=== This is the CP2K Performance-Test === Already up to date. Current branch master is up to date. GIT Revision: 6b94574d37ca4648fbb0f33b2f35874362a23d3a ################# 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 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SIRIUS 7.4.3, 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. # # Last update: 04.10.2023 # # \ 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 ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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)) 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__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 += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.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.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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/01 job id: 49481242 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/02 job id: 49481244 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/03 job id: 49481245 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/04 job id: 49481246 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/05 job id: 49481248 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/06 job id: 49481249 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/07 job id: 49481250 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/08 job id: 49481251 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/09 job id: 49481252 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/10 job id: 49481253 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/11 job id: 49481254 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/12 job id: 49481256 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/13 job id: 49481258 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/14 job id: 49481259 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/15 job id: 49481260 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/16 job id: 49481262 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/17 job id: 49481264 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/18 job id: 49481265 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/19 job id: 49481266 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/20 job id: 49481267 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/21 job id: 49481269 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/22 job id: 49481270 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/23 job id: 49481271 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/24 job id: 49481272 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/25 job id: 49481273 --- 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/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/26 job id: 49481275 --- 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 ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/27 job id: 49481277 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ 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]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 424 8. 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 0.019 0.032 132.542 132.543 farming_run 1 2.0 132.009 132.010 132.512 132.516 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.488785E+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 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 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 0.006 0.026 114.309 114.309 qs_energies 1 2.0 0.000 0.000 114.109 114.111 mp2_main 1 3.0 0.000 0.000 112.034 112.036 mp2_gpw_main 1 4.0 0.028 0.033 111.020 111.022 mp2_ri_gpw_compute_in 1 5.0 0.172 0.176 92.228 92.514 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.645 54.930 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.168 41.042 46.029 get_2c_integrals 1 6.0 0.008 0.009 36.705 37.408 integrate_v_rspace 273 8.0 0.438 0.456 24.610 29.430 pw_transfer 6555 10.6 0.379 0.391 26.741 27.169 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.377 25.872 grid_integrate_task_list 273 9.0 20.516 25.801 20.516 25.801 fft_wrap_pw1pw2_100 2178 12.4 1.096 1.273 22.928 23.428 compute_2c_integrals 1 7.0 0.003 0.003 18.932 18.934 rpa_ri_compute_en 1 5.0 0.027 0.030 18.675 18.800 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.520 18.665 mp2_eri_2c_integrate_gpw 1 9.0 2.386 2.434 18.516 18.661 cp_fm_cholesky_decompose 12 8.2 17.712 18.417 17.712 18.417 cholesky_decomp 1 7.0 0.000 0.000 16.620 17.325 fft3d_s 5443 13.4 16.083 16.333 16.105 16.353 ao_to_mo_and_store_B_mult_1 272 7.0 10.771 15.317 10.771 15.317 calculate_wavefunction 272 8.0 5.413 5.565 12.333 13.037 rpa_num_int 1 6.0 0.000 0.002 10.652 10.662 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.627 10.630 calc_mat_Q 8 8.0 0.000 0.000 9.486 9.566 contract_S_to_Q 8 9.0 0.000 0.000 8.906 8.989 calc_potential_gpw 544 9.5 0.005 0.006 8.210 8.696 parallel_gemm_fm 14 9.1 0.000 0.000 8.492 8.579 parallel_gemm_fm_cosma 14 10.1 8.492 8.579 8.492 8.579 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.029 8.256 potential_pw2rs 545 10.0 0.107 0.109 7.467 8.171 create_integ_mat 1 6.0 0.014 0.026 7.636 7.645 collocate_single_gaussian 272 10.0 0.040 0.041 7.301 7.581 array2fm 1 7.0 0.000 0.000 6.695 7.098 pw_scatter_s 2720 13.7 4.357 4.595 4.357 4.595 pw_gather_s 2722 13.2 3.440 3.960 3.440 3.960 array2fm_buffer_send 1 8.0 2.929 3.146 2.929 3.146 pw_poisson_solve 545 10.5 1.126 1.183 2.194 2.318 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.019818, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2796.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 424 9. 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 0.023 0.036 419.984 419.986 farming_run 1 2.0 419.220 419.228 419.951 419.953 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.244754E+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 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 12276248. 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 0.011 0.032 217.334 217.335 qs_energies 1 2.0 0.000 0.000 217.087 217.094 scf_env_do_scf 1 3.0 0.000 0.000 115.116 115.117 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.217 114.225 rebuild_ks_matrix 4 6.0 0.000 0.000 114.216 114.224 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.062 114.216 114.224 hfx_ks_matrix 4 8.0 0.001 0.001 113.867 113.872 integrate_four_center 4 9.0 0.153 0.478 113.866 113.871 integrate_four_center_main 4 10.0 0.120 0.557 102.021 105.639 integrate_four_center_bin 266 11.0 101.901 105.331 101.901 105.331 mp2_main 1 3.0 0.000 0.000 101.676 101.683 mp2_gpw_main 1 4.0 0.051 0.083 100.834 100.844 init_scf_loop 1 4.0 0.000 0.000 97.278 97.278 mp2_ri_gpw_compute_in 1 5.0 0.069 0.084 73.914 75.036 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.777 54.894 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.160 41.550 46.245 integrate_v_rspace 95 8.0 0.396 0.562 28.000 32.603 pw_transfer 2240 10.6 0.146 0.169 29.475 29.947 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.494 28.877 mp2_ri_gpw_compute_en 1 5.0 0.056 0.064 26.754 28.406 grid_integrate_task_list 95 9.0 23.413 28.218 23.413 28.218 fft_wrap_pw1pw2_100 730 12.4 1.247 1.453 26.251 26.613 ao_to_mo_and_store_B_mult_1 91 7.0 10.549 26.510 10.549 26.510 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.831 1.879 24.990 24.999 get_2c_integrals 1 6.0 0.000 0.000 20.033 20.076 compute_2c_integrals 1 7.0 0.007 0.014 19.027 19.037 compute_2c_integrals_loop_lm 1 8.0 0.004 0.008 18.540 18.894 mp2_eri_2c_integrate_gpw 1 9.0 1.742 1.822 18.536 18.893 fft3d_s 1823 13.4 18.505 18.822 18.519 18.836 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 17.837 17.837 calculate_wavefunction 91 8.0 2.020 2.053 9.635 9.827 mp2_ri_gpw_compute_en_expansio 172 7.0 0.560 0.590 8.753 9.375 potential_pw2rs 186 10.0 0.033 0.035 8.381 8.863 local_gemm 172 8.0 8.193 8.807 8.193 8.807 mp2_ri_gpw_compute_en_comm 22 7.0 0.499 0.516 8.012 8.653 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.104 8.401 calc_potential_gpw 182 9.5 0.002 0.002 7.906 8.259 collocate_single_gaussian 91 10.0 0.017 0.020 7.708 7.915 mp_sync 37 10.5 3.826 6.973 3.826 6.973 integrate_four_center_load 4 10.0 0.000 0.000 6.758 6.762 hfx_load_balance 1 11.0 0.000 0.000 6.758 6.761 mp_sendrecv_dm3 2068 8.0 6.043 6.696 6.043 6.696 mp2_ri_gpw_compute_en_ener 172 7.0 6.350 6.415 6.350 6.415 pw_gather_s 912 13.2 4.473 4.821 4.473 4.821 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.837224, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1504.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 452.546560E+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 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 1521535. 