=== 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: 60432cd460210ee206a5df6c16abecd6840c7cae ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (12.01.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/01 job id: 44315123 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/02 job id: 44315124 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/03 job id: 44315126 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/04 job id: 44315128 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/05 job id: 44315129 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/06 job id: 44315131 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/07 job id: 44315133 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/08 job id: 44315136 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/09 job id: 44315137 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/10 job id: 44315138 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/11 job id: 44315139 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/12 job id: 44315140 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/13 job id: 44315141 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/14 job id: 44315143 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/15 job id: 44315144 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/16 job id: 44315145 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/17 job id: 44315147 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/18 job id: 44315148 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/19 job id: 44315149 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/20 job id: 44315150 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/21 job id: 44315151 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/22 job id: 44315152 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/23 job id: 44315154 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/24 job id: 44315156 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/25 job id: 44315159 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/26 job id: 44315161 --- 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/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.036 134.354 134.355 farming_run 1 2.0 133.788 133.790 134.326 134.329 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.462899E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.026 116.131 116.131 qs_energies 1 2.0 0.000 0.000 115.907 115.909 mp2_main 1 3.0 0.000 0.000 113.978 113.980 mp2_gpw_main 1 4.0 0.020 0.026 113.133 113.135 mp2_ri_gpw_compute_in 1 5.0 0.173 0.183 94.225 94.656 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.462 55.893 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.169 41.785 47.169 get_2c_integrals 1 6.0 0.001 0.015 37.567 38.591 integrate_v_rspace 273 8.0 0.435 0.447 25.104 30.250 pw_transfer 6555 10.6 0.375 0.392 27.500 28.025 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 26.137 26.588 grid_integrate_task_list 273 9.0 20.921 26.541 20.921 26.541 fft_wrap_pw1pw2_100 2178 12.4 1.226 1.464 23.656 24.107 compute_2c_integrals 1 7.0 0.017 0.018 19.872 19.873 compute_2c_integrals_loop_lm 1 8.0 0.023 0.030 19.022 19.556 mp2_eri_2c_integrate_gpw 1 9.0 2.366 2.435 18.999 19.526 rpa_ri_compute_en 1 5.0 0.001 0.002 18.802 18.975 cp_fm_cholesky_decompose 12 8.2 17.680 18.744 17.680 18.744 cholesky_decomp 1 7.0 0.000 0.000 16.554 17.610 fft3d_s 5443 13.4 16.189 16.585 16.210 16.606 ao_to_mo_and_store_B_mult_1 272 7.0 10.846 15.565 10.846 15.565 calculate_wavefunction 272 8.0 5.479 5.622 12.604 13.254 rpa_num_int 1 6.0 0.000 0.001 10.571 10.572 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.543 10.566 calc_mat_Q 8 8.0 0.000 0.000 9.368 9.474 contract_S_to_Q 8 9.0 0.000 0.000 8.788 8.894 calc_potential_gpw 544 9.5 0.005 0.006 8.259 8.699 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.274 8.585 parallel_gemm_fm 14 9.1 0.000 0.000 8.377 8.454 parallel_gemm_fm_cosma 14 10.1 8.377 8.454 8.377 8.454 potential_pw2rs 545 10.0 0.107 0.108 7.738 8.354 collocate_single_gaussian 272 10.0 0.039 0.041 7.538 7.832 create_integ_mat 1 6.0 0.022 0.026 7.704 7.704 array2fm 1 7.0 0.000 0.000 6.684 7.113 pw_scatter_s 2720 13.7 4.416 4.635 4.416 4.635 pw_gather_s 2722 13.2 3.891 4.214 3.891 4.214 array2fm_buffer_send 1 8.0 2.970 3.169 2.970 3.169 pw_poisson_solve 545 10.5 1.116 1.181 2.184 2.398 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.132471, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2735.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.049 0.051 397.747 397.748 farming_run 1 2.0 396.745 396.749 397.686 397.690 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.222394E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 2368424. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.046 209.266 209.266 qs_energies 1 2.0 0.000 0.000 209.026 209.033 scf_env_do_scf 1 3.0 0.000 0.000 106.394 106.394 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.468 105.478 rebuild_ks_matrix 4 6.0 0.000 0.000 105.467 105.477 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.068 105.467 105.477 hfx_ks_matrix 4 8.0 0.001 0.001 105.071 105.075 integrate_four_center 4 9.0 0.145 0.491 105.070 105.074 mp2_main 1 3.0 0.000 0.000 102.343 102.349 mp2_gpw_main 1 4.0 0.042 0.068 101.488 101.496 integrate_four_center_main 4 10.0 0.082 0.380 96.843 99.579 integrate_four_center_bin 263 11.0 96.761 99.560 96.761 99.560 init_scf_loop 1 4.0 0.000 0.000 92.090 92.090 mp2_ri_gpw_compute_in 1 5.0 0.068 0.074 74.773 75.748 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.451 55.434 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.156 42.205 47.176 integrate_v_rspace 95 8.0 0.397 0.571 28.568 33.379 pw_transfer 2240 10.6 0.145 0.166 29.894 30.365 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.914 29.339 grid_integrate_task_list 95 9.0 23.854 28.860 23.854 28.860 mp2_ri_gpw_compute_en 1 5.0 0.055 0.063 26.557 28.231 fft_wrap_pw1pw2_100 730 12.4 1.257 1.408 26.606 27.065 ao_to_mo_and_store_B_mult_1 91 7.0 10.565 25.753 10.565 25.753 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.839 1.894 24.913 24.926 get_2c_integrals 1 6.0 0.000 0.000 20.230 20.268 compute_2c_integrals 1 7.0 0.015 0.029 19.211 19.230 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.804 19.071 mp2_eri_2c_integrate_gpw 1 9.0 1.733 1.852 18.803 19.070 fft3d_s 1823 13.4 18.414 18.780 18.427 18.793 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.303 14.303 calculate_wavefunction 91 8.0 2.045 2.081 9.761 9.978 mp2_ri_gpw_compute_en_expansio 172 7.0 0.554 0.590 8.706 9.403 potential_pw2rs 186 10.0 0.033 0.034 8.623 9.163 local_gemm 172 8.0 8.152 8.813 8.152 8.813 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.213 8.645 mp2_ri_gpw_compute_en_comm 22 7.0 0.499 0.529 7.975 8.565 collocate_single_gaussian 91 10.0 0.017 0.020 7.855 8.152 calc_potential_gpw 182 9.5 0.002 0.002 7.901 8.129 mp_sendrecv_dm3 2068 8.0 6.005 6.599 6.005 6.599 mp2_ri_gpw_compute_en_ener 172 7.0 6.349 6.420 6.349 6.420 pw_gather_s 912 13.2 4.904 5.502 4.904 5.502 mp_sync 38 10.4 2.855 4.918 2.855 4.918 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.485622, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1509.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.136960E+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 578641. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.035 52.246 52.247 qs_mol_dyn_low 1 2.0 0.003 0.004 52.036 52.043 qs_forces 11 3.9 0.002 0.002 51.967 51.968 qs_energies 11 4.9 0.001 0.002 50.476 50.487 scf_env_do_scf 11 5.9 0.000 0.001 44.386 44.387 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 42.340 42.340 dbcsr_multiply_generic 2286 12.5 0.092 0.096 33.569 34.083 qs_scf_new_mos 108 7.5 0.000 0.001 32.070 32.368 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.069 32.368 ot_scf_mini 108 9.5 0.002 0.002 30.413 30.633 multiply_cannon 2286 13.5 0.185 0.192 26.130 27.899 multiply_cannon_loop 2286 14.5 1.514 1.585 25.411 27.151 velocity_verlet 10 3.0 0.007 0.019 25.676 25.677 ot_mini 108 10.5 0.001 0.001 19.345 19.580 qs_ot_get_derivative 108 11.5 0.001 0.001 16.389 16.590 mp_waitall_1 245248 16.5 8.530 15.002 8.530 15.002 multiply_cannon_metrocomm3 54864 15.5 0.068 0.073 6.002 13.354 multiply_cannon_multrec 54864 15.5 4.203 6.465 7.719 11.183 rebuild_ks_matrix 119 8.3 0.000 0.000 8.021 8.153 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.027 8.021 8.152 multiply_cannon_sync_h2d 54864 15.5 5.868 7.834 5.868 7.834 mp_sum_l 7207 12.9 5.421 7.219 5.421 7.219 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.063 7.180 qs_ot_get_p 119 10.4 0.001 0.001 6.450 6.769 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.553 6.005 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.476 5.586 init_scf_run 11 5.9 0.000 0.001 4.883 4.883 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.883 4.883 dbcsr_mm_accdrv_process 76910 16.1 1.195 1.829 3.438 4.773 sum_up_and_integrate 119 10.3 0.012 0.015 4.601 4.620 integrate_v_rspace 119 11.3 0.002 0.002 4.589 4.609 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.286 4.431 calculate_rho_elec 119 8.7 0.011 0.017 4.286 4.430 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.682 3.771 multiply_cannon_metrocomm1 54864 15.5 0.051 0.057 1.816 3.375 calculate_dm_sparse 119 9.5 0.000 0.001 3.079 3.202 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.906 3.102 apply_single 119 13.6 0.000 0.000 2.906 3.102 rs_pw_transfer 974 11.9 0.012 0.013 2.785 2.925 jit_kernel_multiply 13 15.8 2.181 2.880 2.181 2.880 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.864 2.864 calculate_first_density_matrix 1 7.0 0.001 0.001 2.699 2.703 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.657 2.658 ot_diis_step 108 11.5 0.006 0.006 2.657 2.658 cp_fm_redistribute_end 50 14.0 2.419 2.636 2.425 2.638 cp_fm_diag_elpa_base 50 14.0 0.211 2.543 0.212 2.556 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.420 2.489 density_rs2pw 119 9.7 0.004 0.004 2.305 2.412 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.309 2.311 acc_transpose_blocks 54864 15.5 0.234 0.251 1.791 2.200 grid_integrate_task_list 119 12.3 2.028 2.140 2.028 2.140 wfi_extrapolate 11 7.9 0.001 0.001 2.118 2.118 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.044 2.102 init_scf_loop 11 6.9 0.000 0.000 2.030 2.030 pw_transfer 1439 11.6 0.053 0.061 1.894 1.979 potential_pw2rs 119 12.3 0.004 0.004 1.932 1.949 mp_sum_d 4125 12.0 1.292 1.947 1.292 1.947 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.817 1.904 make_m2s 4572 13.5 0.054 0.056 1.711 1.766 make_images 4572 14.5 0.133 0.138 1.628 1.682 fft3d_ps 1201 14.6 0.371 0.474 1.588 1.663 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.584 1.607 mp_alltoall_d11v 2130 13.8 1.306 1.516 1.306 1.516 fft_wrap_pw1pw2_140 487 13.2 0.082 0.095 1.386 1.477 mp_waitany 12084 13.8 1.267 1.411 1.267 1.411 grid_collocate_task_list 119 9.7 1.290 1.363 1.290 1.363 dbcsr_dot_sd 1205 11.9 0.051 0.061 0.718 1.094 acc_transpose_blocks_kernels 54864 16.5 0.252 0.381 0.781 1.045 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.247000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.363636, yerr=0.979121 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_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 489.000960E+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 1031290. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.037 40.084 40.085 qs_mol_dyn_low 1 2.0 0.003 0.005 39.811 39.818 qs_forces 11 3.9 0.002 0.002 39.753 39.755 qs_energies 11 4.9 0.003 0.011 38.048 38.051 scf_env_do_scf 11 5.9 0.001 0.001 32.637 32.638 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 30.008 30.009 dbcsr_multiply_generic 2286 12.5 0.099 0.101 21.850 22.228 qs_scf_new_mos 108 7.5 0.001 0.001 20.469 20.710 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.468 20.710 ot_scf_mini 108 9.5 0.003 0.003 19.556 19.727 velocity_verlet 10 3.0 0.002 0.002 18.735 18.736 multiply_cannon 2286 13.5 0.209 0.221 16.727 18.468 multiply_cannon_loop 2286 14.5 0.907 0.977 15.595 17.175 ot_mini 108 10.5 0.001 0.001 12.128 12.366 mp_waitall_1 200699 16.5 5.963 11.371 5.963 11.371 qs_ot_get_derivative 108 11.5 0.001 0.001 9.647 9.821 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.389 9.747 multiply_cannon_multrec 27432 15.5 1.957 4.251 6.019 8.876 rebuild_ks_matrix 119 8.3 0.000 0.000 7.385 7.529 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.020 7.385 7.528 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.543 6.671 dbcsr_mm_accdrv_process 47894 16.0 3.076 5.149 3.992 6.067 qs_ot_get_p 119 10.4 0.001 0.001 4.532 4.766 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.814 4.666 sum_up_and_integrate 119 10.3 0.024 0.026 4.334 4.339 integrate_v_rspace 119 11.3 0.002 0.002 4.310 4.318 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.057 4.256 apply_single 119 13.6 0.000 0.000 3.057 4.256 mp_sum_l 7207 12.9 2.141 4.166 2.141 4.166 init_scf_run 11 5.9 0.000 0.001 4.132 4.132 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.132 4.132 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.992 4.021 calculate_rho_elec 119 8.7 0.021 0.024 3.991 4.021 rs_pw_transfer 974 11.9 0.010 0.011 2.806 3.259 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.044 3.065 multiply_cannon_sync_h2d 27432 15.5 2.174 2.809 2.174 2.809 make_m2s 4572 13.5 0.053 0.054 2.536 2.771 density_rs2pw 119 9.7 0.004 0.004 2.316 2.768 make_images 4572 14.5 0.199 0.237 2.448 2.679 init_scf_loop 11 6.9 0.001 0.009 2.607 2.608 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.601 2.602 calculate_first_density_matrix 1 7.0 0.001 0.003 2.581 2.582 ot_diis_step 108 11.5 0.011 0.011 2.422 2.423 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.195 2.283 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.253 2.253 cp_fm_redistribute_end 50 14.0 1.861 2.224 1.866 2.225 calculate_dm_sparse 119 9.5 0.000 0.001 2.134 2.206 cp_fm_diag_elpa_base 50 14.0 0.343 2.099 0.357 2.155 pw_transfer 1439 11.6 0.065 0.072 2.089 2.123 potential_pw2rs 119 12.3 0.006 0.006 2.049 2.059 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.040 2.042 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.997 2.034 grid_integrate_task_list 119 12.3 1.837 1.930 1.837 1.930 jit_kernel_multiply 9 16.1 0.864 1.887 0.864 1.887 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.808 1.849 fft3d_ps 1201 14.6 0.515 0.567 1.700 1.733 prepare_preconditioner 11 7.9 0.000 0.000 1.664 1.692 make_preconditioner 11 8.9 0.000 0.001 1.664 1.692 acc_transpose_blocks 27432 15.5 0.109 0.112 1.253 1.650 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.557 1.619 make_images_data 4572 15.5 0.044 0.050 1.164 1.597 fft_wrap_pw1pw2_140 487 13.2 0.079 0.086 1.527 1.562 hybrid_alltoall_any 4725 16.4 0.050 0.112 1.030 1.496 wfi_extrapolate 11 7.9 0.001 0.001 1.494 1.494 mp_alltoall_d11v 2130 13.8 1.232 1.396 1.232 1.396 grid_collocate_task_list 119 9.7 1.228 1.364 1.228 1.364 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.333 1.340 mp_allgather_i34 2286 14.5 0.575 1.328 0.575 1.328 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.183 1.228 mp_sum_d 4125 12.0 0.635 1.094 0.635 1.094 acc_transpose_blocks_kernels 27432 16.5 0.181 0.269 0.714 1.045 mp_waitany 5720 13.7 0.551 1.026 0.551 1.026 rs_pw_transfer_RS2PW_140 130 11.5 0.140 0.148 0.574 1.019 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.978 0.979 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.906 0.921 mp_alltoall_z22v 1201 16.6 0.773 0.873 0.773 0.873 rs_pw_transfer_PW2RS_50 119 14.3 0.587 0.605 0.782 0.830 make_images_sizes 4572 15.5 0.005 0.005 0.559 0.808 mp_alltoall_i44 4572 16.5 0.554 0.803 0.554 0.803 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.085000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.272727, yerr=1.656442 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 522.350592E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.080 0.126 33.231 33.233 qs_mol_dyn_low 1 2.0 0.003 0.004 32.723 32.731 qs_forces 11 3.9 0.002 0.004 32.666 32.666 qs_energies 11 4.9 0.013 0.035 31.057 31.059 scf_env_do_scf 11 5.9 0.001 0.001 26.151 26.151 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 23.472 23.474 dbcsr_multiply_generic 2286 12.5 0.093 0.095 16.696 16.810 velocity_verlet 10 3.0 0.004 0.007 15.572 15.573 qs_scf_new_mos 108 7.5 0.001 0.001 15.082 15.110 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.082 15.109 ot_scf_mini 108 9.5 0.003 0.003 14.345 14.363 multiply_cannon 2286 13.5 0.197 0.206 13.323 14.060 multiply_cannon_loop 2286 14.5 0.639 0.672 12.525 13.312 ot_mini 108 10.5 0.001 0.001 8.872 8.895 multiply_cannon_multrec 18288 15.5 1.997 3.018 7.027 7.372 qs_ot_get_derivative 108 11.5 0.001 0.001 7.349 7.371 rebuild_ks_matrix 119 8.3 0.000 0.001 6.549 6.571 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 6.549 6.570 dbcsr_mm_accdrv_process 38222 16.0 4.121 5.612 4.946 5.820 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.755 5.775 sum_up_and_integrate 119 10.3 0.029 0.030 4.085 4.106 integrate_v_rspace 119 11.3 0.002 0.003 4.056 4.080 mp_waitall_1 158411 16.6 2.649 4.040 2.649 4.040 init_scf_run 11 5.9 0.000 0.001 3.669 3.669 scf_env_initial_rho_setup 11 6.9 0.001 0.003 3.669 3.669 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.541 3.552 calculate_rho_elec 119 8.7 0.030 0.032 3.541 3.551 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.736 3.364 qs_ot_get_p 119 10.4 0.001 0.002 3.294 3.313 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.536 2.794 init_scf_loop 11 6.9 0.010 0.039 2.661 2.666 rs_pw_transfer 974 11.9 0.009 0.010 2.358 2.603 calculate_first_density_matrix 1 7.0 0.001 0.003 2.430 2.431 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.063 2.384 apply_single 119 13.6 0.000 0.000 2.063 2.384 density_rs2pw 119 9.7 0.004 0.004 2.053 2.316 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.202 2.209 pw_transfer 1439 11.6 0.066 0.072 2.050 2.076 jit_kernel_multiply 9 16.1 0.775 1.999 0.775 1.999 calculate_dm_sparse 119 9.5 0.000 0.000 1.981 1.992 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.956 1.985 make_m2s 4572 13.5 0.044 0.045 1.809 1.963 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.934 1.934 grid_integrate_task_list 119 12.3 1.802 1.887 1.802 1.887 make_images 4572 14.5 0.189 0.202 1.723 1.876 potential_pw2rs 119 12.3 0.007 0.008 1.848 1.873 prepare_preconditioner 11 7.9 0.000 0.000 1.826 1.838 make_preconditioner 11 8.9 0.001 0.003 1.826 1.838 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.677 1.768 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.714 1.715 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.704 1.705 cp_fm_redistribute_end 50 14.0 1.271 1.684 1.272 1.684 fft3d_ps 1201 14.6 0.524 0.541 1.636 1.666 cp_fm_diag_elpa_base 50 14.0 0.395 1.601 0.410 1.646 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.585 1.592 multiply_cannon_sync_h2d 18288 15.5 1.385 1.592 1.385 1.592 fft_wrap_pw1pw2_140 487 13.2 0.090 0.094 1.556 1.583 mp_sum_l 7207 12.9 1.192 1.579 1.192 1.579 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.566 1.573 ot_diis_step 108 11.5 0.011 0.011 1.494 1.494 grid_collocate_task_list 119 9.7 1.213 1.354 1.213 1.354 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.337 1.341 acc_transpose_blocks 18288 15.5 0.075 0.077 1.289 1.312 wfi_extrapolate 11 7.9 0.001 0.001 1.190 1.190 make_images_data 4572 15.5 0.044 0.048 0.819 0.996 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.958 0.959 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.907 0.928 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.700 0.903 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.867 0.870 acc_transpose_blocks_kernels 18288 16.5 0.209 0.218 0.845 0.858 mp_alltoall_z22v 1201 16.6 0.712 0.831 0.712 0.831 mp_alltoall_d11v 2130 13.8 0.687 0.812 0.687 0.812 mp_waitany 9880 13.7 0.526 0.788 0.526 0.788 cp_fm_cholesky_invert 11 10.9 0.770 0.774 0.770 0.774 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.123 0.518 0.769 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.649 0.715 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.233000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.090909, yerr=2.193152 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_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 556.490752E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.105 0.124 36.232 36.233 qs_mol_dyn_low 1 2.0 0.003 0.003 35.859 35.867 qs_forces 11 3.9 0.002 0.002 35.799 35.801 qs_energies 11 4.9 0.001 0.001 34.060 34.068 scf_env_do_scf 11 5.9 0.000 0.001 28.974 28.975 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.413 25.413 dbcsr_multiply_generic 2286 12.5 0.099 0.101 18.340 18.447 velocity_verlet 10 3.0 0.002 0.002 18.374 18.376 qs_scf_new_mos 108 7.5 0.001 0.001 16.595 16.649 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.594 16.648 ot_scf_mini 108 9.5 0.002 0.003 15.649 15.703 multiply_cannon 2286 13.5 0.225 0.250 14.500 14.992 multiply_cannon_loop 2286 14.5 0.943 0.970 13.514 13.849 ot_mini 108 10.5 0.001 0.001 9.511 9.579 multiply_cannon_multrec 27432 15.5 2.330 2.984 8.576 8.910 qs_ot_get_derivative 108 11.5 0.001 0.001 7.673 7.724 dbcsr_mm_accdrv_process 47916 15.9 5.153 6.381 6.153 7.342 rebuild_ks_matrix 119 8.3 0.000 0.000 6.818 6.870 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.818 6.870 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.060 6.105 sum_up_and_integrate 119 10.3 0.035 0.038 3.989 3.997 integrate_v_rspace 119 11.3 0.002 0.003 3.954 3.962 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.717 3.750 calculate_rho_elec 119 8.7 0.040 0.046 3.717 3.750 init_scf_run 11 5.9 0.000 0.001 3.651 3.651 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.650 3.651 qs_ot_get_p 119 10.4 0.001 0.001 3.496 3.571 init_scf_loop 11 6.9 0.000 0.000 3.538 3.538 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.749 3.218 prepare_preconditioner 11 7.9 0.000 0.000 2.658 2.668 make_preconditioner 11 8.9 0.000 0.000 2.658 2.667 mp_waitall_1 137007 16.6 2.037 2.638 2.037 2.638 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.261 2.595 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.130 2.593 apply_single 119 13.6 0.000 0.000 2.130 2.593 make_m2s 4572 13.5 0.055 0.056 2.306 2.438 rs_pw_transfer 974 11.9 0.009 0.010 2.244 2.431 make_images 4572 14.5 0.269 0.328 2.199 2.329 density_rs2pw 119 9.7 0.004 0.004 2.128 2.311 pw_transfer 1439 11.6 0.066 0.072 2.199 2.238 calculate_first_density_matrix 1 7.0 0.001 0.001 2.235 2.237 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.160 2.168 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.105 2.149 calculate_dm_sparse 119 9.5 0.000 0.000 2.033 2.094 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.996 2.022 grid_integrate_task_list 119 12.3 1.823 1.915 1.823 1.915 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.837 1.838 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.827 1.828 fft3d_ps 1201 14.6 0.561 0.613 1.779 1.816 jit_kernel_multiply 10 15.9 0.941 1.815 0.941 1.815 ot_diis_step 108 11.5 0.012 0.012 1.795 1.795 fft_wrap_pw1pw2_140 487 13.2 0.089 0.096 1.715 1.765 potential_pw2rs 119 12.3 0.008 0.009 1.735 1.740 mp_sum_l 7207 12.9 1.109 1.645 1.109 1.645 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.626 1.641 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.563 1.564 cp_fm_redistribute_end 50 14.0 1.037 1.539 1.038 1.540 cp_fm_diag_elpa_base 50 14.0 0.479 1.468 0.499 1.507 acc_transpose_blocks 27432 15.5 0.113 0.115 1.466 1.495 wfi_extrapolate 11 7.9 0.001 0.001 1.366 1.366 grid_collocate_task_list 119 9.7 1.221 1.330 1.221 1.330 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.828 1.303 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.216 1.225 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.164 1.166 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.132 1.147 cp_fm_upper_to_full 72 13.5 0.806 1.122 0.806 1.122 dbcsr_complete_redistribute 329 12.2 0.122 0.141 0.837 1.110 make_images_data 4572 15.5 0.045 0.048 0.919 1.065 multiply_cannon_sync_h2d 27432 15.5 0.978 1.021 0.978 1.021 hybrid_alltoall_any 4725 16.4 0.061 0.150 0.809 0.969 mp_alltoall_d11v 2130 13.8 0.789 0.880 0.789 0.880 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.611 0.879 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.791 0.869 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.849 0.854 acc_transpose_blocks_kernels 27432 16.5 0.268 0.276 0.832 0.853 mp_alltoall_z22v 1201 16.6 0.802 0.843 0.802 0.843 cp_fm_cholesky_invert 11 10.9 0.806 0.809 0.806 0.809 mp_alltoall_i22 627 13.8 0.445 0.738 0.445 0.738 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.233000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.272727, yerr=2.831348 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_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 618.360832E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.029 28.668 28.669 qs_mol_dyn_low 1 2.0 0.005 0.035 28.390 28.397 qs_forces 11 3.9 0.002 0.002 28.314 28.316 qs_energies 11 4.9 0.001 0.001 26.588 26.593 scf_env_do_scf 11 5.9 0.000 0.001 21.665 21.665 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.163 19.163 velocity_verlet 10 3.0 0.024 0.025 14.621 14.624 dbcsr_multiply_generic 2286 12.5 0.091 0.094 12.158 12.229 qs_scf_new_mos 108 7.5 0.001 0.001 10.978 11.007 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.978 11.006 ot_scf_mini 108 9.5 0.002 0.002 10.300 10.331 multiply_cannon 2286 13.5 0.230 0.234 9.538 9.997 multiply_cannon_loop 2286 14.5 0.331 0.342 8.615 8.790 rebuild_ks_matrix 119 8.3 0.000 0.000 6.107 6.128 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.106 6.128 multiply_cannon_multrec 9144 15.5 1.580 1.840 5.755 6.002 ot_mini 108 10.5 0.001 0.001 5.681 5.713 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.442 5.462 qs_ot_get_derivative 108 11.5 0.001 0.001 4.368 4.397 dbcsr_mm_accdrv_process 12550 15.8 3.152 3.846 4.074 4.144 sum_up_and_integrate 119 10.3 0.038 0.041 3.742 3.746 integrate_v_rspace 119 11.3 0.003 0.003 3.704 3.709 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.628 3.634 calculate_rho_elec 119 8.7 0.060 0.061 3.627 3.633 init_scf_run 11 5.9 0.000 0.001 3.457 3.457 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.457 3.457 qs_ot_get_p 119 10.4 0.001 0.001 2.809 2.844 init_scf_loop 11 6.9 0.000 0.000 2.482 2.483 calculate_first_density_matrix 