5 0.001 0.002 14.058 14.170 multiply_cannon_multrec 24660 15.4 4.206 6.930 12.624 13.676 rebuild_ks_matrix 110 8.3 0.000 0.000 12.051 12.148 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.017 12.051 12.147 init_scf_loop 11 6.9 0.000 0.000 11.856 11.856 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.659 10.743 prepare_preconditioner 11 7.9 0.000 0.000 10.121 10.137 make_preconditioner 11 8.9 0.000 0.001 10.121 10.137 qs_ot_get_derivative 99 11.5 0.001 0.002 9.896 9.989 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.344 9.804 dbcsr_mm_accdrv_process 52304 16.0 7.000 8.478 8.270 9.219 mp_waitall_1 126806 16.4 4.390 6.495 4.390 6.495 sum_up_and_integrate 110 10.3 0.039 0.041 6.399 6.414 integrate_v_rspace 110 11.3 0.003 0.003 6.360 6.374 make_m2s 4110 13.4 0.059 0.060 5.349 5.769 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.687 5.695 calculate_rho_elec 110 8.6 0.077 0.081 5.686 5.695 qs_ot_get_p 110 10.4 0.001 0.001 5.537 5.661 make_images 4110 14.4 0.578 0.701 5.210 5.626 init_scf_run 11 5.9 0.000 0.001 5.518 5.518 scf_env_initial_rho_setup 11 6.9 0.000 0.003 5.517 5.518 cp_fm_upper_to_full 70 13.8 3.300 4.692 3.300 4.692 ot_diis_step 99 11.5 0.012 0.016 4.123 4.124 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.022 4.088 apply_single 110 13.6 0.000 0.000 4.022 4.087 dbcsr_complete_redistribute 325 12.2 0.428 0.472 2.693 3.858 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.747 3.760 grid_integrate_task_list 110 12.3 3.280 3.425 3.280 3.425 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.426 3.401 multiply_cannon_sync_h2d 24660 15.4 3.181 3.350 3.181 3.350 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.288 3.330 pw_transfer 1331 11.6 0.066 0.073 3.283 3.310 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.129 3.279 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.272 3.273 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.173 3.202 hybrid_alltoall_any 4261 16.3 0.120 0.456 2.324 3.111 make_images_data 4110 15.4 0.046 0.051 2.676 3.107 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.035 3.086 density_rs2pw 110 9.6 0.004 0.004 2.845 3.014 wfi_extrapolate 11 7.9 0.001 0.001 2.969 2.970 calculate_dm_sparse 110 9.5 0.001 0.001 2.907 2.938 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.768 2.892 mp_alltoall_i22 605 13.7 1.653 2.859 1.653 2.859 cp_fm_cholesky_invert 11 10.9 2.822 2.830 2.822 2.830 rs_pw_transfer 902 11.9 0.010 0.011 2.587 2.816 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.796 2.797 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.725 2.727 fft_wrap_pw1pw2_140 451 13.1 0.206 0.215 2.682 2.711 cp_fm_redistribute_end 48 14.0 1.350 2.679 1.351 2.680 cp_fm_diag_elpa_base 48 14.0 1.248 2.547 1.326 2.657 fft3d_ps 1111 14.6 1.061 1.093 2.471 2.492 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.436 2.476 calculate_first_density_matrix 1 7.0 0.000 0.001 2.459 2.462 grid_collocate_task_list 110 9.6 2.222 2.331 2.222 2.331 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.275 2.276 potential_pw2rs 110 12.3 0.013 0.013 2.195 2.206 jit_kernel_multiply 11 16.0 0.937 2.015 0.937 2.015 mp_alltoall_d11v 2046 13.8 1.812 1.966 1.812 1.966 cp_fm_cholesky_decompose 22 10.9 1.729 1.778 1.729 1.778 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.717 1.751 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.625 1.724 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.637 1.648 mp_allgather_i34 2055 14.4 0.630 1.600 0.630 1.600 mp_sum_l 6514 12.8 0.930 1.572 0.930 1.572 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.523 1.538 acc_transpose_blocks 24660 15.4 0.104 0.107 1.498 1.518 multiply_cannon_metrocomm4 20550 15.4 0.057 0.061 0.842 1.457 mp_irecv_dv 62702 16.1 0.744 1.376 0.744 1.376 mp_waitany 13376 13.8 1.044 1.342 1.044 1.342 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.209 1.305 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.015000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=694.272727, yerr=10.198849 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 816.877568E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.030 54.596 54.597 qs_mol_dyn_low 1 2.0 0.003 0.003 54.357 54.365 qs_forces 11 3.9 0.002 0.002 54.291 54.292 qs_energies 11 4.9 0.002 0.002 50.676 50.680 scf_env_do_scf 11 5.9 0.000 0.001 42.226 42.226 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 34.510 34.510 velocity_verlet 10 3.0 0.002 0.002 30.455 30.457 dbcsr_multiply_generic 2055 12.4 0.104 0.105 22.652 22.773 qs_scf_new_mos 99 7.5 0.001 0.001 20.488 20.530 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.487 20.530 ot_scf_mini 99 9.5 0.002 0.002 19.254 19.268 multiply_cannon 2055 13.4 0.249 0.268 17.285 18.578 multiply_cannon_loop 2055 14.4 0.322 0.336 15.952 16.272 rebuild_ks_matrix 110 8.3 0.000 0.000 11.406 11.424 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.406 11.423 ot_mini 99 10.5 0.001 0.001 10.380 10.390 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.152 10.168 multiply_cannon_multrec 8220 15.4 3.225 4.495 7.435 8.403 init_scf_loop 11 6.9 0.000 0.000 7.667 7.669 mp_waitall_1 106626 16.5 6.082 7.657 6.082 7.657 qs_ot_get_derivative 99 11.5 0.001 0.001 6.587 6.599 sum_up_and_integrate 110 10.3 0.048 0.048 6.173 6.184 integrate_v_rspace 110 11.3 0.003 0.003 6.125 6.136 prepare_preconditioner 11 7.9 0.000 0.000 6.059 6.063 make_preconditioner 11 8.9 0.000 0.000 6.059 6.063 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.626 5.707 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.560 5.570 calculate_rho_elec 110 8.6 0.115 0.116 5.559 5.569 init_scf_run 11 5.9 0.000 0.001 5.152 5.152 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.152 5.152 qs_ot_get_p 110 10.4 0.001 0.001 5.017 5.032 dbcsr_mm_accdrv_process 17442 15.9 2.835 3.759 4.081 4.986 make_m2s 4110 13.4 0.038 0.039 4.074 4.321 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.024 4.281 make_images 4110 14.4 0.637 0.700 3.945 4.191 ot_diis_step 99 11.5 0.012 0.012 3.768 3.768 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.677 3.699 apply_single 110 13.6 0.000 0.000 3.677 3.698 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.650 3.654 grid_integrate_task_list 110 12.3 3.366 3.539 3.366 3.539 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.342 3.342 pw_transfer 1331 11.6 0.066 0.071 3.112 3.133 multiply_cannon_sync_h2d 8220 15.4 2.912 3.028 2.912 3.028 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.001 3.027 cp_fm_cholesky_invert 11 10.9 2.930 2.934 2.930 2.934 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.794 2.794 qs_energies_init_hamiltonians 11 5.9 0.038 0.114 2.787 2.788 cp_fm_redistribute_end 48 14.0 0.709 2.762 0.713 2.763 density_rs2pw 110 9.6 0.004 0.004 2.646 2.738 cp_fm_diag_elpa_base 48 14.0 1.867 2.568 2.042 2.728 make_images_data 4110 15.4 0.038 0.043 2.279 2.674 