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 0.008 0.030 52.972 52.974 qs_mol_dyn_low 1 2.0 0.003 0.003 52.767 52.775 qs_forces 11 3.9 0.002 0.003 52.621 52.623 qs_energies 11 4.9 0.002 0.002 51.084 51.093 scf_env_do_scf 11 5.9 0.001 0.001 44.939 44.939 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 42.820 42.821 qs_scf_new_mos 108 7.5 0.000 0.001 32.911 33.182 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.910 33.182 dbcsr_multiply_generic 2286 12.5 0.093 0.097 32.641 32.986 ot_scf_mini 108 9.5 0.002 0.002 31.299 31.492 multiply_cannon 2286 13.5 0.190 0.196 25.717 27.091 multiply_cannon_loop 2286 14.5 1.801 1.920 25.042 26.421 velocity_verlet 10 3.0 0.018 0.060 25.989 26.012 ot_mini 108 10.5 0.001 0.001 18.585 18.835 qs_ot_get_derivative 108 11.5 0.001 0.001 15.640 15.819 mp_waitall_1 245248 16.5 7.959 13.861 7.959 13.861 multiply_cannon_metrocomm3 54864 15.5 0.073 0.080 5.740 12.180 multiply_cannon_multrec 54864 15.5 3.700 5.816 7.609 11.038 qs_ot_get_p 119 10.4 0.001 0.001 8.101 8.366 rebuild_ks_matrix 119 8.3 0.000 0.000 7.889 8.043 qs_ks_build_kohn_sham_matrix 119 9.3 0.041 0.073 7.889 8.042 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.892 7.032 mp_sum_l 7287 12.8 4.954 6.523 4.954 6.523 multiply_cannon_sync_h2d 54864 15.5 5.205 6.050 5.205 6.050 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.372 5.834 dbcsr_mm_accdrv_process 76910 16.1 1.797 2.863 3.822 5.506 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.203 5.264 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.005 5.128 init_scf_run 11 5.9 0.000 0.002 4.700 4.701 scf_env_initial_rho_setup 11 6.9 0.001 0.003 4.700 4.700 sum_up_and_integrate 119 10.3 0.001 0.001 4.449 4.455 integrate_v_rspace 119 11.3 0.002 0.002 4.438 4.445 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.432 4.432 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.220 4.220 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.090 4.201 calculate_rho_elec 119 8.7 0.011 0.016 4.089 4.201 cp_fm_redistribute_end 50 14.0 2.154 4.193 2.161 4.196 cp_fm_diag_elpa_base 50 14.0 2.029 4.078 2.033 4.087 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.916 3.114 apply_single 119 13.6 0.000 0.000 2.915 3.114 calculate_dm_sparse 119 9.5 0.000 0.000 2.855 2.960 acc_transpose_blocks 54864 15.5 0.228 0.257 2.206 2.751 ot_diis_step 108 11.5 0.006 0.006 2.695 2.696 jit_kernel_multiply 13 15.8 1.963 2.567 1.963 2.567 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.420 2.480 calculate_first_density_matrix 1 7.0 0.000 0.001 2.420 2.427 multiply_cannon_metrocomm1 54864 15.5 0.057 0.063 1.558 2.337 density_rs2pw 119 9.7 0.004 0.004 2.160 2.282 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.271 2.273 wfi_extrapolate 11 7.9 0.001 0.001 2.207 2.207 grid_integrate_task_list 119 12.3 2.015 2.138 2.015 2.138 init_scf_loop 11 6.9 0.000 0.000 2.099 2.100 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.941 2.012 mp_sum_d 4135 12.0 1.301 1.922 1.301 1.922 potential_pw2rs 119 12.3 0.004 0.004 1.827 1.837 pw_transfer 1439 11.6 0.051 0.055 1.675 1.741 make_m2s 4572 13.5 0.053 0.055 1.656 1.699 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.600 1.670 make_images 4572 14.5 0.134 0.140 1.575 1.617 transfer_rs2pw 487 10.6 0.005 0.006 1.367 1.460 mp_waitany 12084 13.8 1.266 1.447 1.266 1.447 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.409 1.437 acc_transpose_blocks_sync 164592 16.5 1.191 1.435 1.191 1.435 mp_alltoall_d11v 2130 13.8 1.280 1.430 1.280 1.430 grid_collocate_task_list 119 9.7 1.344 1.427 1.344 1.427 transfer_pw2rs 487 13.2 0.006 0.007 1.355 1.391 fft3d_ps 1201 14.6 0.370 0.475 1.315 1.380 fft_wrap_pw1pw2_140 487 13.2 0.141 0.154 1.237 1.308 dbcsr_dot_sd 1205 11.9 0.048 0.059 0.691 1.094 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.042 1.067 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.974000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.090909, yerr=1.311110 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.620608E+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 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1297115. 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 0.009 0.029 39.503 39.505 qs_mol_dyn_low 1 2.0 0.003 0.004 39.275 39.285 qs_forces 11 3.9 0.002 0.002 39.157 39.158 qs_energies 11 4.9 0.001 0.001 37.463 37.466 scf_env_do_scf 11 5.9 0.000 0.001 32.385 32.386 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.956 29.956 dbcsr_multiply_generic 2286 12.5 0.101 0.103 22.290 22.610 qs_scf_new_mos 108 7.5 0.001 0.001 21.078 21.306 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.078 21.306 ot_scf_mini 108 9.5 0.002 0.003 20.130 20.297 multiply_cannon 2286 13.5 0.208 0.215 17.007 18.770 velocity_verlet 10 3.0 0.001 0.002 18.763 18.765 multiply_cannon_loop 2286 14.5 1.186 1.250 15.805 17.226 ot_mini 108 10.5 0.001 0.001 12.352 12.587 mp_waitall_1 200699 16.5 5.723 10.779 5.723 10.779 qs_ot_get_derivative 108 11.5 0.001 0.001 9.960 10.128 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.109 9.409 multiply_cannon_multrec 27432 15.5 1.831 4.111 6.145 8.998 rebuild_ks_matrix 119 8.3 0.000 0.001 6.878 7.030 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.878 7.030 dbcsr_mm_accdrv_process 47894 16.0 3.501 5.790 4.232 6.331 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.043 6.181 qs_ot_get_p 119 10.4 0.001 0.001 4.809 5.033 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.814 4.658 mp_sum_l 7287 12.8 2.116 4.128 2.116 4.128 init_scf_run 11 5.9 0.000 0.001 3.865 3.865 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.864 3.865 sum_up_and_integrate 119 10.3 0.001 0.001 3.792 3.798 integrate_v_rspace 119 11.3 0.002 0.003 3.778 3.784 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.751 3.774 calculate_rho_elec 119 8.7 0.021 0.024 3.750 3.774 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.009 3.748 apply_single 119 13.6 0.000 0.000 3.009 3.748 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.112 3.131 make_m2s 4572 13.5 0.052 0.053 2.721 3.050 make_images 4572 14.5 0.206 0.244 2.633 2.964 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.710 2.710 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.366 2.456 init_scf_loop 11 6.9 0.000 0.000 2.401 2.402 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.365 2.365 calculate_first_density_matrix 1 7.0 0.000 0.000 2.338 2.341 ot_diis_step 108 11.5 0.011 0.011 2.337 2.338 cp_fm_redistribute_end 50 14.0 1.197 2.334 1.201 2.338 cp_fm_diag_elpa_base 50 14.0 1.102 2.241 1.131 2.280 multiply_cannon_sync_h2d 27432 15.5 1.698 2.188 1.698 2.188 calculate_dm_sparse 119 9.5 0.000 0.001 2.105 2.186 density_rs2pw 119 9.7 0.004 0.004 2.040 2.103 acc_transpose_blocks 27432 15.5 0.113 0.119 1.629 1.983 grid_integrate_task_list 119 12.3 1.835 1.932 1.835 1.932 pw_transfer 1439 11.6 0.065 0.069 1.892 1.928 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.910 1.912 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.800 1.838 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.702 1.742 jit_kernel_multiply 9 16.5 0.673 1.732 0.673 1.732 make_images_data 4572 15.5 0.047 0.054 1.261 1.583 prepare_preconditioner 11 7.9 0.000 0.000 1.529 1.555 make_preconditioner 11 8.9 0.000 0.000 1.529 1.555 potential_pw2rs 119 12.3 0.006 0.007 1.530 1.543 mp_alltoall_d11v 2130 13.8 1.260 1.540 1.260 1.540 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.431 1.487 wfi_extrapolate 11 7.9 0.001 0.001 1.468 1.468 fft3d_ps 1201 14.6 0.519 0.576 1.436 1.468 hybrid_alltoall_any 4725 16.4 0.054 0.115 1.119 1.457 fft_wrap_pw1pw2_140 487 13.2 0.160 0.169 1.407 1.444 grid_collocate_task_list 119 9.7 1.286 1.340 1.286 1.340 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.236 1.284 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.274 1.281 transfer_rs2pw 487 10.6 0.005 0.005 1.171 1.230 mp_sum_d 4135 12.0 0.562 1.049 0.562 1.049 mp_allgather_i34 2286 14.5 0.627 1.034 0.627 1.034 transfer_pw2rs 487 13.2 0.005 0.006 0.984 0.994 acc_transpose_blocks_sync 82296 16.5 0.813 0.943 0.813 0.943 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.927 0.927 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.894 0.908 acc_transpose_blocks_kernels 27432 16.5 0.190 0.279 0.676 0.898 make_images_sizes 4572 15.5 0.005 0.005 0.589 0.840 mp_alltoall_i44 4572 16.5 0.584 0.836 0.584 0.836 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.505000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.000000, yerr=1.705606 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 519.479296E+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 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.026 32.032 32.033 qs_mol_dyn_low 1 2.0 0.003 0.003 31.792 31.811 qs_forces 11 3.9 0.006 0.035 31.682 31.683 qs_energies 11 4.9 0.004 0.024 30.073 30.097 scf_env_do_scf 11 5.9 0.000 0.001 25.383 25.383 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.977 22.977 dbcsr_multiply_generic 2286 12.5 0.095 0.097 16.853 16.903 qs_scf_new_mos 108 7.5 0.001 0.001 15.314 15.334 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.313 15.333 velocity_verlet 10 3.0 0.002 0.002 15.092 15.095 ot_scf_mini 108 9.5 0.002 0.002 14.557 14.569 multiply_cannon 2286 13.5 0.194 0.198 13.604 14.372 multiply_cannon_loop 2286 14.5 0.854 0.900 12.845 13.622 ot_mini 108 10.5 0.001 0.001 8.991 9.003 qs_ot_get_derivative 108 11.5 0.001 0.001 7.514 7.529 multiply_cannon_multrec 18288 15.5 1.898 2.876 6.982 7.277 rebuild_ks_matrix 119 8.3 0.000 0.000 5.961 5.971 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 5.960 5.971 dbcsr_mm_accdrv_process 38222 16.0 4.889 5.689 4.989 5.752 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.252 5.261 sum_up_and_integrate 119 10.3 0.001 0.002 3.546 3.551 integrate_v_rspace 119 11.3 0.003 0.004 3.533 3.540 mp_waitall_1 158411 16.6 2.551 3.479 2.551 3.479 init_scf_run 11 5.9 0.000 0.001 3.470 3.470 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.470 3.470 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.794 