1 7.0 0.001 0.001 2.292 2.294 pw_transfer 1439 11.6 0.066 0.070 2.203 2.213 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.109 2.121 density_rs2pw 119 9.7 0.004 0.004 1.951 2.092 mp_waitall_1 115863 16.7 1.552 2.072 1.552 2.072 make_m2s 4572 13.5 0.034 0.036 1.832 1.977 grid_integrate_task_list 119 12.3 1.849 1.924 1.849 1.924 rs_pw_transfer 974 11.9 0.008 0.008 1.758 1.906 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.888 1.891 make_images 4572 14.5 0.266 0.299 1.743 1.887 fft3d_ps 1201 14.6 0.563 0.574 1.769 1.777 fft_wrap_pw1pw2_140 487 13.2 0.086 0.090 1.735 1.748 prepare_preconditioner 11 7.9 0.000 0.000 1.742 1.747 make_preconditioner 11 8.9 0.000 0.000 1.742 1.747 calculate_dm_sparse 119 9.5 0.000 0.000 1.704 1.725 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.685 1.686 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.628 1.655 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.566 1.568 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.534 1.553 potential_pw2rs 119 12.3 0.010 0.011 1.470 1.473 jit_kernel_multiply 8 15.8 0.885 1.438 0.885 1.438 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.404 1.412 grid_collocate_task_list 119 9.7 1.271 1.391 1.271 1.391 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.388 1.389 cp_fm_redistribute_end 50 14.0 0.694 1.368 0.695 1.369 cp_fm_diag_elpa_base 50 14.0 0.628 1.297 0.673 1.354 ot_diis_step 108 11.5 0.013 0.013 1.298 1.298 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.257 1.260 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.174 1.196 apply_single 119 13.6 0.000 0.000 1.174 1.195 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.170 1.183 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.122 1.130 wfi_extrapolate 11 7.9 0.001 0.001 1.116 1.116 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.840 1.091 make_images_data 4572 15.5 0.039 0.042 0.876 1.066 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.919 acc_transpose_blocks 9144 15.5 0.038 0.040 0.896 0.902 cp_fm_cholesky_invert 11 10.9 0.899 0.902 0.899 0.902 mp_alltoall_d11v 2130 13.8 0.764 0.879 0.764 0.879 multiply_cannon_sync_h2d 9144 15.5 0.715 0.808 0.715 0.808 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.787 0.789 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.715 0.769 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.735 0.742 mp_alltoall_z22v 1201 16.6 0.654 0.710 0.654 0.710 mp_allgather_i34 2286 14.5 0.256 0.710 0.256 0.710 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.332 0.690 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.653 0.656 yz_to_x 606 15.1 0.265 0.275 0.605 0.623 mp_waitany 5200 13.7 0.470 0.596 0.470 0.596 x_to_yz 595 16.2 0.275 0.286 0.590 0.595 qs_create_task_list 11 7.9 0.020 0.020 0.562 0.589 rs_pw_transfer_RS2PW_140 130 11.5 0.100 0.103 0.436 0.582 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.669000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=584.090909, yerr=7.140850 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_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 758.517760E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.130 0.131 42.149 42.149 qs_mol_dyn_low 1 2.0 0.003 0.003 41.712 41.718 qs_forces 11 3.9 0.012 0.015 41.653 41.654 qs_energies 11 4.9 0.001 0.002 39.656 39.660 scf_env_do_scf 11 5.9 0.001 0.001 33.877 33.877 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.914 25.916 velocity_verlet 10 3.0 0.002 0.002 23.716 23.721 dbcsr_multiply_generic 2286 12.5 0.099 0.102 17.467 17.635 qs_scf_new_mos 108 7.5 0.001 0.001 15.914 16.013 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.913 16.013 ot_scf_mini 108 9.5 0.002 0.002 14.827 14.932 multiply_cannon 2286 13.5 0.303 0.313 13.577 14.481 multiply_cannon_loop 2286 14.5 0.345 0.355 12.306 13.221 ot_mini 108 10.5 0.001 0.001 8.804 8.923 multiply_cannon_multrec 9144 15.5 3.374 4.912 8.522 8.590 init_scf_loop 11 6.9 0.000 0.000 7.936 7.940 rebuild_ks_matrix 119 8.3 0.000 0.000 7.458 7.602 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.458 7.602 prepare_preconditioner 11 7.9 0.000 0.000 6.953 6.966 make_preconditioner 11 8.9 0.000 0.000 6.953 6.966 qs_ot_get_derivative 108 11.5 0.001 0.001 6.786 6.890 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.753 6.882 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.495 6.844 dbcsr_mm_accdrv_process 12550 15.8 4.015 5.584 5.024 6.327 cp_fm_upper_to_full 72 14.2 3.209 4.612 3.209 4.612 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.388 4.447 calculate_rho_elec 119 8.7 0.118 0.121 4.388 4.446 sum_up_and_integrate 119 10.3 0.065 0.067 4.113 4.119 integrate_v_rspace 119 11.3 0.003 0.003 4.048 4.055 init_scf_run 11 5.9 0.000 0.001 3.684 3.684 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.684 3.684 mp_waitall_1 94719 16.7 2.469 3.511 2.469 3.511 qs_ot_get_p 119 10.4 0.001 0.001 3.253 3.397 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.522 2.909 pw_transfer 1439 11.6 0.069 0.070 2.861 2.865 dbcsr_complete_redistribute 329 12.2 0.288 0.291 1.996 2.846 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.761 2.766 make_m2s 4572 13.5 0.038 0.038 2.336 2.519 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.688 2.518 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.142 2.453 apply_single 119 13.6 0.000 0.000 2.142 2.453 make_images 4572 14.5 0.350 0.383 2.216 2.397 fft3d_ps 1201 14.6 0.595 0.604 2.386 2.389 fft_wrap_pw1pw2_140 487 13.2 0.095 0.097 2.320 2.328 density_rs2pw 119 9.7 0.004 0.004 2.268 2.316 mp_alltoall_i22 627 13.8 1.435 2.292 1.435 2.292 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.453 2.278 calculate_dm_sparse 119 9.5 0.000 0.000 2.205 2.264 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.351 2.250 calculate_first_density_matrix 1 7.0 0.001 0.001 2.158 2.218 grid_integrate_task_list 119 12.3 2.036 2.080 2.036 2.080 qs_ot_p2m_diag 50 11.0 0.044 0.044 1.998 2.000 ot_diis_step 108 11.5 0.014 0.014 1.991 1.992 qs_energies_init_hamiltonians 11 5.9 0.009 0.009 1.832 1.837 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.789 1.789 mp_sum_l 7207 12.9 1.040 1.779 1.040 1.779 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.713 1.769 jit_kernel_multiply 10 15.4 0.982 1.734 0.982 1.734 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.715 1.716 rs_pw_transfer 974 11.9 0.009 0.009 1.517 1.581 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.535 1.572 cp_fm_cholesky_invert 11 10.9 1.522 1.525 1.522 1.525 grid_collocate_task_list 119 9.7 1.469 1.518 1.469 1.518 potential_pw2rs 119 12.3 0.014 0.014 1.497 1.500 hybrid_alltoall_any 4725 16.4 0.086 0.147 1.169 1.438 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.415 1.416 cp_fm_diag_elpa_base 50 14.0 1.269 1.323 1.413 1.413 wfi_extrapolate 11 7.9 0.001 0.001 1.397 1.397 make_images_data 4572 15.5 0.043 0.046 1.125 1.365 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.267 1.324 mp_alltoall_d11v 2130 13.8 1.184 1.286 1.184 1.286 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.162 1.176 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.154 1.175 multiply_cannon_sync_h2d 9144 15.5 1.040 1.043 1.040 1.043 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.982 1.029 qs_create_task_list 11 7.9 0.032 0.042 0.985 0.995 yz_to_x 606 15.1 0.461 0.473 0.971 0.981 generate_qs_task_list 11 8.9 0.370 0.390 0.953 0.981 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.918 0.932 acc_transpose_blocks 9144 15.5 0.037 0.038 0.897 0.904 mp_alltoall_z22v 1201 16.6 0.836 0.880 0.836 0.880 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.149000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=708.363636, yerr=17.078816 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_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 504.123392E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.055 84.688 84.690 qs_mol_dyn_low 1 2.0 0.005 0.015 84.262 84.272 qs_forces 11 3.9 0.008 0.031 84.042 84.043 qs_energies 11 4.9 0.007 0.035 81.151 81.169 scf_env_do_scf 11 5.9 0.000 0.001 71.889 71.892 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 66.219 66.219 dbcsr_multiply_generic 2055 12.4 0.105 0.109 52.372 52.595 qs_scf_new_mos 99 7.5 0.000 0.001 48.619 48.775 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.618 48.775 ot_scf_mini 99 9.5 0.002 0.003 46.195 46.286 multiply_cannon 2055 13.4 0.180 0.186 42.710 43.552 velocity_verlet 10 3.0 0.004 0.010 42.714 42.717 multiply_cannon_loop 2055 14.4 1.560 1.598 41.715 42.603 ot_mini 99 10.5 0.001 0.001 28.041 28.135 qs_ot_get_derivative 99 11.5 0.001 0.001 21.223 21.311 multiply_cannon_multrec 49320 15.4 12.085 12.791 17.113 17.878 rebuild_ks_matrix 110 8.3 0.000 0.001 14.707 14.828 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.707 14.827 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.921 13.021 mp_waitall_1 220248 16.4 11.725 12.668 11.725 12.668 multiply_cannon_sync_h2d 49320 15.4 9.927 10.434 9.927 10.434 qs_ot_get_p 110 10.4 0.001 0.001 9.642 9.752 multiply_cannon_metrocomm3 49320 15.4 0.080 0.084 7.081 8.295 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.177 7.711 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.173 7.644 apply_single 110 13.6 0.000 0.001 7.173 7.644 sum_up_and_integrate 110 10.3 0.036 0.043 7.168 7.181 integrate_v_rspace 110 11.3 0.003 0.003 7.131 7.151 init_scf_run 11 5.9 0.000 0.001 6.977 6.977 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.976 6.977 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.783 6.947 calculate_rho_elec 110 8.6 0.021 0.026 6.783 6.946 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.790 6.836 ot_diis_step 99 11.5 0.006 0.006 6.532 6.532 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.349 6.373 mp_sum_l 6514 12.8 5.273 6.045 5.273 6.045 init_scf_loop 11 6.9 0.001 0.005 5.643 5.643 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.385 5.385 dbcsr_mm_accdrv_process 87628 16.1 2.089 2.207 4.906 5.300 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.902 4.904 cp_fm_redistribute_end 48 14.0 4.276 4.875 4.280 4.876 cp_fm_diag_elpa_base 48 14.0 0.589 4.727 0.592 4.758 rs_pw_transfer 902 11.9 0.012 0.013 3.852 4.356 density_rs2pw 110 9.6 0.004 0.005 3.556 4.068 make_m2s 4110 13.4 0.061 0.065 3.935 4.059 wfi_extrapolate 11 7.9 0.001 0.001 4.015 4.015 make_images 4110 14.4 0.177 0.190 3.840 3.965 calculate_dm_sparse 110 9.5 0.000 0.001 3.785 3.919 multiply_cannon_metrocomm1 49320 15.4 0.059 0.062 2.728 3.802 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.581 3.585 prepare_preconditioner 11 7.9 0.000 0.000 3.436 3.456 make_preconditioner 11 8.9 0.000 0.002 3.436 3.456 pw_transfer 1331 11.6 0.055 0.063 3.323 3.398 grid_integrate_task_list 110 12.3 3.262 3.391 3.262 3.391 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.270 3.315 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.234 3.311 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.237 3.280 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.212 3.262 fft_wrap_pw1pw2_140 451 13.1 0.171 0.191 2.753 2.837 calculate_first_density_matrix 1 7.0 0.001 0.004 2.825 2.835 fft3d_ps 1111 14.6 0.781 0.882 2.763 2.822 potential_pw2rs 110 12.3 0.006 0.007 2.683 2.705 jit_kernel_multiply 13 15.9 2.533 2.671 2.533 2.671 mp_alltoall_d11v 2046 13.8 2.104 2.657 2.104 2.657 mp_waitany 14300 13.8 1.921 2.478 1.921 2.478 acc_transpose_blocks 49320 15.4 0.226 0.235 2.233 2.322 grid_collocate_task_list 110 9.6 2.089 2.289 2.089 2.289 mp_sum_d 3879 11.9 1.583 2.247 1.583 2.247 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.930 1.945 make_images_data 4110 15.4 0.042 0.045 1.804 1.928 cp_fm_cholesky_invert 11 10.9 1.889 1.893 1.889 1.893 hybrid_alltoall_any 4261 16.3 0.081 0.479 1.566 1.818 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.690000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.727273, yerr=3.193485 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_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 588.832768E+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 1325460. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.087 0.104 70.522 70.523 qs_mol_dyn_low 1 2.0 0.003 0.004 70.073 70.083 qs_forces 11 3.9 0.052 0.402 69.976 69.977 qs_energies 11 4.9 0.003 0.018 66.066 66.069 scf_env_do_scf 11 5.9 0.001 0.001 57.110 57.114 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.324 49.338 dbcsr_multiply_generic 2055 12.4 0.113 0.117 37.787 38.013 velocity_verlet 10 3.0 0.021 0.042 36.558 36.559 qs_scf_new_mos 99 7.5 0.001 0.001 32.846 32.983 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.845 32.982 multiply_cannon 2055 13.4 0.222 0.241 30.960 32.383 ot_scf_mini 99 9.5 0.003 0.004 31.179 31.325 multiply_cannon_loop 2055 14.4 0.932 0.953 29.689 30.594 ot_mini 99 10.5 0.001 0.001 18.344 18.495 multiply_cannon_multrec 24660 15.4 7.595 8.913 13.818 15.056 rebuild_ks_matrix 110 8.3 0.000 0.001 14.086 14.263 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.024 14.086 14.262 qs_ot_get_derivative 99 11.5 0.001 0.001 12.496 12.642 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.258 12.424 mp_waitall_1 176588 16.5 7.790 10.392 7.790 10.392 multiply_cannon_sync_h2d 24660 15.4 6.987 7.968 6.987 7.968 init_scf_loop 11 6.9 0.001 0.003 7.748 7.761 multiply_cannon_metrocomm3 24660 15.4 0.069 0.071 5.191 7.611 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.563 7.210 apply_single 110 13.6 0.000 0.001 6.562 7.210 sum_up_and_integrate 110 10.3 0.051 0.057 6.853 6.885 integrate_v_rspace 110 11.3 0.002 0.003 6.802 6.834 init_scf_run 11 5.9 0.000 0.001 6.502 6.503 scf_env_initial_rho_setup 11 6.9 0.005 0.006 6.502 6.503 qs_ot_get_p 110 10.4 0.001 0.001 6.228 6.385 dbcsr_mm_accdrv_process 52282 16.1 4.761 5.601 6.063 6.346 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.264 6.274 calculate_rho_elec 110 8.6 0.039 0.047 6.264 6.274 ot_diis_step 99 11.5 0.010 0.010 5.785 5.786 prepare_preconditioner 11 7.9 0.000 0.000 5.659 5.677 make_preconditioner 11 8.9 0.000 0.001 5.659 5.677 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.708 5.413 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.246 5.403 make_m2s 4110 13.4 0.057 0.061 4.318 4.742 make_images 4110 14.4 0.395 0.438 4.209 4.632 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.279 4.299 pw_transfer 1331 11.6 0.066 0.075 3.942 4.106 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.834 4.002 density_rs2pw 110 9.6 0.004 0.004 3.380 3.873 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.851 3.851 rs_pw_transfer 902 11.9 0.012 0.013 3.023 3.571 wfi_extrapolate 11 7.9 0.001 0.001 3.540 3.541 fft_wrap_pw1pw2_140 451 13.1 0.203 0.218 3.283 3.455 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.356 3.357 grid_integrate_task_list 110 12.3 3.145 3.352 3.145 3.352 fft3d_ps 1111 14.6 1.105 1.328 3.182 3.337 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.302 3.304 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.229 3.296 cp_fm_redistribute_end 48 14.0 2.470 3.276 2.472 3.277 cp_fm_diag_elpa_base 48 14.0 0.770 3.134 0.802 3.221 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.981 3.049 calculate_dm_sparse 110 9.5 0.001 0.001 2.973 3.004 calculate_first_density_matrix 1 7.0 0.001 0.004 2.862 2.865 make_images_data 4110 15.4 0.046 0.050 2.458 2.863 hybrid_alltoall_any 4261 16.3 0.102 0.436 2.137 2.755 potential_pw2rs 110 12.3 0.008 0.008 2.640 2.671 cp_fm_cholesky_invert 11 10.9 2.661 2.667 2.661 2.667 mp_sum_l 6514 12.8 1.894 2.557 1.894 2.557 grid_collocate_task_list 110 9.6 2.052 2.475 2.052 2.475 mp_alltoall_d11v 2046 13.8 1.754 2.058 1.754 2.058 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.994 2.018 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.969 1.992 qs_energies_init_hamiltonians 11 5.9 0.015 0.020 1.894 1.895 multiply_cannon_metrocomm4 22605 15.4 0.072 0.076 0.803 1.816 mp_waitany 10164 13.8 1.227 1.775 1.227 1.775 jit_kernel_multiply 9 16.3 0.954 1.715 0.954 1.715 mp_allgather_i34 2055 14.4 0.592 1.710 0.592 1.710 mp_irecv_dv 57340 16.2 0.680 1.600 0.680 1.600 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.538 1.549 acc_transpose_blocks 24660 15.4 0.109 0.112 1.526 1.549 mp_alltoall_z22v 1111 16.6 1.399 1.531 1.399 1.531 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.217 0.962 1.496 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.365 1.474 dbcsr_complete_redistribute 325 12.2 0.247 0.337 1.145 1.414 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.523000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.363636, yerr=7.214540 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 662.814720E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.068 0.181 61.621 61.622 qs_mol_dyn_low 1 2.0 0.003 0.004 61.066 61.076 qs_forces 11 3.9 0.002 0.004 61.000 61.002 qs_energies 11 4.9 0.002 0.007 57.773 57.778 scf_env_do_scf 11 5.9 0.001 0.001 49.597 49.597 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.634 40.634 velocity_verlet 10 3.0 0.001 0.002 33.367 33.369 dbcsr_multiply_generic 2055 12.4 0.107 0.112 28.654 28.945 qs_scf_new_mos 99 7.5 0.001 0.001 25.295 25.423 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.295 25.423 ot_scf_mini 99 9.5 0.002 0.003 24.055 24.179 multiply_cannon 2055 13.4 0.213 0.225 22.162 23.413 multiply_cannon_loop 2055 14.4 0.620 0.643 20.968 22.001 ot_mini 99 10.5 0.001 0.001 13.722 13.846 rebuild_ks_matrix 110 8.3 0.000 0.000 12.577 12.725 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 12.577 12.725 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.087 11.219 multiply_cannon_multrec 16440 15.4 3.812 4.764 9.555 10.381 mp_waitall_1 139946 16.5 7.275 10.213 7.275 10.213 qs_ot_get_derivative 99 11.5 0.001 0.001 9.211 9.335 init_scf_loop 11 6.9 0.001 0.007 8.928 8.929 multiply_cannon_metrocomm3 16440 15.4 0.043 0.045 4.558 7.347 prepare_preconditioner 11 7.9 0.000 0.000 7.105 7.126 make_preconditioner 11 8.9 0.000 0.001 7.105 7.126 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.376 6.747 sum_up_and_integrate 110 10.3 0.059 0.061 6.674 6.689 integrate_v_rspace 110 11.3 0.003 0.003 6.614 6.630 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.124 6.135 calculate_rho_elec 110 8.6 0.058 0.059 6.124 6.135 init_scf_run 11 5.9 0.000 0.001 5.729 5.729 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.729 5.729 dbcsr_mm_accdrv_process 34862 16.1 4.607 5.292 5.599 5.706 qs_ot_get_p 110 10.4 0.002 0.009 5.503 5.645 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.016 5.432 apply_single 110 13.6 0.000 0.000 5.016 5.432 make_m2s 4110 13.4 0.050 0.052 4.249 4.585 density_rs2pw 110 9.6 0.004 0.005 3.219 4.479 ot_diis_step 99 11.5 0.011 0.011 4.478 4.478 make_images 4110 14.4 0.389 0.509 4.133 4.470 multiply_cannon_sync_h2d 16440 15.4 3.631 4.178 3.631 4.178 rs_pw_transfer 902 11.9 0.010 0.011 2.798 4.007 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.875 3.879 pw_transfer 1331 11.6 0.066 0.073 3.812 3.820 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.112 3.764 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.704 3.716 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.528 3.529 grid_integrate_task_list 110 12.3 3.161 3.371 3.161 3.371 fft_wrap_pw1pw2_140 451 13.1 0.213 0.217 3.179 3.191 wfi_extrapolate 11 7.9 0.001 0.001 3.016 3.016 fft3d_ps 1111 14.6 1.087 1.098 3.004 3.016 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.011 3.013 cp_fm_redistribute_end 48 14.0 1.880 2.989 1.883 2.989 cp_fm_diag_elpa_base 48 14.0 1.041 2.837 1.101 2.943 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.922 2.924 make_images_data 4110 15.4 0.043 0.047 2.466 2.888 cp_fm_cholesky_invert 11 10.9 2.860 2.866 2.860 2.866 hybrid_alltoall_any 4261 16.3 0.105 0.373 2.186 2.799 calculate_first_density_matrix 1 7.0 0.001 0.004 2.620 2.621 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.545 2.608 calculate_dm_sparse 110 9.5 0.001 0.001 2.517 2.544 mp_waitany 17072 13.8 1.248 2.503 1.248 2.503 qs_ot_get_derivative_diag 47 12.0 0.003 0.013 2.420 2.492 grid_collocate_task_list 110 9.6 2.081 2.475 2.081 2.475 multiply_cannon_metrocomm4 14385 15.4 0.044 0.049 0.896 2.445 potential_pw2rs 110 12.3 0.011 0.011 2.392 2.399 mp_sum_l 6514 12.8 1.698 2.383 1.698 2.383 mp_irecv_dv 48980 15.7 0.826 2.321 0.826 2.321 mp_alltoall_d11v 2046 13.8 1.881 2.318 1.881 2.318 rs_pw_transfer_RS2PW_140 121 11.5 0.177 0.181 0.928 2.135 dbcsr_complete_redistribute 325 12.2 0.330 0.395 1.568 2.062 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.998 1.999 cp_fm_upper_to_full 70 13.6 1.394 1.867 1.394 1.867 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.692 1.711 cp_fm_cholesky_decompose 22 10.9 1.650 1.670 1.650 1.670 mp_allgather_i34 2055 14.4 0.517 1.598 0.517 1.598 copy_fm_to_dbcsr 174 11.2 0.001 0.002 1.081 1.583 jit_kernel_multiply 8 16.3 0.605 1.490 0.605 1.490 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.458 1.473 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.356 1.462 rs_gather_matrices 110 12.3 0.233 0.259 0.984 1.366 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.280 1.286 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.622000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.636364, yerr=9.650932 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 736.866304E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.062 67.112 67.115 qs_mol_dyn_low 1 2.0 0.004 0.004 66.580 66.589 qs_forces 11 3.9 0.002 0.004 66.481 66.482 qs_energies 11 4.9 0.025 0.053 63.028 63.031 scf_env_do_scf 11 5.9 0.001 0.001 54.391 54.394 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 42.382 42.406 velocity_verlet 10 3.0 0.001 0.002 37.746 37.749 dbcsr_multiply_generic 2055 12.4 0.115 0.119 29.966 30.181 qs_scf_new_mos 99 7.5 0.001 0.001 26.878 27.019 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.877 27.018 ot_scf_mini 99 9.5 0.003 0.004 25.248 25.385 multiply_cannon 2055 13.4 0.247 0.263 22.655 24.035 multiply_cannon_loop 2055 14.4 0.888 0.905 21.176 21.718 ot_mini 99 10.5 0.001 0.001 14.442 14.599 multiply_cannon_multrec 24660 15.4 4.236 6.764 12.666 13.907 rebuild_ks_matrix 110 8.3 0.000 0.000 12.528 12.634 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.528 12.633 init_scf_loop 11 6.9 0.002 0.005 11.964 12.011 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.062 11.157 qs_ot_get_derivative 99 11.5 0.001 0.001 10.275 10.417 prepare_preconditioner 11 7.9 0.000 0.000 10.099 10.114 make_preconditioner 11 8.9 0.000 0.001 10.099 10.114 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.335 9.776 dbcsr_mm_accdrv_process 52304 16.0 6.918 8.210 8.283 9.222 sum_up_and_integrate 110 10.3 0.066 0.069 6.692 6.707 integrate_v_rspace 110 11.3 0.003 0.003 6.626 6.643 mp_waitall_1 121746 16.5 4.669 6.557 4.669 6.557 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.350 6.358 calculate_rho_elec 110 8.6 0.077 0.081 6.350 6.357 make_m2s 4110 13.4 0.060 0.062 5.647 5.988 qs_ot_get_p 110 10.4 0.001 0.001 5.809 5.975 make_images 4110 14.4 0.578 0.721 5.506 5.842 init_scf_run 11 5.9 0.000 0.001 5.734 5.735 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.734 5.735 cp_fm_upper_to_full 70 13.8 3.258 4.618 3.258 4.618 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.033 4.125 apply_single 110 13.6 0.000 0.000 4.033 4.125 ot_diis_step 99 11.5 0.011 0.011 4.122 4.123 pw_transfer 1331 11.6 0.066 0.075 4.025 4.055 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.918 3.952 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.931 3.945 dbcsr_complete_redistribute 325 12.2 0.421 0.474 2.722 3.853 density_rs2pw 110 9.6 0.004 0.004 3.236 3.748 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.446 3.446 grid_integrate_task_list 110 12.3 3.267 3.426 3.267 3.426 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.357 3.424 fft_wrap_pw1pw2_140 451 13.1 0.202 0.211 3.344 3.382 multiply_cannon_sync_h2d 24660 15.4 3.192 3.336 3.192 3.336 copy_fm_to_dbcsr 174 11.2 0.001 0.003 2.179 3.296 make_images_data 4110 15.4 0.046 0.051 2.907 3.291 hybrid_alltoall_any 4261 16.3 0.119 0.459 2.415 3.250 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.534 3.246 fft3d_ps 1111 14.6 1.095 1.133 3.211 3.235 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.068 3.134 wfi_extrapolate 11 7.9 0.001 0.001 3.075 3.075 rs_pw_transfer 902 11.9 0.010 0.011 2.530 3.072 calculate_dm_sparse 110 9.5 0.001 0.001 2.946 2.970 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.931 2.936 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.902 2.904 cp_fm_cholesky_invert 11 10.9 2.869 2.878 2.869 2.878 cp_fm_redistribute_end 48 14.0 1.452 2.875 1.453 2.876 cp_fm_diag_elpa_base 48 14.0 1.342 2.740 1.420 2.851 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.754 2.847 mp_alltoall_i22 605 13.7 1.653 2.801 1.653 2.801 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.513 2.561 calculate_first_density_matrix 1 7.0 0.001 0.003 2.558 2.561 grid_collocate_task_list 110 9.6 2.184 2.448 2.184 2.448 qs_energies_init_hamiltonians 11 5.9 0.039 0.061 2.342 2.344 potential_pw2rs 110 12.3 0.012 0.013 2.320 2.329 mp_alltoall_d11v 2046 13.8 1.866 2.308 1.866 2.308 jit_kernel_multiply 12 15.8 1.037 2.185 1.037 2.185 cp_fm_cholesky_decompose 22 10.9 1.730 1.791 1.730 1.791 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.710 1.747 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.713 1.724 mp_allgather_i34 2055 14.4 0.582 1.715 0.582 1.715 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.597 1.693 mp_waitany 13376 13.8 1.078 1.626 1.078 1.626 multiply_cannon_metrocomm4 20550 15.4 0.056 0.060 0.841 1.590 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.553 1.566 mp_sum_l 6514 12.8 0.986 1.564 0.986 1.564 acc_transpose_blocks 24660 15.4 0.105 0.107 1.520 1.544 mp_irecv_dv 62702 16.1 0.744 1.513 0.744 1.513 rs_pw_transfer_RS2PW_140 121 11.5 0.169 0.177 0.820 1.363 mp_alltoall_z22v 1111 16.6 1.257 1.359 1.257 1.359 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.115000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=696.090909, yerr=11.673237 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_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 858.038272E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.049 55.963 55.963 qs_mol_dyn_low 1 2.0 0.003 0.003 55.678 55.687 qs_forces 11 3.9 0.005 0.006 55.615 55.616 qs_energies 11 4.9 0.011 0.014 51.906 51.913 scf_env_do_scf 11 5.9 0.000 0.001 43.264 43.264 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.451 