wfi_extrapolate 11 7.9 0.001 0.001 2.669 2.669 hybrid_alltoall_any 4261 16.3 0.200 0.865 2.216 2.593 fft_wrap_pw1pw2_140 451 13.1 0.213 0.216 2.560 2.589 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.537 2.539 calculate_dm_sparse 110 9.5 0.001 0.001 2.469 2.511 grid_collocate_task_list 110 9.6 2.330 2.501 2.330 2.501 rs_pw_transfer 902 11.9 0.010 0.010 2.345 2.461 calculate_first_density_matrix 1 7.0 0.000 0.000 2.396 2.398 fft3d_ps 1111 14.6 1.113 1.147 2.258 2.285 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.146 2.153 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.789 2.001 potential_pw2rs 110 12.3 0.015 0.016 1.989 2.000 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.882 1.890 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.812 1.818 mp_alltoall_d11v 2046 13.8 1.557 1.781 1.557 1.781 cp_fm_cholesky_decompose 22 10.9 1.683 1.705 1.683 1.705 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.499 1.616 dbcsr_complete_redistribute 325 12.2 0.600 0.634 1.494 1.599 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.586 1.589 mp_allgather_i34 2055 14.4 0.518 1.497 0.518 1.497 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.437 1.449 qs_create_task_list 11 7.9 0.001 0.001 1.226 1.326 generate_qs_task_list 11 8.9 0.380 0.448 1.226 1.325 jit_kernel_multiply 7 15.8 0.933 1.196 0.933 1.196 mp_waitany 9240 13.8 1.031 1.152 1.031 1.152 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.128 1.143 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.482 1.126 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.597000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=770.545455, yerr=10.439119 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 1.191600E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.046 86.094 86.094 qs_mol_dyn_low 1 2.0 0.003 0.004 85.733 85.741 qs_forces 11 3.9 0.002 0.002 85.662 85.665 qs_energies 11 4.9 0.001 0.002 81.601 81.604 scf_env_do_scf 11 5.9 0.001 0.001 71.468 71.469 velocity_verlet 10 3.0 0.002 0.002 54.922 54.928 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.869 42.870 dbcsr_multiply_generic 2055 12.4 0.119 0.121 29.188 29.209 init_scf_loop 11 6.9 0.000 0.000 28.521 28.522 qs_scf_new_mos 99 7.5 0.001 0.001 26.570 26.599 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.569 26.598 prepare_preconditioner 11 7.9 0.000 0.000 26.579 26.585 make_preconditioner 11 8.9 0.000 0.000 26.579 26.585 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.772 26.037 ot_scf_mini 99 9.5 0.002 0.002 24.832 24.862 multiply_cannon 2055 13.4 0.345 0.386 22.035 22.863 multiply_cannon_loop 2055 14.4 0.341 0.345 20.272 20.694 cp_fm_upper_to_full 70 14.2 12.698 18.212 12.698 18.212 ot_mini 99 10.5 0.001 0.001 13.896 13.913 rebuild_ks_matrix 110 8.3 0.000 0.001 13.179 13.205 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.178 13.205 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.980 12.004 dbcsr_complete_redistribute 325 12.2 1.050 1.069 7.427 10.563 multiply_cannon_multrec 8220 15.4 4.597 4.863 10.015 10.115 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.371 9.513 qs_ot_get_derivative 99 11.5 0.001 0.001 9.313 9.341 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.791 8.913 mp_waitall_1 87304 16.6 8.035 8.868 8.035 8.868 mp_alltoall_i22 605 13.7 5.401 8.541 5.401 8.541 sum_up_and_integrate 110 10.3 0.090 0.091 6.560 6.576 integrate_v_rspace 110 11.3 0.003 0.003 6.470 6.487 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.356 6.395 calculate_rho_elec 110 8.6 0.227 0.227 6.355 6.394 init_scf_run 11 5.9 0.000 0.001 5.872 5.872 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.871 5.872 make_m2s 4110 13.4 0.042 0.043 5.375 5.862 qs_ot_get_p 110 10.4 0.001 0.001 5.730 5.750 make_images 4110 14.4 0.876 0.916 5.189 5.675 dbcsr_mm_accdrv_process 11614 15.7 3.176 3.642 5.274 5.502 cp_fm_cholesky_invert 11 10.9 5.480 5.485 5.480 5.485 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.838 5.375 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.684 5.122 apply_single 110 13.6 0.000 0.000 4.684 5.122 ot_diis_step 99 11.5 0.015 0.016 4.557 4.558 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.149 4.156 multiply_cannon_sync_h2d 8220 15.4 3.948 3.953 3.948 3.953 pw_transfer 1331 11.6 0.073 0.073 3.731 3.739 grid_integrate_task_list 110 12.3 3.659 3.708 3.659 3.708 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.703 3.703 hybrid_alltoall_any 4261 16.3 0.257 0.554 2.937 3.699 make_images_data 4110 15.4 0.042 0.045 3.024 3.656 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 3.649 3.651 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.607 3.615 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.934 3.380 wfi_extrapolate 11 7.9 0.001 0.001 3.244 3.245 calculate_dm_sparse 110 9.5 0.001 0.001 3.207 3.233 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.144 3.145 cp_fm_diag_elpa_base 48 14.0 2.595 2.796 3.142 3.143 fft_wrap_pw1pw2_140 451 13.1 0.222 0.223 3.128 3.135 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.033 3.036 density_rs2pw 110 9.6 0.004 0.004 3.002 3.023 fft3d_ps 1111 14.6 1.259 1.273 2.810 2.819 grid_collocate_task_list 110 9.6 2.624 2.649 2.624 2.649 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.543 2.554 calculate_first_density_matrix 1 7.0 0.000 0.000 2.521 2.521 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.379 2.387 rs_pw_transfer 902 11.9 0.011 0.011 2.197 2.237 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.173 2.231 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.103 2.194 potential_pw2rs 110 12.3 0.021 0.021 2.066 2.071 cp_fm_cholesky_decompose 22 10.9 2.058 2.070 2.058 2.070 jit_kernel_multiply 10 15.5 1.894 2.014 1.894 2.014 qs_create_task_list 11 7.9 0.000 0.001 1.898 1.943 generate_qs_task_list 11 8.9 0.738 0.790 1.898 1.942 mp_alltoall_d11v 2046 13.8 1.803 1.850 1.803 1.850 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.802 1.806 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.769 1.784 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.721 1.733 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.094000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1112.818182, yerr=24.154119 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 626.249728E+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 4085 56760. MP_Allreduce 11253 785. MP_Sync 170 MP_Alltoall 2226 2855966. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.119 202.209 202.210 qs_mol_dyn_low 1 2.0 0.003 0.004 201.662 201.674 qs_forces 11 3.9 0.003 0.004 201.577 201.578 qs_energies 11 4.9 0.001 0.003 196.039 196.055 scf_env_do_scf 11 5.9 0.001 0.001 179.623 179.627 scf_env_do_scf_inner_loop 117 6.6 0.002 0.007 159.262 159.264 dbcsr_multiply_generic 2507 12.6 0.175 0.179 123.629 124.562 velocity_verlet 10 3.0 0.001 0.002 121.597 121.598 qs_scf_new_mos 117 7.6 0.001 0.001 120.268 120.584 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.267 120.583 ot_scf_mini 117 9.6 0.003 0.004 113.695 113.978 multiply_cannon 2507 13.6 0.237 0.244 100.538 103.081 multiply_cannon_loop 2507 14.6 2.105 2.168 98.341 100.365 ot_mini 117 10.6 0.001 0.002 65.336 65.675 multiply_cannon_multrec 60168 15.6 33.116 35.749 41.513 43.356 qs_ot_get_derivative 117 11.6 0.001 0.002 40.470 40.745 rebuild_ks_matrix 128 8.3 0.001 0.001 33.564 33.813 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.021 33.563 33.813 mp_waitall_1 291448 16.2 28.069 31.910 28.069 31.910 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.191 30.413 multiply_cannon_sync_h2d 60168 15.6 27.335 30.132 27.335 30.132 qs_ot_get_p 128 10.4 0.001 0.001 26.593 26.928 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.337 24.992 apply_single 128 13.6 0.001 0.001 24.337 24.992 ot_diis_step 117 11.6 0.008 0.011 24.526 24.527 init_scf_loop 11 6.9 0.000 0.000 20.286 20.288 qs_ot_p2m_diag 83 11.4 0.077 0.091 19.956 20.028 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.893 19.102 multiply_cannon_metrocomm3 60168 15.6 0.109 0.115 15.521 17.920 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.384 17.385 prepare_preconditioner 11 7.9 0.000 0.000 15.700 15.738 make_preconditioner 11 8.9 0.000 0.000 15.700 15.738 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.942 15.118 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.613 14.617 cp_fm_redistribute_end 83 14.4 11.545 14.548 11.556 14.550 cp_fm_diag_elpa_base 83 14.4 2.953 14.263 2.984 14.378 sum_up_and_integrate 128 10.3 0.052 0.064 14.158 14.176 integrate_v_rspace 128 11.3 0.003 0.004 14.105 14.125 make_m2s 5014 13.6 0.102 0.109 13.680 14.013 make_images 5014 14.6 0.403 0.423 13.501 13.846 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.705 12.800 calculate_rho_elec 128 8.7 0.045 0.064 12.705 12.799 init_scf_run 11 5.9 0.000 0.001 12.208 12.209 scf_env_initial_rho_setup 11 6.9 0.000 0.002 12.208 12.208 mp_sum_l 7870 13.0 8.177 9.319 8.177 9.319 wfi_extrapolate 11 7.9 0.001 0.003 8.982 8.982 cp_fm_cholesky_invert 11 10.9 8.937 8.945 8.937 8.945 dbcsr_mm_accdrv_process 124484 16.2 3.319 3.503 7.962 8.496 calculate_dm_sparse 128 9.5 0.001 0.001 8.367 8.485 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.813 7.962 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.769 7.886 multiply_cannon_metrocomm1 60168 15.6 0.087 0.091 5.880 7.839 make_images_data 5014 15.6 0.069 0.075 6.760 7.690 grid_integrate_task_list 128 12.3 7.085 7.490 7.085 7.490 hybrid_alltoall_any 5200 16.5 0.290 2.251 5.919 7.365 density_rs2pw 128 9.7 0.006 0.007 6.522 7.084 pw_transfer 1547 11.6 0.075 0.104 6.666 6.969 rs_pw_transfer 1046 11.9 0.017 0.019 6.225 6.870 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.452 6.735 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.659 6.668 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.826 5.934 fft_wrap_pw1pw2_140 523 13.2 0.438 0.492 5.608 5.788 fft3d_ps 1291 14.7 2.094 2.601 5.252 5.495 mp_alltoall_d11v 2415 14.1 4.124 5.147 4.124 5.147 grid_collocate_task_list 128 9.7 4.723 5.132 4.723 5.132 potential_pw2rs 128 12.3 0.009 0.011 4.729 4.749 cp_fm_cholesky_decompose 22 10.9 4.591 4.606 4.591 4.606 mp_sum_d 4464 12.1 3.558 4.330 3.558 4.330 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=202.210000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=593.363636, yerr=6.984043 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 825.200640E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57341. MP_Allreduce 11227 947. MP_Sync 170 MP_Alltoall 1969 6320497. MP_ISendRecv 24064 47072. MP_Wait 37948 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.114 0.255 189.238 189.239 qs_mol_dyn_low 1 2.0 0.003 0.005 188.601 188.634 qs_forces 11 3.9 0.003 0.006 188.514 188.515 qs_energies 11 4.9 0.002 0.005 181.745 181.755 scf_env_do_scf 11 5.9 0.001 0.001 165.446 165.455 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 132.744 132.746 velocity_verlet 10 3.0 0.002 0.002 118.970 118.971 dbcsr_multiply_generic 2507 12.6 0.184 0.188 97.220 98.302 qs_scf_new_mos 117 7.6 0.001 0.001 93.946 94.365 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.945 94.365 ot_scf_mini 117 9.6 0.004 0.005 89.163 89.647 multiply_cannon 2507 13.6 0.475 0.532 77.195 81.862 multiply_cannon_loop 2507 14.6 1.256 1.292 74.012 76.401 ot_mini 117 10.6 0.001 0.001 49.370 49.838 mp_waitall_1 226760 16.4 24.553 37.977 24.553 37.977 multiply_cannon_multrec 30084 15.6 22.206 26.882 31.820 36.700 rebuild_ks_matrix 128 8.3 0.001 0.001 33.012 33.550 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.022 33.012 33.550 init_scf_loop 11 6.9 0.000 0.000 32.613 32.614 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.747 30.241 multiply_cannon_metrocomm3 30084 15.6 0.096 0.100 15.365 28.406 prepare_preconditioner 11 7.9 0.000 0.000 28.207 28.254 make_preconditioner 11 8.9 0.000 0.001 28.207 28.254 qs_ot_get_derivative 117 11.6 0.001 0.002 27.535 28.011 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.899 27.435 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.966 23.062 apply_single 128 13.6 0.001 0.001 21.966 23.061 qs_ot_get_p 128 10.4 0.001 0.001 21.520 22.126 ot_diis_step 117 11.6 0.014 0.018 21.666 21.668 multiply_cannon_sync_h2d 30084 15.6 19.417 21.644 19.417 21.644 qs_ot_p2m_diag 83 11.4 0.188 0.226 16.782 16.825 cp_fm_cholesky_invert 11 10.9 16.458 16.470 16.458 16.470 make_m2s 5014 13.6 0.088 0.095 14.402 15.955 make_images 5014 14.6 1.159 1.342 14.196 15.751 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.648 15.649 sum_up_and_integrate 128 10.3 0.062 0.072 14.322 14.357 integrate_v_rspace 128 11.3 0.003 0.004 14.260 14.298 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.936 12.970 calculate_rho_elec 128 8.7 0.088 0.105 12.935 12.969 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.614 12.621 cp_fm_redistribute_end 83 14.4 7.385 12.556 7.397 12.557 cp_fm_diag_elpa_base 83 14.4 4.925 12.000 5.144 12.464 init_scf_run 11 5.9 0.000 0.001 11.421 11.422 scf_env_initial_rho_setup 11 6.9 0.000 0.002 11.421 11.422 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.805 11.159 make_images_data 5014 15.6 0.067 0.074 8.853 10.760 multiply_cannon_metrocomm4 27577 15.6 0.096 0.111 3.701 10.369 hybrid_alltoall_any 5200 16.5 0.342 1.499 7.482 10.105 mp_irecv_dv 69486 16.3 3.509 9.996 3.509 9.996 dbcsr_mm_accdrv_process 62242 16.2 4.421 5.092 9.074 9.681 wfi_extrapolate 11 7.9 0.001 0.003 8.325 8.325 pw_transfer 1547 11.6 0.083 0.101 7.658 7.716 grid_integrate_task_list 128 12.3 7.152 7.525 7.152 7.525 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.429 7.486 density_rs2pw 128 9.7 0.006 0.006 6.736 7.303 cp_fm_cholesky_decompose 22 10.9 6.906 6.985 6.906 6.985 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.136 6.874 fft_wrap_pw1pw2_140 523 