3.436 qs_ot_get_p 119 10.4 0.001 0.001 3.331 3.349 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.247 3.249 calculate_rho_elec 119 8.7 0.031 0.031 3.246 3.248 init_scf_loop 11 6.9 0.001 0.004 2.390 2.390 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.050 2.326 apply_single 119 13.6 0.000 0.000 2.049 2.326 calculate_first_density_matrix 1 7.0 0.001 0.005 2.263 2.265 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.237 2.241 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.372 2.131 make_m2s 4572 13.5 0.044 0.045 1.810 1.950 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.943 1.943 density_rs2pw 119 9.7 0.004 0.004 1.817 1.918 calculate_dm_sparse 119 9.5 0.001 0.001 1.879 1.891 grid_integrate_task_list 119 12.3 1.811 1.888 1.811 1.888 make_images 4572 14.5 0.193 0.205 1.725 1.863 pw_transfer 1439 11.6 0.065 0.068 1.813 1.821 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.720 1.731 acc_transpose_blocks 18288 15.5 0.079 0.081 1.643 1.722 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.699 1.709 cp_fm_diag_elpa_base 50 14.0 1.676 1.687 1.697 1.708 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.684 1.690 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.672 1.673 prepare_preconditioner 11 7.9 0.000 0.000 1.628 1.631 make_preconditioner 11 8.9 0.000 0.001 1.628 1.631 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.491 1.573 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.526 1.530 ot_diis_step 108 11.5 0.011 0.011 1.464 1.464 fft_wrap_pw1pw2_140 487 13.2 0.212 0.218 1.385 1.395 potential_pw2rs 119 12.3 0.007 0.009 1.338 1.341 grid_collocate_task_list 119 9.7 1.234 1.311 1.234 1.311 fft3d_ps 1201 14.6 0.525 0.541 1.289 1.299 mp_sum_l 7287 12.8 1.054 1.293 1.054 1.293 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.218 1.222 multiply_cannon_sync_h2d 18288 15.5 1.054 1.164 1.054 1.164 wfi_extrapolate 11 7.9 0.001 0.001 1.155 1.155 transfer_rs2pw 487 10.6 0.005 0.005 1.019 1.117 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 0.989 0.991 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.938 0.951 make_images_data 4572 15.5 0.047 0.051 0.780 0.936 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.671 0.868 acc_transpose_blocks_sync 54864 16.5 0.740 0.827 0.740 0.827 transfer_pw2rs 487 13.2 0.004 0.004 0.819 0.823 acc_transpose_blocks_kernels 18288 16.5 0.218 0.225 0.803 0.815 mp_alltoall_d11v 2130 13.8 0.652 0.795 0.652 0.795 jit_kernel_multiply 4 15.9 0.042 0.792 0.042 0.792 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.780 0.782 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.673 0.744 cp_fm_cholesky_invert 11 10.9 0.691 0.694 0.691 0.694 mp_alltoall_z22v 1201 16.6 0.615 0.678 0.615 0.678 mp_waitany 9880 13.7 0.563 0.653 0.563 0.653 transfer_rs2pw_140 130 11.5 0.117 0.119 0.553 0.643 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.033000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.545455, yerr=0.655555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 560.214016E+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 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 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 0.008 0.026 36.418 36.418 qs_mol_dyn_low 1 2.0 0.003 0.003 36.186 36.196 qs_forces 11 3.9 0.002 0.002 36.057 36.058 qs_energies 11 4.9 0.001 0.001 34.364 34.369 scf_env_do_scf 11 5.9 0.000 0.001 29.159 29.161 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.800 25.801 dbcsr_multiply_generic 2286 12.5 0.099 0.102 20.129 20.276 velocity_verlet 10 3.0 0.001 0.002 18.363 18.368 qs_scf_new_mos 108 7.5 0.001 0.001 17.813 17.869 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.812 17.868 ot_scf_mini 108 9.5 0.002 0.003 16.783 16.838 multiply_cannon 2286 13.5 0.218 0.224 16.518 16.817 multiply_cannon_loop 2286 14.5 1.541 1.620 15.600 16.006 ot_mini 108 10.5 0.001 0.001 10.387 10.451 multiply_cannon_multrec 27432 15.5 2.460 3.120 9.240 9.581 qs_ot_get_derivative 108 11.5 0.001 0.001 8.527 8.582 dbcsr_mm_accdrv_process 47916 15.9 6.153 7.880 6.677 8.001 rebuild_ks_matrix 119 8.3 0.000 0.000 6.249 6.300 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.249 6.300 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.538 5.582 init_scf_run 11 5.9 0.000 0.001 3.880 3.880 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.880 3.880 qs_ot_get_p 119 10.4 0.001 0.001 3.568 3.642 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.156 3.557 sum_up_and_integrate 119 10.3 0.001 0.001 3.450 3.460 integrate_v_rspace 119 11.3 0.002 0.003 3.439 3.448 init_scf_loop 11 6.9 0.000 0.000 3.340 3.341 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.281 3.320 calculate_rho_elec 119 8.7 0.040 0.046 3.281 3.319 acc_transpose_blocks 27432 15.5 0.120 0.124 2.532 2.649 prepare_preconditioner 11 7.9 0.000 0.000 2.519 2.526 make_preconditioner 11 8.9 0.000 0.000 2.519 2.526 calculate_first_density_matrix 1 7.0 0.000 0.000 2.512 2.514 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.130 2.453 mp_waitall_1 137007 16.6 1.787 2.369 1.787 2.369 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.312 2.341 make_m2s 4572 13.5 0.054 0.056 2.219 2.326 calculate_dm_sparse 119 9.5 0.000 0.000 2.272 2.325 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.109 2.240 apply_single 119 13.6 0.000 0.000 2.109 2.240 make_images 4572 14.5 0.274 0.336 2.112 2.216 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.173 2.185 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.960 1.961 pw_transfer 1439 11.6 0.064 0.068 1.888 1.921 grid_integrate_task_list 119 12.3 1.842 1.905 1.842 1.905 density_rs2pw 119 9.7 0.004 0.004 1.751 1.841 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.796 1.832 ot_diis_step 108 11.5 0.012 0.012 1.819 1.820 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.799 1.812 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.773 1.774 acc_transpose_blocks_sync 82296 16.5 1.457 1.563 1.457 1.563 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.532 1.541 cp_fm_diag_elpa_base 50 14.0 1.498 1.512 1.530 1.538 fft_wrap_pw1pw2_140 487 13.2 0.246 0.258 1.489 1.527 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.369 1.378 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.857 1.366 grid_collocate_task_list 119 9.7 1.248 1.346 1.248 1.346 fft3d_ps 1201 14.6 0.553 0.603 1.316 1.341 mp_sum_l 7287 12.8 0.966 1.339 0.966 1.339 wfi_extrapolate 11 7.9 0.001 0.001 1.315 1.315 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.285 1.301 potential_pw2rs 119 12.3 0.009 0.010 1.264 1.266 jit_kernel_multiply 6 16.1 0.459 1.228 0.459 1.228 cp_fm_upper_to_full 72 14.2 0.824 1.166 0.824 1.166 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.069 1.069 dbcsr_complete_redistribute 329 12.2 0.117 0.141 0.737 1.013 transfer_rs2pw 487 10.6 0.004 0.005 0.874 0.966 acc_transpose_blocks_kernels 27432 16.5 0.272 0.280 0.926 0.949 make_images_data 4572 15.5 0.048 0.052 0.811 0.921 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.875 hybrid_alltoall_any 4725 16.4 0.066 0.155 0.709 0.862 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.551 0.817 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.800 0.806 mp_alltoall_d11v 2130 13.8 0.689 0.768 0.689 0.768 cp_fm_cholesky_invert 11 10.9 0.741 0.744 0.741 0.744 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.418000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=532.545455, yerr=4.356054 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 620.740608E+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 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 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 0.009 0.029 28.448 28.449 qs_mol_dyn_low 1 2.0 0.003 0.003 28.262 28.287 qs_forces 11 3.9 0.002 0.002 28.140 28.140 qs_energies 11 4.9 0.001 0.001 26.441 26.443 scf_env_do_scf 11 5.9 0.000 0.001 21.783 21.783 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.329 19.330 velocity_verlet 10 3.0 0.002 0.002 14.650 14.657 dbcsr_multiply_generic 2286 12.5 0.092 0.097 13.005 13.101 qs_scf_new_mos 108 7.5 0.001 0.001 11.750 11.781 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.749 11.780 ot_scf_mini 108 9.5 0.002 0.002 11.045 11.072 multiply_cannon 2286 13.5 0.223 0.230 10.443 10.874 multiply_cannon_loop 2286 14.5 0.640 0.663 9.535 9.714 ot_mini 108 10.5 0.001 0.001 6.344 6.377 multiply_cannon_multrec 9144 15.5 1.731 1.983 6.017 6.256 rebuild_ks_matrix 119 8.3 0.000 0.000 5.713 5.737 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.712 5.737 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.083 5.104 qs_ot_get_derivative 108 11.5 0.001 0.001 5.061 5.089 dbcsr_mm_accdrv_process 12550 15.8 3.722 4.199 4.177 4.239 sum_up_and_integrate 119 10.3 0.001 0.001 3.404 3.408 integrate_v_rspace 119 11.3 0.003 0.003 3.393 3.397 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.296 3.303 calculate_rho_elec 119 8.7 0.060 0.061 3.296 3.302 init_scf_run 11 5.9 0.000 0.001 3.233 3.233 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.233 3.233 qs_ot_get_p 119 10.4 0.001 0.001 2.832 2.866 init_scf_loop 11 6.9 0.000 0.000 2.435 2.436 calculate_first_density_matrix 1 7.0 0.000 0.000 2.100 2.102 pw_transfer 1439 11.6 0.065 0.068 1.948 1.957 make_m2s 4572 13.5 0.034 0.035 1.789 1.939 grid_integrate_task_list 119 12.3 1.866 1.939 1.866 1.939 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.855 