35.452 velocity_verlet 10 3.0 0.002 0.002 31.371 31.374 dbcsr_multiply_generic 2055 12.4 0.104 0.106 22.620 22.726 qs_scf_new_mos 99 7.5 0.001 0.001 20.125 20.155 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.124 20.155 ot_scf_mini 99 9.5 0.002 0.002 18.891 18.907 multiply_cannon 2055 13.4 0.246 0.259 17.154 18.561 multiply_cannon_loop 2055 14.4 0.323 0.334 15.776 16.069 rebuild_ks_matrix 110 8.3 0.000 0.000 12.112 12.134 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.112 12.133 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.758 10.776 ot_mini 99 10.5 0.001 0.001 10.098 10.109 multiply_cannon_multrec 8220 15.4 3.191 4.631 7.524 8.520 init_scf_loop 11 6.9 0.000 0.000 7.766 7.767 mp_waitall_1 103326 16.6 5.822 7.427 5.822 7.427 sum_up_and_integrate 110 10.3 0.079 0.081 6.626 6.639 integrate_v_rspace 110 11.3 0.003 0.003 6.547 6.559 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.426 6.437 calculate_rho_elec 110 8.6 0.115 0.116 6.425 6.436 qs_ot_get_derivative 99 11.5 0.001 0.001 6.296 6.312 prepare_preconditioner 11 7.9 0.000 0.000 6.065 6.069 make_preconditioner 11 8.9 0.000 0.000 6.065 6.069 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.638 5.725 init_scf_run 11 5.9 0.000 0.001 5.439 5.439 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.439 5.439 dbcsr_mm_accdrv_process 17442 15.9 2.884 4.110 4.204 5.152 qs_ot_get_p 110 10.4 0.001 0.001 4.899 4.920 make_m2s 4110 13.4 0.039 0.040 4.220 4.500 make_images 4110 14.4 0.633 0.685 4.090 4.370 pw_transfer 1331 11.6 0.067 0.074 4.211 4.234 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 2.860 4.157 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.102 4.129 ot_diis_step 99 11.5 0.012 0.012 3.781 3.781 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.700 3.729 apply_single 110 13.6 0.000 0.001 3.700 3.729 fft_wrap_pw1pw2_140 451 13.1 0.215 0.218 3.550 3.584 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.561 3.565 density_rs2pw 110 9.6 0.004 0.004 3.158 3.530 grid_integrate_task_list 110 12.3 3.356 3.508 3.356 3.508 fft3d_ps 1111 14.6 1.142 1.185 3.348 3.369 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.252 3.253 multiply_cannon_sync_h2d 8220 15.4 2.921 3.032 2.921 3.032 cp_fm_cholesky_invert 11 10.9 2.928 2.932 2.928 2.932 make_images_data 4110 15.4 0.038 0.042 2.417 2.835 hybrid_alltoall_any 4261 16.3 0.199 0.860 2.354 2.819 wfi_extrapolate 11 7.9 0.001 0.001 2.783 2.783 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.741 2.741 cp_fm_redistribute_end 48 14.0 0.697 2.717 0.701 2.718 qs_energies_init_hamiltonians 11 5.9 0.006 0.010 2.697 2.699 cp_fm_diag_elpa_base 48 14.0 1.835 2.523 2.010 2.691 calculate_dm_sparse 110 9.5 0.001 0.001 2.569 2.607 rs_pw_transfer 902 11.9 0.010 0.010 2.202 2.598 calculate_first_density_matrix 1 7.0 0.000 0.001 2.554 2.556 grid_collocate_task_list 110 9.6 2.276 2.532 2.276 2.532 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.514 2.515 potential_pw2rs 110 12.3 0.015 0.015 2.169 2.175 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.143 2.152 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.766 1.984 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.860 1.872 mp_alltoall_d11v 2046 13.8 1.646 1.862 1.646 1.862 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.814 1.825 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.690 1.693 cp_fm_cholesky_decompose 22 10.9 1.677 1.690 1.677 1.690 mp_allgather_i34 2055 14.4 0.554 1.673 0.554 1.673 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.527 1.653 jit_kernel_multiply 9 15.8 1.014 1.585 1.014 1.585 dbcsr_complete_redistribute 325 12.2 0.588 0.611 1.468 1.565 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.538 1.553 mp_waitany 9240 13.8 1.068 1.487 1.068 1.487 qs_create_task_list 11 7.9 0.000 0.001 1.228 1.333 generate_qs_task_list 11 8.9 0.379 0.446 1.228 1.333 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.168 0.798 1.201 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.133 1.156 mp_alltoall_z22v 1111 16.6 1.092 1.132 1.092 1.132 x_to_yz 550 16.1 0.573 0.583 1.103 1.120 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.963000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=811.636364, yerr=10.965008 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_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 1.349657E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.044 88.478 88.478 qs_mol_dyn_low 1 2.0 0.003 0.004 88.054 88.061 qs_forces 11 3.9 0.002 0.002 87.968 87.969 qs_energies 11 4.9 0.001 0.001 83.778 83.779 scf_env_do_scf 11 5.9 0.001 0.001 73.775 73.775 velocity_verlet 10 3.0 0.025 0.029 56.638 56.643 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 44.954 44.955 init_scf_loop 11 6.9 0.000 0.000 28.749 28.751 dbcsr_multiply_generic 2055 12.4 0.118 0.120 28.674 28.716 prepare_preconditioner 11 7.9 0.000 0.000 26.632 26.642 make_preconditioner 11 8.9 0.000 0.000 26.632 26.641 qs_scf_new_mos 99 7.5 0.001 0.001 26.310 26.345 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.309 26.345 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.951 26.095 ot_scf_mini 99 9.5 0.002 0.002 24.546 24.572 multiply_cannon 2055 13.4 0.343 0.385 21.477 22.313 multiply_cannon_loop 2055 14.4 0.342 0.346 19.654 20.037 cp_fm_upper_to_full 70 14.2 12.618 17.999 12.618 17.999 rebuild_ks_matrix 110 8.3 0.001 0.001 14.463 14.497 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.462 14.496 ot_mini 99 10.5 0.001 0.001 13.656 13.679 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.108 13.139 dbcsr_complete_redistribute 325 12.2 1.029 1.052 7.276 10.370 multiply_cannon_multrec 8220 15.4 4.364 4.523 9.531 9.683 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.225 9.323 qs_ot_get_derivative 99 11.5 0.001 0.001 9.058 9.085 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.665 8.738 mp_waitall_1 84994 16.7 7.746 8.549 7.746 8.549 mp_alltoall_i22 605 13.7 5.268 8.354 5.268 8.354 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.889 7.931 calculate_rho_elec 110 8.6 0.227 0.228 7.888 7.930 sum_up_and_integrate 110 10.3 0.150 0.151 7.526 7.546 integrate_v_rspace 110 11.3 0.004 0.004 7.376 7.396 make_m2s 4110 13.4 0.043 0.044 5.459 5.994 make_images 4110 14.4 0.872 0.916 5.271 5.805 init_scf_run 11 5.9 0.000 0.001 5.718 5.718 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.718 5.718 cp_fm_cholesky_invert 11 10.9 5.661 5.665 5.661 5.665 qs_ot_get_p 110 10.4 0.001 0.001 5.634 5.654 pw_transfer 1331 11.6 0.076 0.076 5.435 5.459 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.318 5.342 dbcsr_mm_accdrv_process 11614 15.7 3.343 3.691 5.025 5.302 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.714 5.171 apply_single 110 13.6 0.000 0.000 4.713 5.171 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.759 5.163 fft_wrap_pw1pw2_140 451 13.1 0.222 0.223 4.617 4.645 ot_diis_step 99 11.5 0.015 0.016 4.574 4.574 fft3d_ps 1111 14.6 1.295 1.304 4.512 4.534 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.095 4.103 multiply_cannon_sync_h2d 8220 15.4 3.951 3.957 3.951 3.957 density_rs2pw 110 9.6 0.004 0.004 3.901 3.943 hybrid_alltoall_any 4261 16.3 0.257 0.557 3.046 3.869 make_images_data 4110 15.4 0.042 0.044 3.077 3.769 grid_integrate_task_list 110 12.3 3.674 3.745 3.674 3.745 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.689 3.689 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.649 3.650 wfi_extrapolate 11 7.9 0.001 0.001 3.359 3.359 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.873 3.330 calculate_dm_sparse 110 9.5 0.001 0.001 3.141 3.158 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.035 3.035 cp_fm_diag_elpa_base 48 14.0 2.506 2.695 3.033 3.034 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.837 2.839 grid_collocate_task_list 110 9.6 2.642 2.661 2.642 2.661 potential_pw2rs 110 12.3 0.021 0.021 2.595 2.616 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.415 2.430 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.348 2.360 rs_pw_transfer 902 11.9 0.011 0.011 2.321 2.353 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.224 2.284 calculate_first_density_matrix 1 7.0 0.001 0.001 2.235 2.236 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.087 2.145 mp_alltoall_d11v 2046 13.8 2.024 2.141 2.024 2.141 cp_fm_cholesky_decompose 22 10.9 2.117 2.138 2.117 2.138 qs_create_task_list 11 7.9 0.002 0.015 1.916 1.965 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.962 1.965 generate_qs_task_list 11 8.9 0.736 0.791 1.913 1.965 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.790 1.828 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.478000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1204.181818, yerr=62.512280 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 628.178944E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 2520958. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.039 204.629 204.630 qs_mol_dyn_low 1 2.0 0.003 0.015 204.112 204.125 qs_forces 11 3.9 0.046 0.075 204.014 204.016 qs_energies 11 4.9 0.051 0.060 198.293 198.332 scf_env_do_scf 11 5.9 0.001 0.001 180.642 180.646 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 159.764 159.766 dbcsr_multiply_generic 2507 12.6 0.178 0.195 122.585 123.495 velocity_verlet 10 3.0 0.004 0.023 123.122 123.123 qs_scf_new_mos 117 7.6 0.001 0.001 119.337 119.668 qs_scf_loop_do_ot 117 8.6 0.001 0.001 119.336 119.667 ot_scf_mini 117 9.6 0.003 0.003 112.769 113.074 multiply_cannon 2507 13.6 0.236 0.249 100.227 101.992 multiply_cannon_loop 2507 14.6 2.081 2.197 98.049 99.851 ot_mini 117 10.6 0.001 0.001 64.485 64.858 multiply_cannon_multrec 60168 15.6 33.404 36.472 41.625 43.598 qs_ot_get_derivative 117 11.6 0.001 0.002 39.797 40.096 rebuild_ks_matrix 128 8.3 0.001 0.001 34.004 34.336 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.043 34.004 34.335 mp_waitall_1 267128 16.5 27.311 31.231 27.311 31.231 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.532 30.829 multiply_cannon_sync_h2d 60168 15.6 27.739 29.917 27.739 29.917 qs_ot_get_p 128 10.4 0.001 0.001 26.598 26.962 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.218 25.004 apply_single 128 13.6 0.001 0.001 24.217 25.004 ot_diis_step 117 11.6 0.008 0.008 24.374 24.375 init_scf_loop 11 6.9 0.000 0.000 20.800 20.801 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.056 20.119 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.130 18.362 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.640 17.641 multiply_cannon_metrocomm3 60168 15.6 0.118 0.126 15.036 17.203 prepare_preconditioner 11 7.9 0.000 0.000 16.167 16.227 make_preconditioner 11 8.9 0.000 0.000 16.167 16.227 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.399 15.582 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.825 14.831 cp_fm_redistribute_end 83 14.4 11.718 14.751 11.732 14.757 sum_up_and_integrate 128 10.3 0.089 0.108 14.476 14.508 cp_fm_diag_elpa_base 83 14.4 2.975 14.380 3.010 14.498 integrate_v_rspace 128 11.3 0.003 0.004 14.387 14.421 make_m2s 5014 13.6 0.104 0.113 13.815 14.276 make_images 5014 14.6 0.400 0.417 13.634 14.106 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.916 14.006 calculate_rho_elec 128 8.7 0.045 0.064 13.915 14.005 init_scf_run 11 5.9 0.000 0.001 12.853 12.853 scf_env_initial_rho_setup 11 6.9 0.003 0.003 12.853 12.853 density_rs2pw 128 9.7 0.006 0.007 7.319 10.505 cp_fm_cholesky_invert 11 10.9 9.353 9.361 9.353 9.361 rs_pw_transfer 1046 11.9 0.016 0.019 6.107 9.288 wfi_extrapolate 11 7.9 0.001 0.001 9.124 9.124 mp_sum_l 7870 13.0 7.265 8.619 7.265 8.619 calculate_dm_sparse 128 9.5 0.001 0.001 8.415 8.507 pw_transfer 1547 11.6 0.076 0.096 8.125 8.366 dbcsr_mm_accdrv_process 124484 16.2 3.132 3.443 7.769 8.358 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.046 8.195 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.919 8.169 make_images_data 5014 15.6 0.066 0.071 6.756 7.775 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.630 7.700 grid_integrate_task_list 128 12.3 7.092 7.588 7.092 7.588 multiply_cannon_metrocomm1 60168 15.6 0.089 0.095 5.719 7.436 fft_wrap_pw1pw2_140 523 13.2 0.446 0.510 6.887 7.183 hybrid_alltoall_any 5200 16.5 0.290 2.265 5.919 7.183 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.057 7.065 fft3d_ps 1291 14.7 2.109 2.895 6.707 6.897 mp_waitany 16020 13.9 2.819 6.085 2.819 6.085 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.891 5.989 grid_collocate_task_list 128 9.7 4.552 5.790 4.552 5.790 rs_pw_transfer_RS2PW_140 139 11.5 0.390 0.444 2.331 5.568 mp_alltoall_d11v 2415 14.1 4.196 5.481 4.196 5.481 potential_pw2rs 128 12.3 0.009 0.011 4.816 4.861 cp_fm_cholesky_decompose 22 10.9 4.606 4.621 4.606 4.621 mp_sum_d 4455 12.2 3.553 4.404 3.553 4.404 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.630000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.272727, yerr=6.551765 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 825.671680E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4089 57145. MP_Allreduce 11262 946. MP_Sync 168 MP_Alltoall 1955 5617896. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.033 190.112 190.112 qs_mol_dyn_low 1 2.0 0.003 0.004 189.770 189.782 qs_forces 11 3.9 0.003 0.003 189.689 189.689 qs_energies 11 4.9 0.001 0.002 182.876 182.888 scf_env_do_scf 11 5.9 0.001 0.001 166.578 166.588 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 133.293 133.296 velocity_verlet 10 3.0 0.002 0.002 119.441 119.443 dbcsr_multiply_generic 2485 12.5 0.183 0.190 96.503 97.661 qs_scf_new_mos 116 7.6 0.001 0.001 93.053 93.550 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.052 93.549 ot_scf_mini 116 9.6 0.004 0.004 88.294 88.803 multiply_cannon 2485 13.5 0.478 0.536 76.747 80.345 multiply_cannon_loop 2485 14.5 1.252 1.297 73.586 76.074 ot_mini 116 10.6 0.001 0.001 49.458 49.976 mp_waitall_1 212858 16.6 24.097 38.422 24.097 38.422 multiply_cannon_multrec 29820 15.5 22.002 26.651 31.542 36.496 rebuild_ks_matrix 127 8.3 0.001 0.001 33.405 33.871 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 33.404 33.870 init_scf_loop 11 6.9 0.000 0.000 33.193 33.194 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.043 30.474 prepare_preconditioner 11 7.9 0.000 0.000 28.756 28.818 make_preconditioner 11 8.9 0.000 0.000 28.756 28.818 multiply_cannon_metrocomm3 29820 15.5 0.094 0.098 15.438 28.370 qs_ot_get_derivative 116 11.6 0.001 0.003 27.793 28.294 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.416 27.956 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.785 22.981 apply_single 127 13.6 0.001 0.001 21.784 22.981 multiply_cannon_sync_h2d 29820 15.5 19.241 21.671 19.241 21.671 ot_diis_step 116 11.6 0.014 0.015 21.497 21.499 qs_ot_get_p 127 10.4 0.001 0.001 20.803 21.438 cp_fm_cholesky_invert 11 10.9 16.803 16.816 16.803 16.816 qs_ot_p2m_diag 82 11.4 0.185 0.214 16.043 16.085 make_m2s 4970 13.5 0.090 0.097 14.095 15.547 make_images 4970 14.5 1.134 1.324 13.885 15.336 sum_up_and_integrate 127 10.3 0.113 0.133 14.928 14.956 cp_dbcsr_syevd 82 12.4 0.005 0.006 14.905 14.907 integrate_v_rspace 127 11.3 0.003 0.004 14.814 14.850 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.242 14.274 calculate_rho_elec 127 8.7 0.087 0.105 14.241 14.274 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.814 11.820 cp_fm_redistribute_end 82 14.4 6.901 11.751 6.913 11.752 cp_fm_diag_elpa_base 82 14.4 4.613 11.254 4.823 11.611 init_scf_run 11 5.9 0.000 0.001 11.515 11.516 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.514 11.516 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.848 11.222 multiply_cannon_metrocomm4 27335 15.5 0.093 0.107 3.662 10.229 make_images_data 4970 15.5 0.063 0.071 8.471 10.145 density_rs2pw 127 9.7 0.006 0.007 7.660 10.038 mp_irecv_dv 68888 16.3 3.471 9.844 3.471 9.844 dbcsr_mm_accdrv_process 61726 16.2 4.434 5.270 9.005 9.569 hybrid_alltoall_any 5155 16.4 0.337 1.465 7.297 9.348 pw_transfer 1535 11.6 0.087 0.114 9.118 9.199 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 8.893 8.979 wfi_extrapolate 11 7.9 0.001 0.001 8.423 8.423 rs_pw_transfer 1038 11.9 0.014 0.016 5.889 8.263 fft_wrap_pw1pw2_140 519 13.2 0.469 0.524 7.804 7.912 grid_integrate_task_list 127 12.3 7.104 7.525 7.104 7.525 fft3d_ps 1281 14.7 2.729 2.889 7.276 7.305 cp_fm_cholesky_decompose 22 10.9 7.041 7.124 7.041 7.124 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.249 7.001 calculate_dm_sparse 127 9.5 0.001 0.001 6.410 6.556 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.107 6.116 grid_collocate_task_list 127 9.7 4.662 5.835 4.662 5.835 mp_sum_l 7804 13.0 3.890 5.708 3.890 5.708 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.239 5.407 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.277 5.344 potential_pw2rs 127 12.3 0.015 0.017 5.225 5.254 mp_alltoall_d11v 2401 14.1 4.224 4.908 4.224 4.908 mp_waitany 11660 13.9 2.498 4.893 2.498 4.893 mp_allgather_i34 2485 14.5 1.740 4.541 1.740 4.541 rs_pw_transfer_RS2PW_140 138 11.5 0.351 0.377 2.077 4.447 dbcsr_complete_redistribute 393 12.7 0.778 0.889 3.203 4.064 mp_sum_d 4444 12.2 2.582 3.891 2.582 3.891 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.767 3.811 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.112000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=786.818182, yerr=3.128118 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_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 944.652288E+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 4085 57197. MP_Allreduce 11251 985. MP_Sync 170 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 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 0.053 0.103 180.088 180.089 qs_mol_dyn_low 1 2.0 0.003 0.005 179.564 179.576 qs_forces 11 3.9 0.007 0.030 179.402 179.413 qs_energies 11 4.9 0.010 0.039 172.663 172.682 scf_env_do_scf 11 5.9 0.004 0.015 156.958 156.960 scf_env_do_scf_inner_loop 117 6.6 0.004 0.008 120.991 120.992 velocity_verlet 10 3.0 0.002 0.010 115.472 115.476 dbcsr_multiply_generic 2507 12.6 0.181 0.185 81.837 82.845 qs_scf_new_mos 117 7.6 0.001 0.001 82.009 82.339 qs_scf_loop_do_ot 117 8.6 0.001 0.001 82.008 82.338 ot_scf_mini 117 9.6 0.004 0.006 77.882 78.263 multiply_cannon 2507 13.6 0.499 0.522 61.891 66.029 multiply_cannon_loop 2507 14.6 0.862 0.891 58.783 61.394 ot_mini 117 10.6 0.002 0.008 42.844 43.244 init_scf_loop 11 6.9 0.002 0.008 35.840 35.842 mp_waitall_1 170520 16.6 25.647 35.217 25.647 35.217 rebuild_ks_matrix 128 8.3 0.001 0.001 31.345 31.784 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.041 31.344 31.784 prepare_preconditioner 11 7.9 0.000 0.000 31.711 31.757 make_preconditioner 11 8.9 0.000 0.002 31.711 31.757 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.291 30.721 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.204 28.614 multiply_cannon_metrocomm3 20056 15.6 0.061 0.066 15.714 25.336 multiply_cannon_multrec 20056 15.6 13.377 16.119 21.969 24.655 qs_ot_get_derivative 117 11.6 0.005 0.028 22.856 23.242 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.122 21.138 apply_single 128 13.6 0.001 0.001 20.121 21.137 qs_ot_get_p 128 10.4 0.001 0.001 20.330 20.840 ot_diis_step 117 11.6 0.018 0.022 19.883 19.884 qs_ot_p2m_diag 83 11.4 0.266 0.273 16.141 16.151 make_m2s 5014 13.6 0.082 0.087 14.920 15.859 multiply_cannon_sync_h2d 20056 15.6 14.156 15.804 14.156 15.804 make_images 5014 14.6 1.155 1.244 14.686 15.620 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.118 15.120 sum_up_and_integrate 128 10.3 0.131 0.141 15.002 15.030 integrate_v_rspace 128 11.3 0.004 0.010 14.871 14.900 cp_fm_cholesky_invert 11 10.9 14.845 14.856 14.845 14.856 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.689 14.731 calculate_rho_elec 128 8.7 0.132 0.147 14.689 14.730 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.877 11.878 cp_fm_redistribute_end 83 14.4 4.490 11.807 4.507 11.808 cp_fm_diag_elpa_base 83 14.4 6.862 11.197 7.279 11.667 make_images_data 5014 15.6 0.061 0.069 9.269 10.607 init_scf_run 11 5.9 0.000 0.001 10.573 10.573 scf_env_initial_rho_setup 11 6.9 0.001 0.006 10.572 10.573 hybrid_alltoall_any 5200 16.5 0.433 1.986 7.930 9.679 multiply_cannon_metrocomm4 17549 15.6 0.061 0.072 3.518 9.502 pw_transfer 1547 11.6 0.087 0.108 9.379 9.493 density_rs2pw 128 9.7 0.006 0.006 7.532 9.402 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.153 9.274 mp_irecv_dv 50230 16.2 3.395 9.251 3.395 9.251 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.948 9.228 fft_wrap_pw1pw2_140 523 13.2 0.478 0.524 8.047 8.178 dbcsr_mm_accdrv_process 41502 16.2 4.370 5.243 8.055 8.171 grid_integrate_task_list 128 12.3 7.279 7.800 7.279 7.800 cp_fm_upper_to_full 105 14.5 5.805 7.666 5.805 7.666 cp_fm_cholesky_decompose 22 10.9 7.585 7.621 7.585 7.621 wfi_extrapolate 11 7.9 0.001 0.001 7.530 7.530 fft3d_ps 1291 14.7 2.668 2.892 7.417 7.499 rs_pw_transfer 1046 11.9 0.014 0.015 5.432 7.323 dbcsr_complete_redistribute 395 12.7 1.166 1.201 4.726 6.544 calculate_dm_sparse 128 9.5 0.001 0.001 5.751 5.874 grid_collocate_task_list 128 9.7 4.907 5.755 4.907 5.755 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.440 5.446 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.488 5.303 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.591 5.290 mp_alltoall_d11v 2415 14.1 4.488 5.260 4.488 5.260 potential_pw2rs 128 12.3 0.020 0.022 5.050 5.071 mp_allgather_i34 2507 14.6 1.629 4.798 1.629 4.798 mp_sum_l 7870 13.0 3.257 4.762 3.257 4.762 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.541 4.688 mp_waitany 11748 13.9 2.422 4.360 2.422 4.360 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.400 4.162 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.054 4.093 mp_alltoall_i22 716 14.1 2.007 3.893 2.007 3.893 rs_pw_transfer_RS2PW_140 139 11.5 0.333 0.357 1.940 3.828 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.784 3.786 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.658 3.698 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=180.089000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=889.272727, yerr=11.209500 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 4.320337E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.141264E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837542000 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4077 57301. MP_Allreduce 11231 1068. MP_Sync 168 MP_Alltoall 1700 12496381. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.033 188.189 188.190 qs_mol_dyn_low 1 2.0 0.003 0.004 187.831 187.843 qs_forces 11 3.9 0.003 0.003 187.720 187.727 qs_energies 11 4.9 0.002 0.002 180.475 180.486 scf_env_do_scf 11 5.9 0.001 0.001 163.784 163.791 velocity_verlet 10 3.0 0.002 0.003 124.973 124.976 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 116.786 116.787 qs_scf_new_mos 116 7.6 0.001 0.001 79.144 79.453 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.144 79.452 dbcsr_multiply_generic 2485 12.5 0.187 0.193 78.724 79.405 ot_scf_mini 116 9.6 0.003 0.004 74.682 74.999 multiply_cannon 2485 13.5 0.551 0.577 54.235 57.279 multiply_cannon_loop 2485 14.5 1.179 1.200 50.609 52.334 init_scf_loop 11 6.9 0.000 0.000 46.872 46.873 prepare_preconditioner 11 7.9 0.000 0.000 42.656 42.690 make_preconditioner 11 8.9 0.000 0.000 42.656 42.690 ot_mini 116 10.6 0.001 0.001 42.032 42.350 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.234 41.264 multiply_cannon_multrec 29820 15.5 13.803 19.226 25.650 30.843 rebuild_ks_matrix 127 8.3 0.001 0.001 30.025 30.298 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 30.024 30.297 mp_waitall_1 146592 16.7 17.519 27.999 17.519 27.999 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.996 27.232 qs_ot_get_derivative 116 11.6 0.001 0.002 22.526 22.850 make_m2s 4970 13.5 0.096 0.100 20.176 21.308 make_images 4970 14.5 1.919 2.191 19.870 21.009 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.988 19.519 apply_single 127 13.6 0.001 0.001 18.988 19.519 qs_ot_get_p 127 10.4 0.001 0.001 19.033 19.389 ot_diis_step 116 11.6 0.017 0.018 19.385 19.388 cp_fm_upper_to_full 104 14.7 11.098 16.418 11.098 16.418 cp_fm_cholesky_invert 11 10.9 16.394 16.403 16.394 16.403 multiply_cannon_metrocomm3 29820 15.5 0.047 0.049 6.628 15.303 qs_ot_p2m_diag 82 11.4 0.339 0.385 14.984 15.035 sum_up_and_integrate 127 10.3 0.139 0.152 14.851 14.876 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.720 14.764 calculate_rho_elec 127 8.7 0.173 0.189 14.719 14.764 integrate_v_rspace 127 11.3 0.003 0.004 14.712 14.742 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.718 13.721 dbcsr_complete_redistribute 393 12.7 1.503 1.627 9.192 12.966 make_images_data 4970 15.5 0.063 0.068 10.961 12.895 multiply_cannon_sync_h2d 29820 15.5 11.637 12.458 11.637 12.458 hybrid_alltoall_any 5155 16.4 0.520 2.192 9.872 12.025 dbcsr_mm_accdrv_process 61748 16.2 7.259 8.207 11.422 11.928 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.792 11.570 init_scf_run 11 5.9 0.000 0.001 10.814 10.815 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.814 10.815 cp_fm_diag_elpa 82 13.4 0.000 0.001 10.527 10.531 cp_fm_redistribute_end 82 14.4 1.822 10.460 1.837 10.466 cp_fm_diag_elpa_base 82 14.4 8.023 9.845 8.595 10.337 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.400 10.079 pw_transfer 1535 11.6 0.088 0.105 9.544 9.620 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 9.267 9.507 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 9.318 9.399 mp_alltoall_i22 712 14.1 5.662 9.389 5.662 9.389 density_rs2pw 127 9.7 0.006 0.006 7.185 8.498 fft_wrap_pw1pw2_140 519 13.2 0.478 0.489 8.268 8.374 grid_integrate_task_list 127 12.3 7.477 7.811 7.477 7.811 wfi_extrapolate 11 7.9 0.001 0.001 7.751 7.752 cp_fm_cholesky_decompose 22 10.9 7.612 7.691 7.612 7.691 fft3d_ps 1281 14.7 2.756 2.833 7.556 7.621 multiply_cannon_metrocomm4 24850 15.5 0.074 0.084 2.808 7.262 mp_irecv_dv 75445 16.2 2.665 7.008 2.665 7.008 calculate_dm_sparse 127 9.5 0.001 0.001 6.129 6.220 rs_pw_transfer 1038 11.9 0.013 0.014 4.714 6.080 grid_collocate_task_list 127 9.7 5.040 5.772 5.040 5.772 mp_alltoall_d11v 2401 14.1 5.043 5.749 5.043 5.749 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.425 5.467 potential_pw2rs 127 12.3 0.022 0.023 4.707 4.718 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.376 4.460 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.347 4.454 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.442 4.443 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.151 4.222 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.190000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1074.636364, yerr=19.359154 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.518932E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108613552 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3983 58480. MP_Allreduce 10955 1177. MP_Sync 86 MP_Alltoall 1700 18828162. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.032 177.250 177.250 qs_mol_dyn_low 1 2.0 0.003 0.003 176.788 176.801 qs_forces 11 3.9 0.003 0.003 176.685 176.688 qs_energies 11 4.9 0.006 0.008 169.037 169.049 scf_env_do_scf 11 5.9 0.001 0.001 151.408 151.412 velocity_verlet 10 3.0 0.011 0.014 115.401 115.405 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.059 115.061 qs_scf_new_mos 116 7.6 0.001 0.001 75.954 76.092 qs_scf_loop_do_ot 116 8.6 0.001 0.001 75.953 76.091 dbcsr_multiply_generic 2485 12.5 0.180 0.186 74.821 75.362 ot_scf_mini 116 9.6 0.004 0.004 71.565 71.688 multiply_cannon 2485 13.5 0.579 0.609 54.771 58.743 multiply_cannon_loop 2485 14.5 0.444 0.455 50.103 51.228 ot_mini 116 10.6 0.001 0.001 40.079 40.225 init_scf_loop 11 6.9 0.000 0.000 36.199 36.202 mp_waitall_1 124680 16.7 26.465 34.464 26.465 34.464 prepare_preconditioner 11 7.9 0.000 0.000 32.191 32.230 make_preconditioner 11 8.9 0.000 0.000 32.190 32.230 rebuild_ks_matrix 127 8.3 0.001 0.001 30.260 30.395 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 30.259 30.394 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.007 30.265 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.435 27.559 multiply_cannon_multrec 9940 15.5 10.320 14.875 17.823 21.954 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 12.694 20.242 qs_ot_get_derivative 116 11.6 0.001 0.002 20.049 20.183 ot_diis_step 116 11.6 0.020 0.020 19.959 19.960 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.485 19.924 apply_single 127 13.6 0.001 0.001 19.485 19.924 qs_ot_get_p 127 10.4 0.001 0.001 18.567 18.715 make_m2s 4970 13.5 0.066 0.071 16.362 18.484 cp_fm_cholesky_invert 11 10.9 18.466 18.472 18.466 18.472 make_images 4970 14.5 2.255 2.731 16.058 18.182 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.638 15.677 calculate_rho_elec 127 8.7 0.256 0.266 15.637 15.676 sum_up_and_integrate 127 10.3 0.178 0.189 15.203 15.258 integrate_v_rspace 127 11.3 0.004 0.012 15.024 15.078 qs_ot_p2m_diag 82 11.4 0.490 0.495 14.777 14.791 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.631 13.633 make_images_data 4970 15.5 0.051 0.059 10.004 12.397 hybrid_alltoall_any 5155 16.4 0.840 3.811 9.786 12.236 multiply_cannon_sync_h2d 9940 15.5 11.441 12.183 11.441 12.183 init_scf_run 11 5.9 0.000 0.001 10.801 10.802 scf_env_initial_rho_setup 11 6.9 0.016 0.022 10.801 10.802 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.565 10.577 cp_fm_diag_elpa_base 82 14.4 10.322 10.405 10.561 10.574 pw_transfer 1535 11.6 0.086 0.096 9.987 10.028 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.764 9.811 fft_wrap_pw1pw2_140 519 13.2 0.492 0.514 8.579 8.625 grid_integrate_task_list 127 12.3 7.663 8.207 7.663 8.207 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 8.006 8.099 cp_fm_cholesky_decompose 22 10.9 7.962 8.078 7.962 8.078 density_rs2pw 127 9.7 0.005 0.006 7.201 8.053 fft3d_ps 1281 14.7 2.688 2.766 7.936 7.955 multiply_cannon_metrocomm1 9940 15.5 0.028 0.029 4.456 7.931 dbcsr_mm_accdrv_process 20590 16.1 2.622 3.531 7.140 7.810 wfi_extrapolate 11 7.9 0.001 0.001 7.561 7.561 mp_allgather_i34 2485 14.5 2.886 7.128 2.886 7.128 mp_alltoall_d11v 2401 14.1 5.238 6.334 5.238 6.334 calculate_dm_sparse 127 9.5 0.001 0.001 6.118 6.221 grid_collocate_task_list 127 9.7 5.354 6.012 5.354 6.012 dbcsr_complete_redistribute 393 12.7 2.166 2.241 5.337 5.729 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.369 5.383 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 5.297 5.298 rs_pw_transfer 1038 11.9 0.012 0.013 4.281 5.163 multiply_cannon_metrocomm4 7455 15.5 0.023 0.027 1.859 5.149 mp_irecv_dv 28618 15.9 1.822 5.062 1.822 5.062 potential_pw2rs 127 12.3 0.026 0.027 4.698 4.712 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.231 4.279 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.548 3.867 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.506 3.821 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.721 3.758 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.725 3.738 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.598 3.701 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.543 3.586 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=177.250000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1417.545455, yerr=50.865891 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1389586876416 0.0% 0.0% 100.0% flops 32 x 32 x 32 1898375544832 0.0% 0.0% 100.0% flops 22 x 9 x 32 1929499951104 0.0% 0.0% 100.0% flops 9 x 22 x 32 1935094689792 0.0% 0.0% 100.0% flops 22 x 22 x 32 2675276685312 0.0% 0.0% 100.0% flops 32 x 32 x 9 4300336005120 0.0% 0.0% 100.0% flops 32 x 32 x 22 5255966228480 0.0% 0.0% 100.0% flops 9 x 32 x 32 5299016564736 0.0% 0.0% 100.0% flops 22 x 32 x 32 6476575801344 0.0% 0.0% 100.0% flops 9 x 32 x 9 11276327374848 0.0% 0.0% 100.0% flops 22 x 32 x 9 14799245598720 0.0% 0.0% 100.0% flops 9 x 32 x 22 14799245598720 0.0% 0.0% 100.0% flops 22 x 32 x 22 19338589388800 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 91.373136E+12 0.0% 0.0% 100.0% flops max/rank 11.428386E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6604820320 0.0% 0.0% 100.0% number of processed stacks 1917536 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.4 marketing flops 141.377438E+12 ------------------------------------------------------------------------------- # multiplications 2447 max memory usage/rank 3.016339E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 97880 MPI messages size (bytes): total size 1.110077E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.341202E+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 44128 34259075072 4194304 < size <= 16777216 43248 370189271040 16777216 < size 9888 705625874544 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4015 58954. MP_Allreduce 11035 1515. MP_Sync 86 MP_Alltoall 1676 36913754. MP_SendRecv 1750 218624. MP_ISendRecv 1750 218624. MP_Wait 9580 MP_ISend 6264 1080795. MP_IRecv 6264 1080795. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.056 297.331 297.332 qs_mol_dyn_low 1 2.0 0.003 0.003 296.784 296.795 qs_forces 11 3.9 0.003 0.003 296.701 296.704 qs_energies 11 4.9 0.002 0.002 287.303 287.314 scf_env_do_scf 11 5.9 0.001 0.001 265.112 265.129 velocity_verlet 10 3.0 0.002 0.002 212.741 212.748 scf_env_do_scf_inner_loop 114 6.6 0.003 0.008 139.826 139.827 init_scf_loop 11 6.9 0.000 0.000 125.033 125.034 prepare_preconditioner 11 7.9 0.000 0.000 119.925 119.948 make_preconditioner 11 8.9 0.000 0.000 119.925 119.948 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.533 117.077 qs_scf_new_mos 114 7.6 0.001 0.001 89.698 89.847 qs_scf_loop_do_ot 114 8.6 0.001 0.001 89.697 89.846 ot_scf_mini 114 9.6 0.003 0.004 85.044 85.144 dbcsr_multiply_generic 2447 12.5 0.206 0.214 81.275 81.872 cp_fm_upper_to_full 104 14.8 52.368 75.650 52.368 75.650 multiply_cannon 2447 13.5 0.690 0.728 57.094 57.738 multiply_cannon_loop 2447 14.5 0.463 0.474 53.548 55.068 ot_mini 114 10.6 0.001 0.001 43.578 43.709 dbcsr_complete_redistribute 393 12.7 3.975 4.048 30.122 43.418 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.685 39.902 rebuild_ks_matrix 125 8.3 0.001 0.001 37.893 37.968 qs_ks_build_kohn_sham_matrix 125 9.3 0.017 0.017 37.893 37.967 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.349 37.423 mp_alltoall_i22 708 14.1 22.172 35.437 22.172 35.437 qs_ks_update_qs_env 125 7.6 0.001 0.001 34.794 34.875 cp_fm_cholesky_invert 11 10.9 33.842 33.849 33.842 33.849 mp_waitall_1 101196 16.8 27.616 31.070 27.616 31.070 qs_ot_get_p 125 10.4 0.001 0.001 26.436 26.635 qs_ot_get_derivative 114 11.6 0.002 0.002 23.836 23.937 qs_ot_p2m_diag 82 11.4 0.868 0.872 22.378 22.406 cp_dbcsr_syevd 82 12.4 0.006 0.006 20.641 20.642 qs_rho_update_rho_low 125 7.7 0.001 0.001 20.481 20.513 calculate_rho_elec 125 8.7 0.472 0.472 20.480 20.513 make_m2s 4894 13.5 0.075 0.078 19.545 20.419 make_images 4894 14.5 3.632 3.825 19.076 19.952 ot_diis_step 114 11.6 0.021 0.022 19.700 19.700 sum_up_and_integrate 125 10.3 0.313 0.316 19.555 19.622 multiply_cannon_metrocomm3 9788 15.5 0.023 0.024 18.285 19.616 integrate_v_rspace 125 11.3 0.004 0.004 19.241 19.308 apply_preconditioner_dbcsr 125 12.6 0.000 0.000 18.970 19.248 apply_single 125 13.6 0.001 0.001 18.969 19.247 multiply_cannon_multrec 9788 15.5 10.205 11.980 17.473 17.532 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.390 17.391 cp_fm_diag_elpa_base 82 14.4 13.133 14.679 17.386 17.386 multiply_cannon_sync_h2d 9788 15.5 15.298 15.312 15.298 15.312 pw_transfer 1511 11.6 0.091 0.092 12.227 12.231 init_scf_run 11 5.9 0.000 0.001 12.226 12.228 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.226 12.227 make_images_data 4894 15.5 0.059 0.063 10.493 12.181 hybrid_alltoall_any 5079 16.4 1.271 2.959 10.567 12.162 fft_wrap_pw1pw2 1261 12.7 0.011 0.011 11.997 12.000 fft_wrap_pw1pw2_140 511 13.2 0.530 0.532 10.624 10.633 fft3d_ps 1261 14.7 2.676 2.691 10.079 10.086 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.375 9.441 mp_alltoall_d11v 2385 14.1 8.380 9.409 8.380 9.409 wfi_extrapolate 11 7.9 0.001 0.001 9.150 9.151 cp_fm_cholesky_decompose 22 10.9 9.011 9.033 9.011 9.033 dbcsr_mm_accdrv_process 20270 16.0 3.755 5.494 7.035 8.820 grid_integrate_task_list 125 12.3 8.392 8.561 8.392 8.561 density_rs2pw 125 9.7 0.005 0.005 8.309 8.456 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.953 7.954 calculate_dm_sparse 125 9.5 0.001 0.001 6.486 6.597 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.291 6.362 grid_collocate_task_list 125 9.7 6.217 6.248 6.217 6.248 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.139 6.188 rs_scatter_matrices 136 9.7 3.522 4.430 5.933 6.148 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=297.332000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2650.454545, yerr=179.329237 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_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 1.262248E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 6805335. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.082 85.330 85.331 qs_energies 1 2.0 0.000 0.000 84.796 84.801 ls_scf 1 3.0 0.000 0.000 83.762 83.767 dbcsr_multiply_generic 111 6.7 0.014 0.016 72.649 72.805 multiply_cannon 111 7.7 0.017 0.020 55.889 56.947 multiply_cannon_loop 111 8.7 0.214 0.229 52.482 53.660 ls_scf_main 1 4.0 0.000 0.000 52.373 52.373 density_matrix_trs4 2 5.0 0.002 0.003 46.835 46.911 ls_scf_init_scf 1 4.0 0.000 0.000 28.344 28.346 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.199 27.252 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.101 25.120 mp_waitall_1 11031 10.9 22.389 24.804 22.389 24.804 multiply_cannon_multrec 2664 9.7 8.194 8.942 15.512 17.256 multiply_cannon_sync_h2d 2664 9.7 13.757 15.186 13.757 15.186 make_m2s 222 7.7 0.008 0.012 13.162 13.610 make_images 222 8.7 0.099 0.109 13.140 13.587 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.530 12.780 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.456 8.542 make_images_data 222 9.7 0.004 0.005 7.736 8.340 dbcsr_mm_accdrv_process 4760 10.4 0.516 0.616 6.937 7.938 hybrid_alltoall_any 227 10.6 0.215 1.838 6.600 7.863 dbcsr_mm_accdrv_process_sort 4760 11.4 6.219 7.147 6.219 7.147 calculate_norms 4752 9.8 5.546 6.294 5.546 6.294 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.034 5.177 mp_sum_l 807 5.4 3.081 4.534 3.081 4.534 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.065 3.655 mp_irecv_dv 6231 10.9 2.048 3.635 2.048 3.635 make_images_sizes 222 9.7 0.000 0.000 0.748 3.585 mp_alltoall_i44 222 10.7 0.748 3.585 0.748 3.585 arnoldi_extremal 4 6.8 0.000 0.000 3.335 3.359 arnoldi_normal_ev 4 7.8 0.001 0.003 3.335 3.359 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.274 3.190 build_subspace 16 8.4 0.009 0.012 3.140 3.144 ls_scf_post 1 4.0 0.000 0.000 3.045 3.050 ls_scf_store_result 1 5.0 0.000 0.000 2.857 2.901 dbcsr_special_finalize 555 9.7 0.005 0.006 2.326 2.845 dbcsr_merge_single_wm 555 10.7 0.459 0.573 2.318 2.836 make_images_pack 222 9.7 2.206 2.633 2.208 2.635 dbcsr_matrix_vector_mult 304 9.0 0.005 0.013 2.320 2.555 dbcsr_sort_data 658 11.4 2.117 2.546 2.117 2.546 dbcsr_matrix_vector_mult_local 304 10.0 2.070 2.456 2.072 2.458 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.299 2.402 buffer_matrices_ensure_size 222 8.7 1.752 2.069 1.752 2.069 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.793 1.794 rebuild_ks_matrix 3 7.3 0.000 0.000 1.783 1.785 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.783 1.785 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.331000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1143.