13.2 0.461 0.499 6.531 6.584 calculate_dm_sparse 128 9.5 0.001 0.001 6.435 6.561 rs_pw_transfer 1046 11.9 0.014 0.017 5.794 6.350 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.141 6.152 fft3d_ps 1291 14.7 2.779 2.931 5.801 5.842 mp_sum_l 7870 13.0 3.973 5.766 3.973 5.766 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.314 5.434 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.296 5.363 grid_collocate_task_list 128 9.7 4.910 5.275 4.910 5.275 mp_allgather_i34 2507 14.6 1.756 5.188 1.756 5.188 potential_pw2rs 128 12.3 0.016 0.018 4.839 4.858 mp_alltoall_d11v 2415 14.1 4.081 4.613 4.081 4.613 mp_sum_d 4459 12.1 2.595 3.901 2.595 3.901 dbcsr_complete_redistribute 395 12.7 0.781 0.862 3.070 3.892 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.239000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=786.090909, yerr=2.874798 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/17/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 2.928533E+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 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 929.529856E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+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 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56951. MP_Allreduce 11296 984. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.049 0.109 177.617 177.618 qs_mol_dyn_low 1 2.0 0.003 0.004 177.145 177.158 qs_forces 11 3.9 0.003 0.003 177.048 177.052 qs_energies 11 4.9 0.003 0.008 170.487 170.496 scf_env_do_scf 11 5.9 0.001 0.002 154.801 154.804 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 119.584 119.585 velocity_verlet 10 3.0 0.002 0.002 113.642 113.644 dbcsr_multiply_generic 2507 12.6 0.182 0.187 82.284 83.466 qs_scf_new_mos 117 7.6 0.001 0.001 82.542 82.881 qs_scf_loop_do_ot 117 8.6 0.001 0.001 82.542 82.880 ot_scf_mini 117 9.6 0.004 0.004 78.387 78.748 multiply_cannon 2507 13.6 0.496 0.517 61.960 66.349 multiply_cannon_loop 2507 14.6 0.862 0.891 58.852 61.565 ot_mini 117 10.6 0.001 0.002 42.949 43.330 mp_waitall_1 178456 16.5 26.192 35.273 26.192 35.273 init_scf_loop 11 6.9 0.000 0.000 35.113 35.114 prepare_preconditioner 11 7.9 0.000 0.000 31.098 31.152 make_preconditioner 11 8.9 0.000 0.001 31.098 31.152 rebuild_ks_matrix 128 8.3 0.001 0.001 30.465 30.913 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 30.465 30.913 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.725 30.098 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.505 27.920 multiply_cannon_metrocomm3 20056 15.6 0.058 0.062 15.724 24.799 multiply_cannon_multrec 20056 15.6 13.341 16.056 22.015 24.788 qs_ot_get_derivative 117 11.6 0.001 0.002 22.933 23.300 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.125 21.153 apply_single 128 13.6 0.001 0.001 20.125 21.152 qs_ot_get_p 128 10.4 0.001 0.002 20.648 21.084 ot_diis_step 117 11.6 0.018 0.021 19.903 19.904 qs_ot_p2m_diag 83 11.4 0.265 0.272 16.374 16.391 make_m2s 5014 13.6 0.080 0.085 15.209 16.130 make_images 5014 14.6 1.186 1.291 14.978 15.897 multiply_cannon_sync_h2d 20056 15.6 14.110 15.744 14.110 15.744 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.339 15.341 cp_fm_cholesky_invert 11 10.9 14.605 14.613 14.605 14.613 sum_up_and_integrate 128 10.3 0.072 0.076 14.257 14.282 integrate_v_rspace 128 11.3 0.003 0.003 14.185 14.210 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.190 13.223 calculate_rho_elec 128 8.7 0.132 0.146 13.189 13.223 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.147 12.148 cp_fm_redistribute_end 83 14.4 4.591 12.068 4.606 12.070 cp_fm_diag_elpa_base 83 14.4 7.032 11.447 7.444 11.960 make_images_data 5014 15.6 0.061 0.069 9.488 10.905 init_scf_run 11 5.9 0.000 0.001 10.561 10.561 scf_env_initial_rho_setup 11 6.9 0.001 0.006 10.561 10.561 hybrid_alltoall_any 5200 16.5 0.431 1.967 8.153 9.640 multiply_cannon_metrocomm4 17549 15.6 0.063 0.072 3.526 9.521 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.012 9.275 mp_irecv_dv 50230 16.2 3.402 9.271 3.402 9.271 dbcsr_mm_accdrv_process 41502 16.2 4.303 4.917 8.141 8.255 pw_transfer 1547 11.6 0.083 0.102 7.794 7.907 grid_integrate_task_list 128 12.3 7.285 7.758 7.285 7.758 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.569 7.688 cp_fm_upper_to_full 105 14.5 5.759 7.463 5.759 7.463 wfi_extrapolate 11 7.9 0.001 0.001 7.415 7.415 cp_fm_cholesky_decompose 22 10.9 7.277 7.314 7.277 7.314 density_rs2pw 128 9.7 0.005 0.006 6.644 7.060 fft_wrap_pw1pw2_140 523 13.2 0.470 0.510 6.630 6.752 dbcsr_complete_redistribute 395 12.7 1.196 1.234 4.690 6.463 fft3d_ps 1291 14.7 2.698 2.911 5.844 5.928 rs_pw_transfer 1046 11.9 0.013 0.014 5.417 5.863 calculate_dm_sparse 128 9.5 0.001 0.001 5.763 5.859 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.530 5.534 grid_collocate_task_list 128 9.7 5.072 5.482 5.072 5.482 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.547 5.248 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.465 5.236 mp_alltoall_d11v 2415 14.1 4.382 4.976 4.382 4.976 mp_allgather_i34 2507 14.6 1.632 4.874 1.632 4.874 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.593 4.748 potential_pw2rs 128 12.3 0.021 0.023 4.676 4.687 mp_sum_l 7870 13.0 3.320 4.682 3.320 4.682 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.354 4.101 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.067 4.098 mp_alltoall_i22 716 14.1 1.959 3.911 1.959 3.911 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 3.829 3.830 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.618000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=882.636364, yerr=8.657868 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 1.148670E+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_ISendRecv 11776 75008. MP_Wait 28330 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.076 188.888 188.889 qs_mol_dyn_low 1 2.0 0.003 0.004 188.490 188.502 qs_forces 11 3.9 0.003 0.003 188.379 188.392 qs_energies 11 4.9 0.002 0.003 181.283 181.296 scf_env_do_scf 11 5.9 0.001 0.001 164.273 164.286 velocity_verlet 10 3.0 0.002 0.002 125.408 125.411 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 117.114 117.117 qs_scf_new_mos 117 7.6 0.001 0.001 81.041 81.352 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.040 81.352 dbcsr_multiply_generic 2507 12.6 0.190 0.194 79.954 80.651 ot_scf_mini 117 9.6 0.003 0.004 76.508 76.815 multiply_cannon 2507 13.6 0.559 0.586 54.836 58.388 multiply_cannon_loop 2507 14.6 1.180 1.210 51.094 52.872 init_scf_loop 11 6.9 0.000 0.000 47.023 47.027 prepare_preconditioner 11 7.9 0.000 0.000 42.912 42.942 make_preconditioner 11 8.9 0.000 0.001 42.912 42.942 ot_mini 117 10.6 0.001 0.001 42.570 42.889 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.296 41.340 multiply_cannon_multrec 30084 15.6 14.044 19.697 25.943 30.985 rebuild_ks_matrix 128 8.3 0.001 0.001 29.423 29.746 