1.866 make_images 4572 14.5 0.270 0.303 1.700 1.850 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.825 1.827 mp_waitall_1 115863 16.7 1.324 1.789 1.324 1.789 density_rs2pw 119 9.7 0.003 0.004 1.675 1.786 calculate_dm_sparse 119 9.5 0.000 0.000 1.768 1.786 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.709 1.729 prepare_preconditioner 11 7.9 0.000 0.000 1.712 1.717 make_preconditioner 11 8.9 0.000 0.000 1.712 1.717 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.602 1.627 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.581 1.582 fft_wrap_pw1pw2_140 487 13.2 0.323 0.333 1.531 1.542 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.533 1.534 acc_transpose_blocks 9144 15.5 0.042 0.043 1.435 1.471 grid_collocate_task_list 119 9.7 1.297 1.409 1.297 1.409 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.376 1.384 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.340 1.356 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.297 1.305 cp_fm_diag_elpa_base 50 14.0 1.271 1.287 1.296 1.304 fft3d_ps 1201 14.6 0.556 0.568 1.268 1.278 ot_diis_step 108 11.5 0.013 0.013 1.269 1.269 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.215 1.250 apply_single 119 13.6 0.000 0.000 1.215 1.250 potential_pw2rs 119 12.3 0.010 0.011 1.216 1.219 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.214 1.215 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.159 1.164 wfi_extrapolate 11 7.9 0.001 0.001 1.081 1.081 jit_kernel_multiply 6 15.8 0.416 1.035 0.416 1.035 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.767 0.977 make_images_data 4572 15.5 0.042 0.046 0.768 0.932 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.876 0.927 mp_alltoall_d11v 2130 13.8 0.788 0.878 0.788 0.878 cp_fm_cholesky_invert 11 10.9 0.856 0.859 0.856 0.859 transfer_rs2pw 487 10.6 0.004 0.004 0.761 0.855 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.815 0.822 acc_transpose_blocks_sync 27432 16.5 0.720 0.756 0.720 0.756 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.751 0.753 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.663 0.713 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.344 0.689 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.657 0.660 transfer_pw2rs 487 13.2 0.003 0.004 0.651 0.653 mp_allgather_i34 2286 14.5 0.227 0.642 0.227 0.642 mp_alltoall_z22v 1201 16.6 0.586 0.617 0.586 0.617 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.449000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=588.000000, yerr=5.169843 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 775.843840E+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 63723. MP_Allreduce 10154 429. 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 0.011 0.032 42.716 42.717 qs_mol_dyn_low 1 2.0 0.003 0.003 42.507 42.514 qs_forces 11 3.9 0.002 0.003 42.420 42.421 qs_energies 11 4.9 0.001 0.001 40.400 40.402 scf_env_do_scf 11 5.9 0.001 0.001 34.635 34.635 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.740 26.742 velocity_verlet 10 3.0 0.001 0.002 24.184 24.193 dbcsr_multiply_generic 2286 12.5 0.099 0.103 19.063 19.228 qs_scf_new_mos 108 7.5 0.001 0.001 17.317 17.417 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.316 17.417 ot_scf_mini 108 9.5 0.002 0.002 16.159 16.259 multiply_cannon 2286 13.5 0.299 0.308 14.993 15.900 multiply_cannon_loop 2286 14.5 0.849 0.868 13.729 14.633 ot_mini 108 10.5 0.001 0.001 9.822 9.939 multiply_cannon_multrec 9144 15.5 3.443 4.951 8.832 8.911 init_scf_loop 11 6.9 0.000 0.000 7.868 7.872 qs_ot_get_derivative 108 11.5 0.001 0.001 7.751 7.851 rebuild_ks_matrix 119 8.3 0.000 0.000 7.126 7.267 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.126 7.267 prepare_preconditioner 11 7.9 0.000 0.000 6.888 6.902 make_preconditioner 11 8.9 0.000 0.000 6.888 6.902 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.474 6.780 dbcsr_mm_accdrv_process 12550 15.8 4.422 5.833 5.254 6.558 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.422 6.551 cp_fm_upper_to_full 72 14.2 3.188 4.544 3.188 4.544 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.038 4.042 calculate_rho_elec 119 8.7 0.118 0.121 4.038 4.041 sum_up_and_integrate 119 10.3 0.001 0.001 3.778 3.784 integrate_v_rspace 119 11.3 0.003 0.003 3.768 3.774 init_scf_run 11 5.9 0.000 0.001 3.718 3.718 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.718 3.718 qs_ot_get_p 119 10.4 0.001 0.001 3.394 3.532 mp_waitall_1 94719 16.7 2.314 3.309 2.314 3.309 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.810 3.264 dbcsr_complete_redistribute 329 12.2 0.286 0.291 1.962 2.785 pw_transfer 1439 11.6 0.068 0.069 2.607 2.613 make_m2s 4572 13.5 0.037 0.038 2.445 2.603 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.268 2.536 apply_single 119 13.6 0.000 0.000 2.268 2.536 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.509 2.514 make_images 4572 14.5 0.353 0.385 2.325 2.485 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.645 2.452 mp_alltoall_i22 627 13.8 1.452 2.290 1.452 2.290 calculate_dm_sparse 119 9.5 0.000 0.000 2.275 2.288 calculate_first_density_matrix 1 7.0 0.000 0.000 2.247 2.249 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.271 2.231 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.409 2.211 grid_integrate_task_list 119 12.3 2.082 2.135 2.082 2.135 fft_wrap_pw1pw2_140 487 13.2 0.575 0.580 2.111 2.120 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.051 2.104 density_rs2pw 119 9.7 0.003 0.004 2.053 2.076 ot_diis_step 108 11.5 0.014 0.015 2.045 2.045 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.001 2.002 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.949 1.950 acc_transpose_blocks 9144 15.5 0.044 0.045 1.801 1.830 mp_sum_l 7287 12.8 1.081 1.815 1.081 1.815 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.767 1.768 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.624 1.651 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.644 1.644 fft3d_ps 1201 14.6 0.594 0.604 1.606 1.610 grid_collocate_task_list 119 9.7 1.523 1.562 1.523 1.562 cp_fm_cholesky_invert 11 10.9 1.477 1.480 1.477 1.480 wfi_extrapolate 11 7.9 0.001 0.001 1.408 1.408 hybrid_alltoall_any 4725 16.4 0.090 0.150 1.151 1.392 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.366 1.366 cp_fm_diag_elpa_base 50 14.0 1.220 1.272 1.364 1.364 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.314 1.332 make_images_data 4572 15.5 0.046 0.049 1.096 1.298 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.280 1.286 potential_pw2rs 119 12.3 0.014 0.015 1.269 1.270 mp_alltoall_d11v 2130 13.8 1.198 1.228 1.198 1.228 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.096 1.108 acc_transpose_blocks_sync 27432 16.5 1.080 1.106 1.080 1.106 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.997 1.064 jit_kernel_multiply 6 15.5 0.803 1.030 0.803 1.030 qs_create_task_list 11 7.9 0.000 0.001 0.927 0.939 generate_qs_task_list 11 8.9 0.366 0.385 0.926 0.939 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.915 0.928 mp_alltoall_z22v 1201 16.6 0.876 0.895 0.876 0.895 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.717000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=726.000000, yerr=15.207654 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 502.374400E+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 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1069673. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 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 0.010 0.027 82.277 82.279 qs_mol_dyn_low 1 2.0 0.003 0.006 81.974 82.015 qs_forces 11 3.9 0.003 0.003 81.880 81.881 qs_energies 11 4.9 0.001 0.002 79.046 79.060 scf_env_do_scf 11 5.9 0.001 0.004 70.169 70.171 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.706 64.707 dbcsr_multiply_generic 2055 12.4 0.105 0.106 51.444 51.659 qs_scf_new_mos 99 7.5 0.000 0.001 47.663 47.776 qs_scf_loop_do_ot 99 8.5 0.000 0.001 47.663 47.775 ot_scf_mini 99 9.5 0.002 0.002 45.263 45.349 multiply_cannon 2055 13.4 0.185 0.189 42.845 43.625 velocity_verlet 10 3.0 0.001 0.001 43.159 43.160 multiply_cannon_loop 2055 14.4 1.780 1.823 41.896 42.679 ot_mini 99 10.5 0.001 0.001 27.005 27.080 qs_ot_get_derivative 99 11.5 0.001 0.001 20.192 20.294 multiply_cannon_multrec 49320 15.4 11.345 12.079 17.525 18.105 rebuild_ks_matrix 110 8.3 0.000 0.000 14.446 14.520 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.445 14.520 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.672 12.734 mp_waitall_1 220248 16.4 10.981 11.691 10.981 11.691 multiply_cannon_sync_h2d 49320 15.4 9.539 10.092 9.539 10.092 qs_ot_get_p 110 10.4 0.001 0.001 9.739 9.796 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.623 8.194 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.194 7.678 apply_single 110 13.6 0.000 0.000 7.194 7.678 multiply_cannon_metrocomm3 49320 15.4 0.082 0.085 6.699 7.666 sum_up_and_integrate 110 10.3 0.002 0.003 6.963 6.978 integrate_v_rspace 110 11.3 0.003 0.003 6.938 6.958 init_scf_run 11 5.9 0.000 0.001 6.843 6.843 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.842 6.843 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.578 6.610 ot_diis_step 99 11.5 0.006 0.006 6.552 6.552 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.285 6.422 calculate_rho_elec 110 8.6 0.020 0.024 6.284 6.421 dbcsr_mm_accdrv_process 87628 16.1 3.046 3.139 6.050 6.339 