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_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 2.102407E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.043 90.400 90.401 qs_energies 1 2.0 0.000 0.000 89.962 89.966 ls_scf 1 3.0 0.000 0.000 88.657 88.661 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.635 75.002 multiply_cannon 111 7.7 0.029 0.042 53.040 56.630 ls_scf_main 1 4.0 0.000 0.000 54.546 54.552 multiply_cannon_loop 111 8.7 0.116 0.124 49.853 53.209 density_matrix_trs4 2 5.0 0.002 0.003 48.784 49.016 ls_scf_init_scf 1 4.0 0.000 0.000 30.600 30.601 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.407 29.498 mp_waitall_1 9105 10.9 20.690 29.479 20.690 29.479 multiply_cannon_multrec 1332 9.7 13.213 16.985 22.433 27.348 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.986 27.001 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.606 20.402 make_m2s 222 7.7 0.006 0.007 14.971 15.547 make_images 222 8.7 1.575 1.969 14.941 15.517 dbcsr_mm_accdrv_process 4041 10.4 0.293 0.446 8.823 10.379 dbcsr_mm_accdrv_process_sort 4041 11.4 8.415 9.936 8.415 9.936 make_images_data 222 9.7 0.004 0.004 8.478 9.308 hybrid_alltoall_any 227 10.6 0.523 2.489 7.931 8.950 mp_sum_l 807 5.4 5.307 8.136 5.307 8.136 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.199 7.699 mp_irecv_dv 3311 11.0 3.181 7.635 3.181 7.635 calculate_norms 2376 9.8 6.001 6.807 6.001 6.807 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.064 6.391 multiply_cannon_sync_h2d 1332 9.7 4.798 6.209 4.798 6.209 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.076 5.306 arnoldi_extremal 4 6.8 0.000 0.000 4.720 4.744 arnoldi_normal_ev 4 7.8 0.001 0.004 4.720 4.744 build_subspace 16 8.4 0.014 0.021 4.462 4.464 ls_scf_post 1 4.0 0.000 0.000 3.511 3.514 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.193 3.427 ls_scf_store_result 1 5.0 0.000 0.000 3.220 3.336 dbcsr_matrix_vector_mult_local 304 10.0 2.771 3.256 2.773 3.258 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.238 3.024 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.626 2.753 make_images_pack 222 9.7 2.027 2.431 2.029 2.433 mp_allgather_i34 111 8.7 0.894 2.146 0.894 2.146 dbcsr_sort_data 436 11.2 1.850 2.088 1.850 2.088 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.888 1.891 rebuild_ks_matrix 3 7.3 0.000 0.000 1.876 1.878 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.876 1.878 dbcsr_data_new 4174 10.1 1.597 1.841 1.597 1.841 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.401000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1748.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_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 2.716787E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.044 94.101 94.102 qs_energies 1 2.0 0.000 0.000 93.550 93.553 ls_scf 1 3.0 0.000 0.000 92.154 92.157 dbcsr_multiply_generic 111 6.7 0.016 0.019 76.817 77.091 ls_scf_main 1 4.0 0.000 0.000 57.827 57.833 multiply_cannon 111 7.7 0.043 0.110 52.857 56.461 multiply_cannon_loop 111 8.7 0.100 0.107 49.184 53.579 density_matrix_trs4 2 5.0 0.002 0.003 51.765 51.916 mp_waitall_1 7281 11.0 23.809 34.574 23.809 34.574 ls_scf_init_scf 1 4.0 0.000 0.000 30.793 30.796 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.589 29.665 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.170 27.183 multiply_cannon_multrec 888 9.7 12.643 15.457 21.176 24.589 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.926 23.725 make_m2s 222 7.7 0.006 0.007 16.945 18.290 make_images 222 8.7 1.970 2.302 16.907 18.250 hybrid_alltoall_any 227 10.6 0.620 2.874 9.333 10.964 make_images_data 222 9.7 0.003 0.004 9.734 10.871 dbcsr_mm_accdrv_process 3754 10.4 0.252 0.417 8.052 9.259 mp_sum_l 807 5.4 5.440 9.136 5.440 9.136 dbcsr_mm_accdrv_process_sort 3754 11.4 7.683 8.842 7.683 8.842 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.217 7.307 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.769 7.301 multiply_cannon_sync_h2d 888 9.7 5.998 7.168 5.998 7.168 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.435 6.701 mp_irecv_dv 2335 11.1 2.418 6.662 2.418 6.662 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.973 5.165 arnoldi_extremal 4 6.8 0.000 0.000 5.093 5.108 arnoldi_normal_ev 4 7.8 0.001 0.004 5.093 5.108 build_subspace 16 8.4 0.014 0.020 4.789 4.794 calculate_norms 1584 9.8 4.264 4.562 4.264 4.562 mp_allgather_i34 111 8.7 1.500 3.896 1.500 3.896 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.473 3.776 dbcsr_matrix_vector_mult_local 304 10.0 3.037 3.602 3.039 3.604 ls_scf_post 1 4.0 0.000 0.000 3.534 3.538 ls_scf_store_result 1 5.0 0.000 0.000 3.297 3.362 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.895 3.013 make_images_pack 222 9.7 1.822 2.136 1.825 2.139 dbcsr_sort_data 325 11.1 1.889 2.129 1.889 2.129 make_images_sizes 222 9.7 0.000 0.000 0.893 2.063 mp_alltoall_i44 222 10.7 0.892 2.062 0.892 2.062 dbcsr_data_release 9322 10.9 1.321 2.000 1.321 2.000 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.937 1.939 rebuild_ks_matrix 3 7.3 0.000 0.000 1.919 1.921 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.919 1.921 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.102000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2225.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_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 3.363795E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.043 97.082 97.083 qs_energies 1 2.0 0.000 0.000 96.372 96.377 ls_scf 1 3.0 0.000 0.000 94.714 94.721 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.490 78.753 ls_scf_main 1 4.0 0.000 0.000 58.928 58.929 multiply_cannon 111 7.7 0.052 0.145 51.782 55.651 density_matrix_trs4 2 5.0 0.002 0.003 52.807 52.915 multiply_cannon_loop 111 8.7 0.115 0.125 46.763 49.802 ls_scf_init_scf 1 4.0 0.000 0.000 32.543 32.544 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.310 31.378 mp_waitall_1 6369 11.0 23.129 30.970 23.129 30.970 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.847 28.860 multiply_cannon_multrec 1332 9.7 14.100 16.950 21.949 25.012 make_m2s 222 7.7 0.007 0.008 21.181 22.564 make_images 222 8.7 3.130 3.580 21.130 22.516 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.600 17.301 make_images_data 222 9.7 0.004 0.004 11.872 13.479 hybrid_alltoall_any 227 10.6 0.796 3.810 11.230 12.924 dbcsr_mm_accdrv_process 3641 10.4 0.204 0.407 7.494 9.015 dbcsr_mm_accdrv_process_sort 3641 11.4 7.121 8.600 7.121 8.600 mp_sum_l 807 5.4 3.939 6.782 3.939 6.782 multiply_cannon_sync_h2d 1332 9.7 5.526 6.211 5.526 6.211 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.075 5.950 mp_irecv_dv 3229 10.9 2.053 5.874 2.053 5.874 arnoldi_extremal 4 6.8 0.000 0.000 5.228 5.244 arnoldi_normal_ev 4 7.8 0.001 0.004 5.228 5.244 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.951 5.177 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.487 4.984 build_subspace 16 8.4 0.014 0.021 4.888 4.895 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.518 4.695 calculate_norms 2376 9.8 4.181 4.535 4.181 4.535 mp_allgather_i34 111 8.7 2.154 4.470 2.154 4.470 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.603 3.895 dbcsr_matrix_vector_mult_local 304 10.0 3.214 3.710 3.216 3.712 dbcsr_sort_data 658 11.4 3.055 3.402 3.055 3.402 ls_scf_post 1 4.0 0.000 0.000 3.243 3.248 dbcsr_special_finalize 555 9.7 0.006 0.007 2.789 3.199 dbcsr_merge_single_wm 555 10.7 0.529 0.662 2.781 3.191 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.056 3.125 ls_scf_store_result 1 5.0 0.000 0.000 2.994 3.051 dbcsr_data_release 10477 10.7 1.573 2.424 1.573 2.424 dbcsr_finalize 304 7.8 0.049 0.061 1.805 1.991 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.975 1.977 rebuild_ks_matrix 3 7.3 0.000 0.000 1.952 1.954 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.952 1.954 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.083000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2740.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_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 4.629926E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.057 99.395 99.396 qs_energies 1 2.0 0.000 0.000 98.625 98.630 ls_scf 1 3.0 0.000 0.000 96.683 96.693 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.859 78.035 ls_scf_main 1 4.0 0.000 0.000 62.370 62.371 multiply_cannon 111 7.7 0.084 0.168 55.284 60.609 density_matrix_trs4 2 5.0 0.002 0.003 54.994 55.107 multiply_cannon_loop 111 8.7 0.069 0.077 50.892 52.300 mp_waitall_1 5436 11.0 26.759 32.243 26.759 32.243 ls_scf_init_scf 1 4.0 0.000 0.000 30.728 30.733 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.346 29.403 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.167 27.178 multiply_cannon_multrec 444 9.7 13.993 16.628 21.021 23.110 make_m2s 222 7.7 0.004 0.005 17.892 20.275 make_images 222 8.7 3.710 4.407 17.830 20.214 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.132 16.486 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.336 14.133 make_images_data 222 9.7 0.003 0.004 10.091 12.471 hybrid_alltoall_any 227 10.6 0.787 3.753 9.933 12.355 dbcsr_mm_accdrv_process 3003 10.4 0.176 0.347 6.739 7.864 multiply_cannon_sync_h2d 444 9.7 6.560 7.653 6.560 7.653 dbcsr_mm_accdrv_process_sort 3003 11.4 6.424 7.521 6.424 7.521 mp_allgather_i34 111 8.7 2.621 7.015 2.621 7.015 arnoldi_extremal 4 6.8 0.000 0.000 5.804 5.816 arnoldi_normal_ev 4 7.8 0.001 0.004 5.804 5.816 build_subspace 16 8.4 0.015 0.020 5.419 5.430 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.654 4.723 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.585 4.719 mp_irecv_dv 1241 11.2 1.635 4.691 1.635 4.691 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.164 4.354 mp_sum_l 807 5.4 2.734 4.307 2.734 4.307 dbcsr_matrix_vector_mult_local 304 10.0 3.679 4.143 3.682 4.145 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.767 3.851 calculate_norms 792 9.8 3.551 3.652 3.551 3.652 make_images_sizes 222 9.7 0.000 0.000 1.170 3.589 mp_alltoall_i44 222 10.7 1.170 3.589 1.170 3.589 ls_scf_post 1 4.0 0.000 0.000 3.584 3.589 ls_scf_store_result 1 5.0 0.000 0.000 3.365 3.408 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.858 3.156 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.264 2.266 dbcsr_finalize 304 7.8 0.062 0.078 2.197 2.246 rebuild_ks_matrix 3 7.3 0.000 0.000 2.232 2.234 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.232 2.234 dbcsr_merge_all 275 8.9 0.475 0.521 2.047 2.103 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.396000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3601.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60432cd460210ee206a5df6c16abecd6840c7cae_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 8.741380E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.103 0.137 108.432 108.432 qs_energies 1 2.0 0.000 0.000 106.979 106.992 ls_scf 1 3.0 0.000 0.000 104.017 104.029 dbcsr_multiply_generic 111 6.7 0.023 0.026 77.352 77.459 ls_scf_main 1 4.0 0.000 0.000 66.072 66.072 density_matrix_trs4 2 5.0 0.002 0.003 56.672 56.724 multiply_cannon 111 7.7 0.167 0.245 49.838 51.950 multiply_cannon_loop 111 8.7 0.068 0.070 46.396 47.316 ls_scf_init_scf 1 4.0 0.000 0.000 34.260 34.261 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.574 32.591 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.757 29.766 mp_waitall_1 4527 11.1 22.243 26.017 22.243 26.017 make_m2s 222 7.7 0.005 0.005 23.953 24.921 make_images 222 8.7 4.592 4.984 23.847 24.812 multiply_cannon_multrec 444 9.7 17.841 18.502 22.432 22.953 hybrid_alltoall_any 227 10.6 1.662 3.628 12.956 15.621 make_images_data 222 9.7 0.003 0.003 13.147 15.555 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.512 10.901 multiply_cannon_sync_h2d 444 9.7 8.842 8.890 8.842 8.890 arnoldi_extremal 4 6.8 0.000 0.000 7.338 7.346 arnoldi_normal_ev 4 7.8 0.002 0.008 7.338 7.346 build_subspace 16 8.4 0.026 0.036 6.785 6.794 dbcsr_matrix_vector_mult 304 9.0 0.016 0.033 5.401 5.537 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.329 5.428 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.041 5.291 dbcsr_matrix_vector_mult_local 304 10.0 4.924 5.197 4.927 5.200 dbcsr_mm_accdrv_process 1814 10.4 0.232 0.321 4.430 4.565 dbcsr_mm_accdrv_process_sort 1814 11.4 4.128 4.266 4.128 4.266 ls_scf_post 1 4.0 0.000 0.000 3.685 3.698 make_images_sizes 222 9.7 0.000 0.000 1.469 3.620 mp_alltoall_i44 222 10.7 1.469 3.620 1.469 3.620 mp_allgather_i34 111 8.7 1.089 3.558 1.089 3.558 ls_scf_store_result 1 5.0 0.000 0.000 3.426 3.434 calculate_norms 792 9.8 3.245 3.284 3.245 3.284 dbcsr_finalize 304 7.8 0.082 0.089 3.099 3.183 dbcsr_merge_all 275 8.9 0.890 0.918 2.887 2.964 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 2.932 2.932 dbcsr_complete_redistribute 5 7.6 1.432 1.470 2.779 2.900 dbcsr_data_release 12724 10.6 2.331 2.828 2.331 2.828 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.637 2.638 rebuild_ks_matrix 3 7.3 0.000 0.000 2.571 2.572 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.571 2.572 matrix_ls_to_qs 2 6.0 0.000 0.000 2.416 2.545 dbcsr_sort_data 325 11.1 2.448 2.507 2.448 2.507 dbcsr_new_transposed 4 7.5 0.245 0.258 2.305 2.316 dbcsr_frobenius_norm 74 6.6 2.058 2.138 2.197 2.237 dbcsr_add_d 103 6.2 0.000 0.000 2.130 2.202 dbcsr_add_anytype 103 7.2 0.859 0.891 2.129 2.201 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.432000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6862.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 60432cd460210ee206a5df6c16abecd6840c7cae Summary: empty Status: OK