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 29.423 29.746 mp_waitall_1 153770 16.5 18.193 27.586 18.193 27.586 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.605 26.900 qs_ot_get_derivative 117 11.6 0.001 0.002 22.851 23.154 make_m2s 5014 13.6 0.094 0.099 20.691 21.834 make_images 5014 14.6 1.960 2.296 20.389 21.533 qs_ot_get_p 128 10.4 0.001 0.002 19.994 20.322 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.044 19.607 apply_single 128 13.6 0.001 0.001 19.044 19.607 ot_diis_step 117 11.6 0.018 0.019 19.590 19.591 cp_fm_upper_to_full 105 14.7 11.232 16.612 11.232 16.612 cp_fm_cholesky_invert 11 10.9 16.322 16.332 16.322 16.332 qs_ot_p2m_diag 83 11.4 0.343 0.390 15.849 15.901 multiply_cannon_metrocomm3 30084 15.6 0.047 0.049 6.649 15.024 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.557 14.558 sum_up_and_integrate 128 10.3 0.079 0.087 14.165 14.194 integrate_v_rspace 128 11.3 0.003 0.003 14.086 14.116 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.348 13.385 calculate_rho_elec 128 8.7 0.174 0.190 13.347 13.384 dbcsr_complete_redistribute 395 12.7 1.541 1.657 9.426 13.236 make_images_data 5014 15.6 0.064 0.069 11.279 13.218 multiply_cannon_sync_h2d 30084 15.6 11.678 12.694 11.678 12.694 hybrid_alltoall_any 5200 16.5 0.526 2.205 9.975 12.346 dbcsr_mm_accdrv_process 62264 16.2 7.392 8.233 11.470 11.991 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.985 11.794 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.319 11.321 cp_fm_redistribute_end 83 14.4 1.969 11.250 1.984 11.255 cp_fm_diag_elpa_base 83 14.4 8.648 10.618 9.240 11.135 init_scf_run 11 5.9 0.000 0.001 10.969 10.971 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.969 10.970 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.593 10.292 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.498 9.719 mp_alltoall_i22 716 14.1 5.649 9.430 5.649 9.430 pw_transfer 1547 11.6 0.083 0.097 7.926 8.007 grid_integrate_task_list 128 12.3 7.524 7.844 7.524 7.844 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.698 7.783 cp_fm_cholesky_decompose 22 10.9 7.637 7.735 7.637 7.735 wfi_extrapolate 11 7.9 0.001 0.001 7.726 7.726 multiply_cannon_metrocomm4 25070 15.6 0.075 0.085 2.823 7.410 mp_irecv_dv 76098 16.2 2.679 7.143 2.679 7.143 fft_wrap_pw1pw2_140 523 13.2 0.472 0.483 6.823 6.931 density_rs2pw 128 9.7 0.005 0.006 6.414 6.788 calculate_dm_sparse 128 9.5 0.001 0.001 6.201 6.281 fft3d_ps 1291 14.7 2.798 2.887 5.939 6.011 mp_alltoall_d11v 2415 14.1 5.287 5.955 5.287 5.955 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.537 5.582 grid_collocate_task_list 128 9.7 5.209 5.554 5.209 5.554 rs_pw_transfer 1046 11.9 0.013 0.014 4.903 5.277 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.426 4.558 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.422 4.515 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 4.505 4.505 potential_pw2rs 128 12.3 0.024 0.024 4.377 4.388 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.272 4.325 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.889000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1082.272727, yerr=21.011213 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 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 1.521656E+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 5.532237E+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 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58352. MP_Allreduce 10977 1175. MP_Sync 87 MP_Alltoall 1712 18838210. MP_ISendRecv 7680 122880. MP_Wait 19962 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 0.018 0.033 171.052 171.052 qs_mol_dyn_low 1 2.0 0.003 0.003 170.668 170.680 qs_forces 11 3.9 0.003 0.003 170.567 170.571 qs_energies 11 4.9 0.001 0.002 163.255 163.263 scf_env_do_scf 11 5.9 0.001 0.001 145.824 145.839 velocity_verlet 10 3.0 0.002 0.002 112.180 112.184 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 109.831 109.832 qs_scf_new_mos 117 7.6 0.001 0.001 74.554 74.645 qs_scf_loop_do_ot 117 8.6 0.001 0.001 74.553 74.644 dbcsr_multiply_generic 2507 12.6 0.180 0.186 73.751 74.037 ot_scf_mini 117 9.6 0.004 0.004 70.150 70.189 multiply_cannon 2507 13.6 0.584 0.616 54.507 58.680 multiply_cannon_loop 2507 14.6 0.447 0.458 49.755 50.239 ot_mini 117 10.6 0.001 0.001 39.407 39.447 init_scf_loop 11 6.9 0.000 0.000 35.837 35.840 mp_waitall_1 129618 16.6 25.297 32.914 25.297 32.914 prepare_preconditioner 11 7.9 0.000 0.000 31.992 32.020 make_preconditioner 11 8.9 0.000 0.000 31.992 32.020 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.813 30.088 rebuild_ks_matrix 128 8.3 0.001 0.001 28.347 28.439 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 28.346 28.438 qs_ks_update_qs_env 128 7.6 0.001 0.001 25.857 25.937 multiply_cannon_multrec 10028 15.6 10.477 14.421 17.997 20.677 ot_diis_step 117 11.6 0.019 0.020 19.808 19.809 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.416 19.663 apply_single 128 13.6 0.001 0.001 19.416 19.663 qs_ot_get_derivative 117 11.6 0.002 0.002 19.518 19.561 multiply_cannon_metrocomm3 10028 15.6 0.022 0.023 12.148 18.552 cp_fm_cholesky_invert 11 10.9 18.250 18.256 18.250 18.256 make_m2s 5014 13.6 0.066 0.070 15.639 18.164 qs_ot_get_p 128 10.4 0.001 0.001 17.952 18.014 make_images 5014 14.6 2.317 2.860 15.335 17.861 qs_ot_p2m_diag 83 11.4 0.495 0.501 14.282 14.299 sum_up_and_integrate 128 10.3 0.103 0.109 13.870 13.920 integrate_v_rspace 128 11.3 0.003 0.004 13.766 13.821 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.128 13.129 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.022 13.054 calculate_rho_elec 128 8.7 0.257 0.268 13.021 13.053 multiply_cannon_sync_h2d 10028 15.6 11.660 12.118 11.660 12.118 make_images_data 5014 15.6 0.053 0.060 9.412 11.976 hybrid_alltoall_any 5200 16.5 0.812 3.550 9.203 11.841 init_scf_run 11 5.9 0.000 0.001 10.562 10.562 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.562 10.562 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.097 10.103 cp_fm_diag_elpa_base 83 14.4 9.845 9.932 10.089 10.096 cp_fm_cholesky_decompose 22 10.9 7.997 8.107 7.997 8.107 grid_integrate_task_list 128 12.3 7.796 8.031 7.796 8.031 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.797 7.827 dbcsr_mm_accdrv_process 20762 16.1 2.636 3.454 7.146 7.776 pw_transfer 1547 11.6 0.082 0.091 7.453 7.476 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.227 7.255 wfi_extrapolate 11 7.9 0.001 0.001 7.227 7.227 multiply_cannon_metrocomm1 10028 15.6 0.029 0.030 4.245 6.976 mp_allgather_i34 2507 14.6 2.953 6.911 2.953 6.911 density_rs2pw 128 9.7 0.005 0.006 5.896 6.386 fft_wrap_pw1pw2_140 523 13.2 0.493 0.508 6.342 6.379 calculate_dm_sparse 128 9.5 0.001 0.001 6.145 6.203 grid_collocate_task_list 128 9.7 5.474 5.864 5.474 5.864 dbcsr_complete_redistribute 395 12.7 2.171 2.272 5.259 5.643 fft3d_ps 1291 14.7 2.721 2.793 5.393 5.417 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.277 