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.688 5.689 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.537 5.580 init_scf_loop 11 6.9 0.000 0.000 5.436 5.437 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.109 5.132 cp_fm_diag_elpa_base 48 14.0 5.096 5.120 5.107 5.130 mp_sum_l 6594 12.7 4.217 5.117 4.217 5.117 wfi_extrapolate 11 7.9 0.001 0.001 4.043 4.043 make_m2s 4110 13.4 0.060 0.066 3.930 4.028 make_images 4110 14.4 0.177 0.189 3.835 3.936 calculate_dm_sparse 110 9.5 0.001 0.001 3.797 3.887 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.638 3.642 density_rs2pw 110 9.6 0.004 0.005 3.300 3.417 grid_integrate_task_list 110 12.3 3.253 3.388 3.253 3.388 multiply_cannon_metrocomm1 49320 15.4 0.065 0.067 2.458 3.353 prepare_preconditioner 11 7.9 0.000 0.000 3.282 3.300 make_preconditioner 11 8.9 0.000 0.000 3.282 3.300 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.239 3.295 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.248 3.283 pw_transfer 1331 11.6 0.054 0.070 3.064 3.148 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.068 3.117 fft_wrap_pw1pw2 1111 12.6 0.007 0.009 2.976 3.058 jit_kernel_multiply 13 15.9 2.726 2.778 2.726 2.778 calculate_first_density_matrix 1 7.0 0.000 0.001 2.700 2.707 acc_transpose_blocks 49320 15.4 0.213 0.221 2.586 2.655 fft_wrap_pw1pw2_140 451 13.1 0.391 0.430 2.536 2.628 potential_pw2rs 110 12.3 0.005 0.006 2.562 2.577 mp_alltoall_d11v 2046 13.8 2.086 2.551 2.086 2.551 fft3d_ps 1111 14.6 0.796 0.892 2.270 2.338 grid_collocate_task_list 110 9.6 2.157 2.267 2.157 2.267 mp_waitany 14300 13.8 1.817 2.001 1.817 2.001 transfer_rs2pw 451 10.6 0.005 0.006 1.856 1.975 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.902 1.918 mp_sum_d 3889 11.9 1.371 1.902 1.371 1.902 make_images_data 4110 15.4 0.043 0.047 1.776 1.901 hybrid_alltoall_any 4261 16.3 0.085 0.485 1.549 1.803 cp_fm_cholesky_invert 11 10.9 1.758 1.761 1.758 1.761 transfer_pw2rs 451 13.1 0.006 0.007 1.730 1.742 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.674 1.698 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.279000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.181818, yerr=3.511492 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 595.058688E+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 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2984657. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 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 0.010 0.028 68.811 68.812 qs_mol_dyn_low 1 2.0 0.003 0.004 68.543 68.594 qs_forces 11 3.9 0.003 0.003 68.426 68.426 qs_energies 11 4.9 0.001 0.002 65.108 65.112 scf_env_do_scf 11 5.9 0.000 0.001 56.579 56.582 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.048 49.048 dbcsr_multiply_generic 2055 12.4 0.114 0.117 38.114 38.296 velocity_verlet 10 3.0 0.002 0.002 36.013 36.015 qs_scf_new_mos 99 7.5 0.001 0.001 33.238 33.392 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.237 33.391 multiply_cannon 2055 13.4 0.223 0.243 31.510 32.709 ot_scf_mini 99 9.5 0.003 0.003 31.576 31.728 multiply_cannon_loop 2055 14.4 1.159 1.185 30.217 31.095 ot_mini 99 10.5 0.001 0.001 18.462 18.616 multiply_cannon_multrec 24660 15.4 7.018 8.447 14.081 15.464 rebuild_ks_matrix 110 8.3 0.000 0.000 13.515 13.690 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.515 13.690 qs_ot_get_derivative 99 11.5 0.001 0.001 12.642 12.793 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.887 12.045 mp_waitall_1 176588 16.5 7.569 10.275 7.569 10.275 multiply_cannon_metrocomm3 24660 15.4 0.071 0.074 5.189 8.045 init_scf_loop 11 6.9 0.000 0.000 7.495 7.495 multiply_cannon_sync_h2d 24660 15.4 6.438 7.376 6.438 7.376 dbcsr_mm_accdrv_process 52282 16.1 5.444 6.198 6.892 7.181 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.535 7.156 apply_single 110 13.6 0.000 0.001 6.534 7.156 qs_ot_get_p 110 10.4 0.001 0.001 6.442 6.634 sum_up_and_integrate 110 10.3 0.001 0.002 6.334 6.344 integrate_v_rspace 110 11.3 0.002 0.003 6.307 6.318 init_scf_run 11 5.9 0.000 0.002 6.163 6.164 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.163 6.164 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.850 5.861 calculate_rho_elec 110 8.6 0.039 0.047 5.849 5.861 ot_diis_step 99 11.5 0.010 0.010 5.773 5.773 prepare_preconditioner 11 7.9 0.000 0.000 5.485 5.509 make_preconditioner 11 8.9 0.000 0.000 5.485 5.509 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.757 5.504 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.083 5.238 make_m2s 4110 13.4 0.057 0.061 4.219 4.696 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.566 4.587 make_images 4110 14.4 0.409 0.466 4.109 4.584 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.108 4.108 pw_transfer 1331 11.6 0.066 0.073 3.505 3.648 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.544 3.554 cp_fm_diag_elpa_base 48 14.0 3.497 3.511 3.541 3.551 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.399 3.545 wfi_extrapolate 11 7.9 0.001 0.001 3.503 3.503 grid_integrate_task_list 110 12.3 3.164 3.355 3.164 3.355 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.261 3.325 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.310 3.312 density_rs2pw 110 9.6 0.004 0.005 3.096 3.251 fft_wrap_pw1pw2_140 451 13.1 0.460 0.476 2.917 3.066 calculate_dm_sparse 110 9.5 0.001 0.001 2.997 3.028 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.983 3.023 make_images_data 4110 15.4 0.048 0.052 2.308 2.822 hybrid_alltoall_any 4261 16.3 0.107 0.455 2.018 2.758 fft3d_ps 1111 14.6 1.107 1.337 2.494 2.647 cp_fm_cholesky_invert 11 10.9 2.577 2.584 2.577 2.584 calculate_first_density_matrix 1 7.0 0.000 0.000 2.564 2.567 mp_sum_l 6594 12.7 1.756 2.530 1.756 2.530 grid_collocate_task_list 110 9.6 2.166 2.326 2.166 2.326 potential_pw2rs 110 12.3 0.009 0.009 2.217 2.229 acc_transpose_blocks 24660 15.4 0.114 0.118 1.999 2.032 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.965 1.986 jit_kernel_multiply 10 16.3 1.085 1.941 1.085 1.941 mp_alltoall_d11v 2046 13.8 1.718 1.900 1.718 1.900 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.860 1.861 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.784 1.798 mp_allgather_i34 2055 14.4 0.604 1.671 0.604 1.671 multiply_cannon_metrocomm4 22605 15.4 0.077 0.081 0.770 1.611 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.569 1.579 transfer_rs2pw 451 10.6 0.007 0.007 1.411 1.573 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.002 1.388 1.490 mp_irecv_dv 57340 16.2 0.643 1.395 0.643 1.395 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.812000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.272727, yerr=8.464803 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 664.203264E+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 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.034 60.076 60.077 qs_mol_dyn_low 1 2.0 0.003 0.004 59.725 59.748 qs_forces 11 3.9 0.003 0.003 59.651 59.652 qs_energies 11 4.9 0.002 0.008 56.480 56.484 scf_env_do_scf 11 5.9 0.000 0.001 48.407 48.408 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.779 39.779 velocity_verlet 10 3.0 0.001 0.002 32.592 32.594 dbcsr_multiply_generic 2055 12.4 0.106 0.110 28.976 29.278 qs_scf_new_mos 99 7.5 0.001 0.001 25.220 25.332 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.220 25.331 ot_scf_mini 99 9.5 0.002 0.002 23.972 24.095 multiply_cannon 2055 13.4 0.213 0.225 22.651 23.828 multiply_cannon_loop 2055 14.4 0.817 0.863 21.475 22.574 ot_mini 99 10.5 0.001 0.001 13.780 13.906 rebuild_ks_matrix 110 8.3 0.000 0.000 12.104 12.269 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.103 12.268 multiply_cannon_multrec 16440 15.4 3.622 4.704 9.886 10.872 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.654 10.805 mp_waitall_1 139946 16.5 7.000 10.011 7.000 10.011 qs_ot_get_derivative 99 11.5 0.001 0.001 9.278 9.403 init_scf_loop 11 6.9 0.001 0.002 8.595 8.596 multiply_cannon_metrocomm3 16440 15.4 0.046 0.048 4.422 7.241 prepare_preconditioner 11 7.9 0.000 0.000 6.830 6.852 make_preconditioner 11 8.9 0.000 0.001 6.829 6.852 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.137 6.491 dbcsr_mm_accdrv_process 34862 16.1 5.416 5.792 6.108 6.335 sum_up_and_integrate 110 10.3 0.001 0.002 6.242 6.257 integrate_v_rspace 110 11.3 0.003 0.003 6.216 6.230 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.699 5.707 calculate_rho_elec 110 8.6 0.059 0.059 5.698 5.707 init_scf_run 11 5.9 0.000 0.001 5.655 5.655 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.655 5.655 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.002 5.499 apply_single 110 13.6 0.000 0.000 5.001 5.499 qs_ot_get_p 110 10.4 0.001 0.001 5.343 5.492 make_m2s 4110 13.4 0.049 0.051 4.230 4.617 make_images 4110 14.4 0.400 0.525 4.115 4.502 ot_diis_step 99 11.5 0.011 0.011 4.471 4.471 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.182 3.825 multiply_cannon_sync_h2d 16440 15.4 3.221 3.778 3.221 3.778 qs_ot_p2m_diag 48 11.0 0.041 0.044 3.725 3.729 pw_transfer 1331 11.6 0.065 0.073 3.525 3.534 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.419 3.430 grid_integrate_task_list 110 12.3 3.204 3.386 3.204 3.386 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.356 3.357 density_rs2pw 110 9.6 0.004 0.005 2.936 3.093 fft_wrap_pw1pw2_140 451 13.1 0.579 0.588 2.964 2.975 wfi_extrapolate 11 7.9 0.001 0.001 2.971 2.971 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.959 2.961 hybrid_alltoall_any 4261 16.3 0.109 0.383 2.168 2.903 make_images_data 4110 15.4 0.045 0.050 2.385 2.894 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.820 2.833 cp_fm_diag_elpa_base 48 14.0 2.754 2.785 2.819 2.831 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.601 2.659 cp_fm_cholesky_invert 11 10.9 2.645 2.652 2.645 2.652 calculate_first_density_matrix 1 7.0 0.001 0.004 2.593 2.595 calculate_dm_sparse 110 9.5 0.001 0.001 2.560 2.592 multiply_cannon_metrocomm4 14385 15.4 0.048 0.052 0.899 2.508 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.430 2.498 mp_irecv_dv 48980 15.7 0.824 2.374 0.824 2.374 grid_collocate_task_list 110 9.6 2.217 2.370 2.217 2.370 fft3d_ps 1111 14.6 1.088 1.108 2.359 2.368 mp_sum_l 6594 12.7 1.482 2.133 1.482 2.133 potential_pw2rs 110 12.3 0.011 0.012 2.086 2.091 mp_alltoall_d11v 2046 13.8 1.853 2.083 1.853 2.083 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.980 1.981 dbcsr_complete_redistribute 325 12.2 0.333 0.365 1.514 1.954 cp_fm_upper_to_full 70 14.2 1.383 1.744 1.383 1.744 acc_transpose_blocks 16440 15.4 0.077 0.078 1.642 1.701 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.642 1.656 mp_allgather_i34 2055 14.4 0.491 1.602 0.491 1.602 cp_fm_cholesky_decompose 22 10.9 1.575 1.598 1.575 1.598 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.497 1.508 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.376 1.474 transfer_rs2pw 451 10.6 0.005 0.006 1.283 1.470 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.018 1.462 mp_waitany 17072 13.8 1.143 1.378 1.143 1.378 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.285 1.290 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.077000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=629.636364, yerr=8.025785 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 735.875072E+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 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 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 0.012 0.030 65.436 65.437 qs_mol_dyn_low 1 2.0 0.003 0.003 65.177 65.190 qs_forces 11 3.9 0.003 0.003 65.099 65.100 qs_energies 11 4.9 0.001 0.001 61.722 61.726 scf_env_do_scf 11 5.9 0.000 0.001 53.408 53.410 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.850 41.850 velocity_verlet 10 3.0 0.002 0.002 36.974 36.976 dbcsr_multiply_generic 2055 12.4 0.113 0.120 30.842 31.146 qs_scf_new_mos 99 7.5 0.001 0.001 27.495 27.606 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.494 27.606 ot_scf_mini 99 9.5 0.002 0.003 25.812 25.920 multiply_cannon 2055 13.4 0.237 0.256 23.868 24.931 multiply_cannon_loop 2055 14.4 1.432 1.495 22.552 23.383 ot_mini 99 10.5 0.001 0.001 15.058 15.162 multiply_cannon_multrec 24660 15.4 4.101 6.803 13.083 14.060 rebuild_ks_matrix 110 8.3 0.000 0.000 11.794 11.902 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.793 11.902 init_scf_loop 11 6.9 0.000 0.000 11.519 11.520 qs_ot_get_derivative 99 11.5 0.001 0.001 10.870 10.970 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.415 10.509 prepare_preconditioner 11 7.9 0.000 0.000 9.813 9.828 make_preconditioner 11 8.9 0.000 0.000 9.813 9.828 dbcsr_mm_accdrv_process 52304 16.0 7.573 9.101 8.825 9.713 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.048 9.514 mp_waitall_1 121746 16.5 4.262 6.250 4.262 6.250 sum_up_and_integrate 110 10.3 0.001 0.002 6.150 6.161 integrate_v_rspace 110 11.3 0.003 0.004 6.124 6.135 qs_ot_get_p 110 10.4 0.001 0.002 5.678 5.829 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.713 5.725 calculate_rho_elec 110 8.6 0.078 0.081 5.713 5.725 init_scf_run 11 5.9 0.000 0.001 5.602 5.603 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.602 5.602 make_m2s 4110 13.4 0.059 0.061 5.317 5.588 make_images 4110 14.4 0.579 0.703 5.176 5.441 cp_fm_upper_to_full 70 14.2 3.367 4.871 3.367 4.871 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.054 4.179 apply_single 110 13.6 0.000 0.000 4.054 4.179 ot_diis_step 99 11.5 0.011 0.011 4.148 4.149 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.844 3.862 dbcsr_complete_redistribute 325 12.2 0.424 0.462 2.661 3.786 pw_transfer 1331 11.6 0.065 0.073 3.587 3.618 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.516 3.574 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.481 3.514 grid_integrate_task_list 110 12.3 3.279 3.444 3.279 3.444 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.352 3.402 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.284 3.284 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.101 3.217 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.550 3.209 calculate_dm_sparse 110 9.5 0.001 0.001 3.076 3.119 density_rs2pw 110 9.6 0.004 0.004 2.859 3.062 fft_wrap_pw1pw2_140 451 13.1 0.610 0.633 3.006 3.039 wfi_extrapolate 11 7.9 0.001 0.001 2.956 2.956 make_images_data 4110 15.4 0.049 0.053 2.593 2.899 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.863 2.864 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.755 2.851 mp_alltoall_i22 605 13.7 1.644 2.826 1.644 2.826 hybrid_alltoall_any 4261 16.3 0.123 0.463 2.242 2.803 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.761 2.771 cp_fm_diag_elpa_base 48 14.0 2.614 2.675 2.759 2.769 acc_transpose_blocks 24660 15.4 0.114 0.117 2.565 2.724 cp_fm_cholesky_invert 11 10.9 2.631 2.639 2.631 2.639 calculate_first_density_matrix 1 7.0 0.000 0.000 2.552 2.555 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.514 2.547 multiply_cannon_sync_h2d 24660 15.4 2.357 2.463 2.357 2.463 grid_collocate_task_list 110 9.6 2.262 2.402 2.262 2.402 fft3d_ps 1111 14.6 1.083 1.114 2.361 2.381 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.222 2.222 potential_pw2rs 110 12.3 0.013 0.014 1.983 1.988 mp_alltoall_d11v 2046 13.8 1.701 1.941 1.701 1.941 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.807 1.840 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.614 1.719 mp_sum_l 6594 12.7 0.970 1.700 0.970 1.700 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.637 1.654 cp_fm_cholesky_decompose 22 10.9 1.602 1.642 1.602 1.642 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.619 1.632 acc_transpose_blocks_sync 73980 16.4 1.441 1.600 1.441 1.600 multiply_cannon_metrocomm4 20550 15.4 0.062 0.065 0.864 1.482 mp_allgather_i34 2055 14.4 0.407 1.448 0.407 1.448 mp_irecv_dv 62702 16.1 0.759 1.400 0.759 1.400 transfer_rs2pw 451 10.6 0.005 0.006 1.143 1.353 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.437000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.909091, yerr=10.448615 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 860.635136E+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 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 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 0.011 0.031 55.666 55.667 qs_mol_dyn_low 1 2.0 0.003 0.003 55.402 55.427 qs_forces 11 3.9 0.003 0.003 55.321 55.322 qs_energies 11 4.9 0.001 0.001 51.682 51.687 scf_env_do_scf 11 5.9 0.000 0.001 43.423 43.423 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.565 35.566 velocity_verlet 10 3.0 0.002 0.002 31.341 31.345 dbcsr_multiply_generic 2055 12.4 0.105 0.110 23.450 23.592 qs_scf_new_mos 99 7.5 0.001 0.001 21.128 21.205 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.127 21.204 ot_scf_mini 99 9.5 0.002 0.002 19.880 19.946 multiply_cannon 2055 13.4 0.237 0.252 17.880 19.036 multiply_cannon_loop 2055 14.4 0.601 0.620 16.633 16.979 rebuild_ks_matrix 110 8.3 0.000 0.000 11.538 11.603 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.538 11.602 ot_mini 99 10.5 0.001 0.001 10.954 11.022 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.257 10.315 multiply_cannon_multrec 8220 15.4 3.211 4.792 7.627 8.834 init_scf_loop 11 6.9 0.000 0.000 7.809 7.810 mp_waitall_1 103326 16.6 6.024 7.549 6.024 7.549 qs_ot_get_derivative 99 11.5 0.001 0.001 7.196 7.263 sum_up_and_integrate 110 10.3 0.001 0.002 6.191 6.203 integrate_v_rspace 110 11.3 0.003 0.003 6.164 6.176 prepare_preconditioner 11 7.9 0.000 0.000 6.152 6.159 make_preconditioner 11 8.9 0.000 0.000 6.152 6.159 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.939 5.954 calculate_rho_elec 110 8.6 0.114 0.114 5.938 5.953 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.743 5.814 dbcsr_mm_accdrv_process 17442 15.9 3.065 4.144 4.276 5.209 init_scf_run 11 5.9 0.000 0.002 5.141 5.141 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.141 5.141 qs_ot_get_p 110 10.4 0.002 0.002 4.982 5.044 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 3.146 4.451 make_m2s 4110 13.4 0.038 0.039 4.187 4.377 make_images 4110 14.4 0.649 0.714 4.058 4.244 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.800 3.866 apply_single 110 13.6 0.000 0.000 3.800 3.865 pw_transfer 1331 11.6 0.065 0.070 3.835 3.849 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.728 3.744 ot_diis_step 99 11.5 0.012 0.012 3.735 3.735 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.575 3.579 grid_integrate_task_list 110 12.3 3.383 3.519 3.383 3.519 fft_wrap_pw1pw2_140 451 13.1 0.777 0.787 3.219 3.239 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.231 3.232 density_rs2pw 110 9.6 0.004 0.004 2.928 3.048 cp_fm_cholesky_invert 11 10.9 3.010 3.014 3.010 3.014 wfi_extrapolate 11 7.9 0.001 0.001 2.726 2.726 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.653 2.661 cp_fm_diag_elpa_base 48 14.0 2.598 2.620 2.651 