5.277 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.202 5.209 mp_alltoall_d11v 2415 14.1 4.537 4.957 4.537 4.957 rs_pw_transfer 1046 11.9 0.013 0.013 4.285 4.791 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.124 4.156 potential_pw2rs 128 12.3 0.027 0.028 4.072 4.086 multiply_cannon_metrocomm4 7521 15.6 0.024 0.026 1.852 4.038 mp_irecv_dv 28860 15.9 1.816 3.967 1.816 3.967 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.596 3.887 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.462 3.796 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.682 3.708 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.544 3.612 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.469 3.483 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=171.052000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1428.545455, yerr=51.706099 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 11.786061E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 2.292199E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+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 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4045 58536. MP_Allreduce 11109 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_ISendRecv 3612 218624. MP_Wait 11682 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.042 292.003 292.003 qs_mol_dyn_low 1 2.0 0.003 0.004 291.304 291.316 qs_forces 11 3.9 0.003 0.003 291.218 291.220 qs_energies 11 4.9 0.002 0.002 282.557 282.571 scf_env_do_scf 11 5.9 0.001 0.001 260.988 261.006 velocity_verlet 10 3.0 0.002 0.002 209.802 209.809 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 136.834 136.836 init_scf_loop 11 6.9 0.000 0.000 123.897 123.898 prepare_preconditioner 11 7.9 0.000 0.000 119.169 119.204 make_preconditioner 11 8.9 0.000 0.000 119.169 119.204 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.347 116.275 qs_scf_new_mos 118 7.6 0.001 0.001 93.971 94.072 qs_scf_loop_do_ot 118 8.6 0.001 0.001 93.970 94.071 ot_scf_mini 118 9.6 0.004 0.004 89.117 89.185 dbcsr_multiply_generic 2529 12.6 0.218 0.224 84.675 85.149 cp_fm_upper_to_full 106 14.8 52.424 75.219 52.424 75.219 multiply_cannon 2529 13.6 0.699 0.756 59.966 60.395 multiply_cannon_loop 2529 14.6 0.475 0.483 56.170 57.399 ot_mini 118 10.6 0.001 0.001 45.333 45.396 dbcsr_complete_redistribute 397 12.7 4.084 4.125 29.485 42.216 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.073 38.812 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.776 36.427 rebuild_ks_matrix 129 8.3 0.001 0.001 34.212 34.255 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 34.212 34.254 mp_alltoall_i22 720 14.1 21.593 34.165 21.593 34.165 cp_fm_cholesky_invert 11 10.9 33.708 33.714 33.708 33.714 mp_waitall_1 106386 16.7 29.283 33.051 29.283 33.051 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.962 31.998 qs_ot_get_p 129 10.4 0.001 0.001 27.958 28.014 qs_ot_get_derivative 118 11.6 0.002 0.002 24.617 24.679 qs_ot_p2m_diag 84 11.4 0.890 0.895 23.732 23.761 cp_dbcsr_syevd 84 12.4 0.006 0.006 21.957 21.960 make_m2s 5058 13.6 0.077 0.078 19.853 20.893 multiply_cannon_metrocomm3 10116 15.6 0.023 0.023 19.758 20.882 ot_diis_step 118 11.6 0.022 0.023 20.667 20.667 make_images 5058 14.6 3.783 3.888 19.376 20.418 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.868 20.005 apply_single 129 13.6 0.001 0.001 19.867 20.005 cp_fm_diag_elpa 84 13.4 0.000 0.000 18.681 18.683 cp_fm_diag_elpa_base 84 14.4 14.199 15.860 18.677 18.679 multiply_cannon_multrec 10116 15.6 10.614 12.400 18.102 18.204 sum_up_and_integrate 129 10.3 0.195 0.197 15.796 15.902 multiply_cannon_sync_h2d 10116 15.6 15.777 15.782 15.777 15.782 integrate_v_rspace 129 11.3 0.004 0.004 15.600 15.707 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.196 15.210 calculate_rho_elec 129 8.7 0.486 0.487 15.196 15.210 make_images_data 5058 15.6 0.061 0.065 10.592 12.401 hybrid_alltoall_any 5245 16.5 1.311 3.057 10.630 12.358 init_scf_run 11 5.9 0.000 0.001 11.700 11.701 scf_env_initial_rho_setup 11 6.9 0.000 0.000 11.700 11.701 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.809 9.854 dbcsr_mm_accdrv_process 20934 16.1 3.891 5.600 7.250 9.089 cp_fm_cholesky_decompose 22 10.9 8.982 9.026 8.982 9.026 grid_integrate_task_list 129 12.3 8.614 8.781 8.614 8.781 wfi_extrapolate 11 7.9 0.001 0.001 8.604 8.605 pw_transfer 1559 11.6 0.094 0.095 8.509 8.513 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.268 8.273 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.909 7.909 fft_wrap_pw1pw2_140 527 13.2 0.530 0.534 7.302 7.314 mp_alltoall_d11v 2429 14.1 6.920 7.058 6.920 7.058 calculate_dm_sparse 129 9.5 0.001 0.001 6.717 6.792 grid_collocate_task_list 129 9.7 6.386 6.422 6.386 6.422 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.282 6.366 fft3d_ps 1301 14.7 2.783 2.813 6.307 6.321 density_rs2pw 129 9.7 0.005 0.005 6.200 6.229 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.052 6.142 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=292.003000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2154.545455, yerr=54.880637 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 1.260167E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.064 85.162 85.163 qs_energies 1 2.0 0.000 0.000 84.639 84.653 ls_scf 1 3.0 0.000 0.003 83.688 83.701 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.587 72.752 multiply_cannon 111 7.7 0.018 0.021 55.883 57.305 multiply_cannon_loop 111 8.7 0.209 0.222 52.444 54.099 ls_scf_main 1 4.0 0.000 0.002 52.320 52.321 density_matrix_trs4 2 5.0 0.002 0.003 46.818 46.877 ls_scf_init_scf 1 4.0 0.000 0.001 28.307 28.309 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.219 27.276 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.065 25.094 mp_waitall_1 11316 10.9 22.242 24.632 22.242 24.632 multiply_cannon_multrec 2664 9.7 8.158 8.960 15.558 17.369 multiply_cannon_sync_h2d 2664 9.7 13.674 15.402 13.674 15.402 make_m2s 222 7.7 0.008 0.011 13.004 13.521 make_images 222 8.7 0.099 0.108 12.982 13.502 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.583 11.907 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.413 8.324 hybrid_alltoall_any 227 10.6 0.216 1.840 6.474 8.156 make_images_data 222 9.7 0.004 0.005 7.557 8.154 dbcsr_mm_accdrv_process 4760 10.4 0.510 0.610 7.021 7.987 dbcsr_mm_accdrv_process_sort 4760 11.4 6.312 7.229 6.312 7.229 calculate_norms 4752 9.8 5.534 6.109 5.534 6.109 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.062 5.195 mp_sum_l 807 5.4 3.171 4.657 3.171 4.657 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.366 3.468 make_images_sizes 222 9.7 0.000 0.000 0.832 3.467 mp_alltoall_i44 222 10.7 0.832 3.467 0.832 3.467 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.067 3.367 mp_irecv_dv 6231 10.9 2.049 3.338 2.049 3.338 arnoldi_extremal 4 6.8 0.000 0.003 3.269 3.298 arnoldi_normal_ev 4 7.8 0.001 