2.659 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.656 2.656 hybrid_alltoall_any 4261 16.3 0.200 0.856 2.223 2.587 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.550 2.552 calculate_dm_sparse 110 9.5 0.001 0.001 2.503 2.547 grid_collocate_task_list 110 9.6 2.370 2.523 2.370 2.523 multiply_cannon_sync_h2d 8220 15.4 2.341 2.480 2.341 2.480 make_images_data 4110 15.4 0.040 0.046 2.209 2.478 fft3d_ps 1111 14.6 1.132 1.141 2.394 2.409 calculate_first_density_matrix 1 7.0 0.000 0.000 2.319 2.320 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.175 2.198 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.045 2.093 potential_pw2rs 110 12.3 0.016 0.016 1.989 1.993 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.792 1.987 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.954 1.987 mp_alltoall_d11v 2046 13.8 1.715 1.922 1.715 1.922 cp_fm_cholesky_decompose 22 10.9 1.707 1.719 1.707 1.719 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.609 1.614 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.484 1.602 dbcsr_complete_redistribute 325 12.2 0.555 0.581 1.443 1.523 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.461 1.475 mp_allgather_i34 2055 14.4 0.420 1.459 0.420 1.459 acc_transpose_blocks 8220 15.4 0.039 0.041 1.397 1.424 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.480 1.401 jit_kernel_multiply 8 15.6 0.898 1.397 0.898 1.397 mp_irecv_dv 24056 15.7 0.453 1.356 0.453 1.356 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.827 1.331 qs_create_task_list 11 7.9 0.000 0.000 1.207 1.307 generate_qs_task_list 11 8.9 0.372 0.439 1.206 1.307 transfer_rs2pw 451 10.6 0.005 0.005 1.124 1.264 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.221 1.258 mp_waitany 9240 13.8 1.041 1.196 1.041 1.196 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.667000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=807.090909, yerr=14.859674 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 1.401041E+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 67098. MP_Allreduce 9752 812. 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 0.017 0.036 86.148 86.149 qs_mol_dyn_low 1 2.0 0.003 0.003 85.795 85.829 qs_forces 11 3.9 0.003 0.004 85.695 85.695 qs_energies 11 4.9 0.002 0.002 81.587 81.588 scf_env_do_scf 11 5.9 0.000 0.001 71.524 71.525 velocity_verlet 10 3.0 0.002 0.002 54.981 54.991 scf_env_do_scf_inner_loop 99 6.5 0.003 0.009 43.580 43.581 dbcsr_multiply_generic 2055 12.4 0.118 0.125 29.429 29.520 init_scf_loop 11 6.9 0.000 0.000 27.874 27.877 qs_scf_new_mos 99 7.5 0.001 0.001 26.517 26.551 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.516 26.550 prepare_preconditioner 11 7.9 0.000 0.000 25.903 25.911 make_preconditioner 11 8.9 0.000 0.000 25.903 25.911 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.150 25.374 ot_scf_mini 99 9.5 0.002 0.002 24.727 24.756 multiply_cannon 2055 13.4 0.334 0.357 22.488 23.265 multiply_cannon_loop 2055 14.4 0.816 0.828 20.715 20.986 cp_fm_upper_to_full 70 14.2 12.617 18.087 12.617 18.087 ot_mini 99 10.5 0.001 0.001 13.962 13.992 rebuild_ks_matrix 110 8.3 0.001 0.001 13.434 13.470 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.434 13.469 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.186 12.219 dbcsr_complete_redistribute 325 12.2 1.016 1.036 7.292 10.452 multiply_cannon_multrec 8220 15.4 4.086 4.242 9.862 9.984 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.290 9.451 qs_ot_get_derivative 99 11.5 0.001 0.001 9.410 9.440 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.737 8.874 mp_alltoall_i22 605 13.7 5.349 8.490 5.349 8.490 mp_waitall_1 84994 16.7 7.331 8.188 7.331 8.188 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.985 7.019 calculate_rho_elec 110 8.6 0.224 0.225 6.985 7.018 sum_up_and_integrate 110 10.3 0.002 0.002 6.643 6.659 integrate_v_rspace 110 11.3 0.004 0.004 6.615 6.631 dbcsr_mm_accdrv_process 11614 15.7 3.834 4.062 5.629 5.842 init_scf_run 11 5.9 0.000 0.001 5.810 5.811 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.810 5.810 make_m2s 4110 13.4 0.044 0.044 5.260 5.754 make_images 4110 14.4 0.884 0.937 5.071 5.564 qs_ot_get_p 110 10.4 0.001 0.002 5.496 5.528 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.705 5.184 apply_single 110 13.6 0.000 0.000 4.705 5.184 cp_fm_cholesky_invert 11 10.9 5.057 5.061 5.057 5.061 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.754 5.049 pw_transfer 1331 11.6 0.074 0.074 4.849 4.860 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.733 4.743 ot_diis_step 99 11.5 0.015 0.015 4.537 4.537 fft_wrap_pw1pw2_140 451 13.1 1.279 1.289 4.144 4.154 qs_ot_p2m_diag 48 11.0 0.151 0.155 3.962 3.969 grid_integrate_task_list 110 12.3 3.670 3.735 3.670 3.735 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 3.647 3.648 hybrid_alltoall_any 4261 16.3 0.263 0.567 2.795 3.498 density_rs2pw 110 9.6 0.004 0.004 3.457 3.471 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.448 3.449 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.981 3.445 make_images_data 4110 15.4 0.045 0.049 2.754 3.353 wfi_extrapolate 11 7.9 0.001 0.001 3.271 3.271 calculate_dm_sparse 110 9.5 0.001 0.001 3.196 3.227 multiply_cannon_sync_h2d 8220 15.4 3.138 3.147 3.138 3.147 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.009 3.011 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.871 2.872 cp_fm_diag_elpa_base 48 14.0 2.332 2.525 2.869 2.869 fft3d_ps 1111 14.6 1.296 1.302 2.786 2.796 grid_collocate_task_list 110 9.6 2.662 2.678 2.662 2.678 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.611 2.627 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.517 2.531 calculate_first_density_matrix 1 7.0 0.000 0.000 2.427 2.428 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.168 2.230 potential_pw2rs 110 12.3 0.022 0.022 2.197 2.200 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.101 2.186 mp_alltoall_d11v 2046 13.8 2.021 2.068 2.021 2.068 cp_fm_cholesky_decompose 22 10.9 1.937 1.962 1.937 1.962 qs_create_task_list 11 7.9 0.001 0.001 1.875 1.922 generate_qs_task_list 11 8.9 0.729 0.784 1.874 1.921 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.848 1.852 jit_kernel_multiply 10 15.2 1.599 1.847 1.599 1.847 acc_transpose_blocks 8220 15.4 0.041 0.041 1.786 1.799 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.752 1.786 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.149000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1262.181818, yerr=63.854690 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 631.271424E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57761. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 2529087. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.029 205.454 205.454 qs_mol_dyn_low 1 2.0 0.003 0.004 204.996 205.013 qs_forces 11 3.9 0.005 0.005 204.888 204.889 qs_energies 11 4.9 0.002 0.002 199.311 199.326 scf_env_do_scf 11 5.9 0.001 0.001 182.701 182.705 scf_env_do_scf_inner_loop 117 6.6 0.003 0.007 162.054 162.057 dbcsr_multiply_generic 2507 12.6 0.180 0.184 124.427 125.060 velocity_verlet 10 3.0 0.001 0.002 124.110 124.111 qs_scf_new_mos 117 7.6 0.001 0.001 122.489 122.821 qs_scf_loop_do_ot 117 8.6 0.001 0.001 122.489 122.820 ot_scf_mini 117 9.6 0.003 0.003 115.816 116.133 multiply_cannon 2507 13.6 0.240 0.251 101.568 103.544 multiply_cannon_loop 2507 14.6 2.329 2.433 99.388 101.305 ot_mini 117 10.6 0.001 0.001 65.240 65.551 multiply_cannon_multrec 60168 15.6 32.248 34.735 41.882 43.614 qs_ot_get_derivative 117 11.6 0.001 0.001 40.250 40.546 rebuild_ks_matrix 128 8.3 0.001 0.001 33.717 34.258 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.018 33.716 34.257 mp_waitall_1 267128 16.5 28.175 30.813 28.175 30.813 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.289 30.762 multiply_cannon_sync_h2d 60168 15.6 26.887 29.269 26.887 29.269 qs_ot_get_p 128 10.4 0.001 0.001 28.468 28.852 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.544 25.418 apply_single 128 13.6 0.001 0.001 24.544 25.418 ot_diis_step 117 11.6 0.008 0.008 24.751 24.752 qs_ot_p2m_diag 83 11.4 0.079 0.092 21.835 21.916 init_scf_loop 11 6.9 0.000 0.000 20.573 20.575 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.412 19.413 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.113 18.306 multiply_cannon_metrocomm3 60168 15.6 0.119 0.123 15.606 17.796 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.419 16.448 cp_fm_diag_elpa_base 83 14.4 16.350 16.387 16.416 16.446 prepare_preconditioner 11 7.9 0.000 0.000 15.929 15.976 make_preconditioner 11 8.9 0.000 0.000 15.929 15.976 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.157 15.341 make_m2s 5014 13.6 0.105 0.114 13.752 14.156 sum_up_and_integrate 128 10.3 0.002 0.004 14.032 14.047 integrate_v_rspace 128 11.3 0.003 0.004 13.973 13.992 make_images 5014 14.6 0.396 0.417 13.568 13.982 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.143 13.207 calculate_rho_elec 128 8.7 0.045 0.064 13.142 13.206 init_scf_run 11 5.9 0.000 0.002 12.436 12.437 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.436 12.436 dbcsr_mm_accdrv_process 124484 16.2 4.482 4.761 9.190 9.797 mp_sum_l 7950 12.9 7.837 9.314 7.837 9.314 cp_fm_cholesky_invert 11 10.9 9.161 9.170 9.161 9.170 wfi_extrapolate 11 7.9 0.001 0.001 9.072 9.072 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.641 8.794 calculate_dm_sparse 128 9.5 0.001 0.001 8.532 8.646 