0.006 3.269 3.298 build_subspace 16 8.4 0.009 0.013 3.168 3.171 ls_scf_post 1 4.0 0.000 0.003 3.060 3.073 ls_scf_store_result 1 5.0 0.000 0.000 2.872 2.911 dbcsr_special_finalize 555 9.7 0.005 0.006 2.262 2.744 dbcsr_merge_single_wm 555 10.7 0.456 0.568 2.254 2.736 make_images_pack 222 9.7 2.207 2.629 2.209 2.630 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.334 2.568 dbcsr_sort_data 658 11.4 2.055 2.492 2.055 2.492 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.456 2.069 2.458 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.242 2.330 buffer_matrices_ensure_size 222 8.7 1.752 2.104 1.752 2.104 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.780 1.784 rebuild_ks_matrix 3 7.3 0.000 0.000 1.770 1.774 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.007 1.770 1.774 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.163000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 2.108445E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.102 0.125 91.890 91.890 qs_energies 1 2.0 0.000 0.000 91.166 91.172 ls_scf 1 3.0 0.000 0.000 89.834 89.841 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.868 76.197 multiply_cannon 111 7.7 0.028 0.039 53.220 57.571 ls_scf_main 1 4.0 0.000 0.000 55.488 55.493 multiply_cannon_loop 111 8.7 0.116 0.123 49.991 53.466 density_matrix_trs4 2 5.0 0.002 0.003 49.776 50.000 ls_scf_init_scf 1 4.0 0.000 0.000 30.804 30.805 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.646 29.759 mp_waitall_1 9246 10.9 21.345 29.285 21.345 29.285 multiply_cannon_multrec 1332 9.7 13.099 17.075 22.409 27.510 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.235 27.246 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.763 20.577 make_m2s 222 7.7 0.006 0.007 15.543 16.219 make_images 222 8.7 1.583 1.924 15.513 16.189 dbcsr_mm_accdrv_process 4041 10.4 0.275 0.451 8.912 10.518 dbcsr_mm_accdrv_process_sort 4041 11.4 8.506 10.067 8.506 10.067 make_images_data 222 9.7 0.004 0.004 8.977 9.851 hybrid_alltoall_any 227 10.6 0.519 2.447 8.373 9.380 mp_sum_l 807 5.4 5.801 8.883 5.801 8.883 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.219 7.813 mp_irecv_dv 3311 11.0 3.199 7.760 3.199 7.760 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.508 7.142 calculate_norms 2376 9.8 6.001 6.655 6.001 6.655 multiply_cannon_sync_h2d 1332 9.7 4.788 5.862 4.788 5.862 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.059 5.278 arnoldi_extremal 4 6.8 0.000 0.000 4.659 4.678 arnoldi_normal_ev 4 7.8 0.001 0.005 4.659 4.678 build_subspace 16 8.4 0.014 0.021 4.401 4.405 ls_scf_post 1 4.0 0.000 0.000 3.542 3.548 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.154 3.369 ls_scf_store_result 1 5.0 0.000 0.000 3.246 3.360 dbcsr_matrix_vector_mult_local 304 10.0 2.742 3.226 2.744 3.228 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.554 2.635 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.228 2.452 mp_allgather_i34 111 8.7 0.926 2.433 0.926 2.433 make_images_pack 222 9.7 2.026 2.414 2.028 2.416 dbcsr_sort_data 436 11.2 1.836 2.082 1.836 2.082 dbcsr_data_new 4174 10.1 1.612 1.844 1.612 1.844 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.890000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1718.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 2.699579E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.103 93.622 93.624 qs_energies 1 2.0 0.000 0.001 92.884 92.891 ls_scf 1 3.0 0.004 0.016 91.466 91.471 dbcsr_multiply_generic 111 6.7 0.015 0.016 76.246 76.590 ls_scf_main 1 4.0 0.000 0.001 57.069 57.073 multiply_cannon 111 7.7 0.048 0.172 52.795 56.178 multiply_cannon_loop 111 8.7 0.100 0.106 49.149 53.094 density_matrix_trs4 2 5.0 0.002 0.003 51.140 51.313 mp_waitall_1 7374 11.0 23.453 33.298 23.453 33.298 ls_scf_init_scf 1 4.0 0.000 0.001 30.736 30.738 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.548 29.647 matrix_sqrt_Newton_Schulz 2 6.5 0.003 0.015 27.132 27.146 multiply_cannon_multrec 888 9.7 12.634 15.331 21.256 24.561 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.805 22.643 make_m2s 222 7.7 0.006 0.007 16.680 17.998 make_images 222 8.7 1.973 2.306 16.642 17.957 make_images_data 222 9.7 0.003 0.004 9.522 10.658 hybrid_alltoall_any 227 10.6 0.620 2.863 9.234 10.553 dbcsr_mm_accdrv_process 3754 10.4 0.259 0.422 8.147 9.402 dbcsr_mm_accdrv_process_sort 3754 11.4 7.777 8.980 7.777 8.980 mp_sum_l 807 5.4 5.268 8.632 5.268 8.632 multiply_cannon_sync_h2d 888 9.7 6.018 7.135 6.018 7.135 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.448 6.775 mp_irecv_dv 2335 11.1 2.434 6.734 2.434 6.734 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.745 6.732 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.954 6.485 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.985 5.209 arnoldi_extremal 4 6.8 0.000 0.000 5.121 5.149 arnoldi_normal_ev 4 7.8 0.002 0.016 5.121 5.149 build_subspace 16 8.4 0.014 0.020 4.800 4.808 calculate_norms 1584 9.8 4.254 4.537 4.254 4.537 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.461 3.791 mp_allgather_i34 111 8.7 1.480 3.778 1.480 3.778 ls_scf_post 1 4.0 0.003 0.023 3.658 3.663 dbcsr_matrix_vector_mult_local 304 10.0 3.028 3.604 3.030 3.606 ls_scf_store_result 1 5.0 0.000 0.000 3.379 3.485 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.772 2.881 dbcsr_sort_data 325 11.1 1.901 2.150 1.901 2.150 make_images_pack 222 9.7 1.813 2.132 1.816 2.135 dbcsr_data_release 9322 10.9 1.299 1.963 1.299 1.963 dbcsr_finalize 304 7.8 0.026 0.032 1.614 1.896 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.890 1.893 rebuild_ks_matrix 3 7.3 0.000 0.000 1.872 1.875 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.007 1.872 1.875 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.624000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2252.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 3.369075E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.083 0.130 97.394 97.395 qs_energies 1 2.0 0.000 0.000 96.723 96.730 ls_scf 1 3.0 0.000 0.000 95.060 95.066 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.913 79.182 ls_scf_main 1 4.0 0.000 0.000 59.205 59.206 multiply_cannon 111 7.7 0.057 0.124 51.720 56.139 density_matrix_trs4 2 5.0 0.002 0.003 53.109 53.262 multiply_cannon_loop 111 8.7 0.114 0.126 46.706 50.123 ls_scf_init_scf 1 4.0 0.000 0.000 32.554 32.556 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.414 31.484 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.948 28.960 mp_waitall_1 6438 11.0 22.829 28.665 22.829 28.665 multiply_cannon_multrec 1332 9.7 14.291 17.229 22.222 25.449 make_m2s 222 7.7 0.006 0.008 21.339 22.719 make_images 222 8.7 3.142 3.596 21.289 22.671 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.173 16.959 make_images_data 222 9.7 0.004 0.004 11.910 13.452 hybrid_alltoall_any 227 10.6 0.797 3.823 11.290 12.945 dbcsr_mm_accdrv_process 3641 10.4 0.211 0.395 7.572 9.138 dbcsr_mm_accdrv_process_sort 3641 11.4 7.201 8.716 7.201 8.716 mp_sum_l 807 5.4 4.284 8.380 4.284 8.380 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.254 6.618 multiply_cannon_sync_h2d 1332 9.7 5.523 6.203 5.523 6.203 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.077 5.996 mp_irecv_dv 3229 10.9 2.054 5.913 2.054 5.913 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.574 5.648 arnoldi_extremal 4 6.8 0.000 0.000 5.242 5.257 arnoldi_normal_ev 4 7.8 0.001 0.005 5.242 5.257 build_subspace 16 8.4 0.014 0.021 4.892 4.900 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.584 4.747 calculate_norms 2376 9.8 4.196 4.548 4.196 4.548 mp_allgather_i34 111 8.7 2.128 4.365 2.128 4.365 dbcsr_matrix_vector_mult 304 9.0 0.006 0.017 3.578 3.884 dbcsr_matrix_vector_mult_local 304 10.0 3.198 3.701 3.200 3.702 dbcsr_sort_data 658 11.4 3.082 3.457 3.082 3.457 ls_scf_post 1 4.0 0.000 0.000 3.302 3.309 dbcsr_special_finalize 555 9.7 0.006 0.007 2.833 3.258 dbcsr_merge_single_wm 555 10.7 0.536 0.667 2.825 3.250 ls_scf_store_result 1 5.0 0.000 0.000 3.011 3.102 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.919 2.962 dbcsr_data_release 10477 10.7 1.584 2.421 1.584 2.421 dbcsr_finalize 304 7.8 0.049 0.061 1.801 1.976 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.395000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2722.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 4.625564E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.068 98.232 98.233 qs_energies 1 2.0 0.000 0.000 97.350 97.356 ls_scf 1 3.0 0.000 0.000 95.406 95.411 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.081 77.320 ls_scf_main 1 4.0 0.000 0.000 61.405 61.406 multiply_cannon 111 7.7 0.085 0.127 54.929 59.396 density_matrix_trs4 2 5.0 0.002 0.003 54.377 54.447 multiply_cannon_loop 111 8.7 0.069 0.077 50.333 52.060 ls_scf_init_scf 1 4.0 0.000 0.000 30.414 30.419 mp_waitall_1 5481 11.0 25.714 30.418 25.714 30.418 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.212 29.245 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.038 27.047 multiply_cannon_multrec 444 9.7 14.012 16.552 21.125 22.870 make_m2s 222 7.7 0.004 0.005 17.436 20.032 make_images 222 8.7 3.713 4.429 17.375 19.972 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.742 15.832 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.074 14.548 make_images_data 222 9.7 0.003 0.004 9.697 12.193 hybrid_alltoall_any 227 10.6 0.789 3.772 9.443 11.967 dbcsr_mm_accdrv_process 3003 10.4 0.164 0.325 6.821 7.954 multiply_cannon_sync_h2d 444 9.7 6.593 7.736 6.593 7.736 dbcsr_mm_accdrv_process_sort 3003 11.4 6.505 7.613 6.505 7.613 mp_allgather_i34 111 8.7 2.830 6.983 2.830 6.983 arnoldi_extremal 4 6.8 0.000 0.000 5.830 5.843 arnoldi_normal_ev 4 7.8 0.002 0.004 5.830 5.843 build_subspace 16 8.4 0.015 0.020 5.437 5.447 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.629 4.806 mp_sum_l 807 5.4 2.759 4.564 2.759 4.564 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.200 4.388 dbcsr_matrix_vector_mult_local 304 10.0 3.730 4.203 3.732 4.205 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.599 3.914 mp_irecv_dv 1241 11.2 1.579 3.891 1.579 3.891 calculate_norms 792 9.8 3.549 3.667 3.549 3.667 ls_scf_post 1 4.0 0.000 0.000 3.587 3.592 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.383 3.489 ls_scf_store_result 1 5.0 0.000 0.000 3.356 3.416 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.833 3.317 make_images_sizes 222 9.7 0.000 0.000 1.104 3.315 mp_alltoall_i44 222 10.7 1.104 3.315 1.104 3.315 dbcsr_finalize 304 7.8 0.062 0.078 2.204 2.341 dbcsr_merge_all 275 8.9 0.476 0.541 2.047 2.169 dbcsr_data_release 10123 10.8 1.335 2.009 1.335 2.009 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.992 1.993 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.233000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3628.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fb438e6dbebd23a036a1436a9c376208ffa38d38_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 8.712258E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.079 0.100 107.875 107.876 qs_energies 1 2.0 0.000 0.000 106.458 106.473 ls_scf 1 3.0 0.000 0.000 103.522 103.536 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.493 77.618 ls_scf_main 1 4.0 0.000 0.000 65.869 65.870 density_matrix_trs4 2 5.0 0.002 0.003 56.969 57.026 multiply_cannon 111 7.7 0.159 0.226 50.005 51.865 multiply_cannon_loop 111 8.7 0.067 0.069 46.494 47.522 ls_scf_init_scf 1 4.0 0.000 0.000 33.937 33.937 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.583 32.597 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.752 29.763 mp_waitall_1 4569 11.1 22.274 26.158 22.274 26.158 make_m2s 222 7.7 0.005 0.005 23.921 24.946 make_images 222 8.7 4.577 4.962 23.815 24.838 multiply_cannon_multrec 444 9.7 17.878 18.525 22.539 23.119 hybrid_alltoall_any 227 10.6 1.657 3.609 12.964 15.680 make_images_data 222 9.7 0.003 0.003 13.173 15.640 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.515 11.147 multiply_cannon_sync_h2d 444 9.7 8.849 8.893 8.849 8.893 arnoldi_extremal 4 6.8 0.000 0.000 7.497 7.509 arnoldi_normal_ev 4 7.8 0.003 0.009 7.497 7.509 build_subspace 16 8.4 0.026 0.037 6.941 6.953 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.574 5.726 dbcsr_matrix_vector_mult_local 304 10.0 5.117 5.425 5.119 5.428 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.023 5.288 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.846 4.940 dbcsr_mm_accdrv_process 1814 10.4 0.207 0.324 4.485 4.611 dbcsr_mm_accdrv_process_sort 1814 11.4 4.188 4.320 4.188 4.320 ls_scf_post 1 4.0 0.000 0.000 3.717 3.731 mp_allgather_i34 111 8.7 1.143 3.597 1.143 3.597 make_images_sizes 222 9.7 0.000 0.000 1.423 3.475 mp_alltoall_i44 222 10.7 1.423 3.474 1.423 3.474 ls_scf_store_result 1 5.0 0.000 0.000 3.421 3.430 calculate_norms 792 9.8 3.239 3.272 3.239 3.272 dbcsr_finalize 304 7.8 0.082 0.089 3.077 3.122 dbcsr_complete_redistribute 5 7.6 1.478 1.515 2.811 2.948 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.905 2.906 dbcsr_data_release 12724 10.6 2.323 2.905 2.323 2.905 dbcsr_merge_all 275 8.9 0.892 0.915 2.862 2.902 matrix_ls_to_qs 2 6.0 0.000 0.000 2.452 2.599 dbcsr_sort_data 325 11.1 2.434 2.500 2.434 2.500 dbcsr_new_transposed 4 7.5 0.245 0.254 2.268 2.281 dbcsr_frobenius_norm 74 6.6 2.054 2.137 2.201 2.242 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.228 2.230 dbcsr_add_d 103 6.2 0.000 0.000 2.146 2.222 dbcsr_add_anytype 103 7.2 0.859 0.891 2.145 2.221 rebuild_ks_matrix 3 7.3 0.000 0.000 2.162 2.163 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.000 2.162 2.163 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.876000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6835.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: fb438e6dbebd23a036a1436a9c376208ffa38d38 Summary: empty Status: OK