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.919 8.080 multiply_cannon_metrocomm1 60168 15.6 0.097 0.101 6.060 7.967 pw_transfer 1547 11.6 0.073 0.088 7.613 7.792 make_images_data 5014 15.6 0.066 0.073 6.737 7.648 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.412 7.593 density_rs2pw 128 9.7 0.006 0.007 6.839 7.585 grid_integrate_task_list 128 12.3 7.002 7.488 7.002 7.488 hybrid_alltoall_any 5200 16.5 0.297 2.279 5.903 7.234 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.758 6.768 fft_wrap_pw1pw2_140 523 13.2 1.123 1.164 6.495 6.671 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.886 6.088 fft3d_ps 1291 14.7 2.192 2.840 5.422 5.751 mp_alltoall_d11v 2415 14.1 4.295 5.564 4.295 5.564 grid_collocate_task_list 128 9.7 4.842 5.254 4.842 5.254 cp_fm_cholesky_decompose 22 10.9 4.582 4.597 4.582 4.597 potential_pw2rs 128 12.3 0.009 0.010 4.545 4.565 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.454000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.272727, yerr=6.743224 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 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 11444707676160 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 92.796579E+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 6705500928 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 842.752000E+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 1453843137296 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4005 58170. MP_Allreduce 11091 960. MP_Sync 86 MP_Alltoall 1955 5646024. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 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 0.011 0.030 187.864 187.864 qs_mol_dyn_low 1 2.0 0.003 0.004 187.438 187.474 qs_forces 11 3.9 0.004 0.005 187.257 187.259 qs_energies 11 4.9 0.001 0.002 180.505 180.516 scf_env_do_scf 11 5.9 0.001 0.001 164.201 164.212 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 131.585 131.587 velocity_verlet 10 3.0 0.001 0.002 117.709 117.712 dbcsr_multiply_generic 2485 12.5 0.187 0.194 96.905 98.170 qs_scf_new_mos 116 7.6 0.001 0.001 93.523 94.083 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.522 94.083 ot_scf_mini 116 9.6 0.004 0.004 88.743 89.316 multiply_cannon 2485 13.5 0.497 0.559 76.795 81.316 multiply_cannon_loop 2485 14.5 1.553 1.620 73.440 76.126 ot_mini 116 10.6 0.001 0.001 50.091 50.657 mp_waitall_1 212858 16.6 24.134 38.390 24.134 38.390 multiply_cannon_multrec 29820 15.5 20.774 24.991 31.308 35.961 rebuild_ks_matrix 127 8.3 0.001 0.001 31.952 32.700 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 31.951 32.700 init_scf_loop 11 6.9 0.000 0.000 32.527 32.527 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.700 29.366 qs_ot_get_derivative 116 11.6 0.001 0.001 28.144 28.712 multiply_cannon_metrocomm3 29820 15.5 0.095 0.099 15.738 28.340 prepare_preconditioner 11 7.9 0.000 0.000 28.233 28.315 make_preconditioner 11 8.9 0.000 0.000 28.232 28.315 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.913 27.441 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 22.147 23.309 apply_single 127 13.6 0.001 0.001 22.147 23.308 ot_diis_step 116 11.6 0.014 0.015 21.773 21.775 qs_ot_get_p 127 10.4 0.001 0.001 20.825 21.507 multiply_cannon_sync_h2d 29820 15.5 17.775 20.377 17.775 20.377 cp_fm_cholesky_invert 11 10.9 16.496 16.508 16.496 16.508 qs_ot_p2m_diag 82 11.4 0.185 0.214 16.016 16.050 make_m2s 4970 13.5 0.089 0.096 14.184 15.861 make_images 4970 14.5 1.162 1.360 13.971 15.648 cp_dbcsr_syevd 82 12.4 0.005 0.006 14.823 14.824 sum_up_and_integrate 127 10.3 0.002 0.004 13.832 13.858 integrate_v_rspace 127 11.3 0.003 0.003 13.772 13.802 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.200 13.237 calculate_rho_elec 127 8.7 0.087 0.103 13.199 13.236 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.698 11.724 cp_fm_diag_elpa_base 82 14.4 11.450 11.550 11.693 11.714 init_scf_run 11 5.9 0.000 0.002 11.511 11.513 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.511 11.513 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.028 11.417 multiply_cannon_metrocomm4 27335 15.5 0.103 0.117 3.844 10.755 dbcsr_mm_accdrv_process 61726 16.2 5.443 6.144 9.989 10.576 mp_irecv_dv 68888 16.3 3.642 10.361 3.642 10.361 make_images_data 4970 15.5 0.066 0.075 8.389 10.350 hybrid_alltoall_any 5155 16.4 0.349 1.552 7.104 9.788 pw_transfer 1535 11.6 0.084 0.103 8.508 8.563 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.287 8.344 wfi_extrapolate 11 7.9 0.001 0.001 8.335 8.335 grid_integrate_task_list 127 12.3 7.127 7.478 7.127 7.478 density_rs2pw 127 9.7 0.006 0.007 6.971 7.471 fft_wrap_pw1pw2_140 519 13.2 1.206 1.224 7.348 7.420 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.402 7.162 cp_fm_cholesky_decompose 22 10.9 6.828 6.913 6.828 6.913 calculate_dm_sparse 127 9.5 0.001 0.001 6.446 6.573 mp_sum_l 7884 12.9 4.138 6.355 4.138 6.355 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.141 6.151 fft3d_ps 1281 14.7 2.821 2.985 5.862 5.912 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.238 5.451 grid_collocate_task_list 127 9.7 4.960 5.413 4.960 5.413 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.352 5.413 mp_allgather_i34 2485 14.5 1.894 5.025 1.894 5.025 mp_alltoall_d11v 2401 14.1 4.083 4.586 4.083 4.586 potential_pw2rs 127 12.3 0.016 0.018 4.432 4.449 dbcsr_complete_redistribute 393 12.7 0.772 0.858 3.177 4.027 mp_sum_d 4453 12.1 2.589 3.978 2.589 3.978 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=187.864000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=802.727273, yerr=4.024717 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 2.951022E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 4017216 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.3 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 937.705472E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1052064 MPI messages size (bytes): total size 2.737208E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.601750E+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 284544 37295751168 131072 < size <= 4194304 665856 1004519030784 4194304 < size <= 16777216 66080 937891119088 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57629. MP_Allreduce 11184 996. MP_Sync 88 MP_Alltoall 1724 9394234. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.028 176.022 176.022 qs_mol_dyn_low 1 2.0 0.003 0.004 175.623 175.637 qs_forces 11 3.9 0.004 0.006 175.464 175.466 qs_energies 11 4.9 0.001 0.002 168.935 168.946 scf_env_do_scf 11 5.9 0.001 0.001 153.459 153.462 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 118.790 118.791 velocity_verlet 10 3.0 0.001 0.002 113.075 113.079 dbcsr_multiply_generic 2529 12.6 0.182 0.186 82.720 83.985 qs_scf_new_mos 118 7.6 0.001 0.001 81.380 81.696 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.379 81.695 ot_scf_mini 118 9.6 0.003 0.004 77.101 77.482 multiply_cannon 2529 13.6 0.514 0.542 63.026 67.784 multiply_cannon_loop 2529 14.6 1.143 1.225 59.945 62.661 ot_mini 118 10.6 0.001 0.001 43.206 43.573 init_scf_loop 11 6.9 0.000 0.000 34.571 34.573 mp_waitall_1 172006 16.6 24.698 34.488 24.698 34.488 rebuild_ks_matrix 129 8.3 0.001 0.001 30.387 30.837 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.017 30.386 30.836 prepare_preconditioner 11 7.9 0.000 0.000 30.560 30.598 make_preconditioner 11 8.9 0.000 0.000 30.560 30.598 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.231 29.615 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.415 27.819 multiply_cannon_multrec 20232 15.6 13.197 16.088 22.697 25.776 multiply_cannon_metrocomm3 20232 15.6 0.062 0.065 15.403 25.356 qs_ot_get_derivative 118 11.6 0.001 0.002 23.557 23.941 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.698 20.879 apply_single 129 13.6 0.001 0.001 19.698 20.879 qs_ot_get_p 129 10.4 0.001 0.002 19.155 19.602 ot_diis_step 118 11.6 0.018 0.018 19.535 19.535 make_m2s 5058 13.6 0.080 0.086 14.606 15.997 make_images 5058 14.6 1.160 1.258 14.370 15.758 multiply_cannon_sync_h2d 20232 15.6 13.848 15.073 13.848 15.073 qs_ot_p2m_diag 84 11.4 0.268 0.275 14.801 14.809 cp_fm_cholesky_invert 11 10.9 14.221 14.230 14.221 14.230 sum_up_and_integrate 129 10.3 0.002 0.003 14.062 14.087 integrate_v_rspace 129 11.3 0.003 0.004 14.002 14.026 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.699 13.740 calculate_rho_elec 129 8.7 0.131 0.146 13.699 13.739 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.736 13.738 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.497 10.521 cp_fm_diag_elpa_base 84 14.4 10.083 10.253 10.494 10.518 init_scf_run 11 5.9 0.000 0.001 10.468 10.468 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.468 10.468 make_images_data 5058 15.6 0.063 0.074 8.656 10.442 hybrid_alltoall_any 5245 16.5 0.453 2.057 7.560 9.749 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.252 9.526 multiply_cannon_metrocomm4 17703 15.6 0.067 0.077 3.487 9.358 dbcsr_mm_accdrv_process 41846 16.2 5.728 6.148 8.946 9.108 mp_irecv_dv 50659 16.2 3.358 9.094 3.358 9.094 pw_transfer 1559 11.6 0.085 0.102 8.764 8.879 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.540 8.659 grid_integrate_task_list 129 12.3 7.388 7.848 7.388 7.848 fft_wrap_pw1pw2_140 527 13.2 1.311 1.341 7.566 7.698 wfi_extrapolate 11 7.9 0.001 0.001 7.431 7.431 density_rs2pw 129 9.7 0.006 0.006 6.986 7.352 cp_fm_cholesky_decompose 22 10.9 7.252 7.279 7.252 7.279 cp_fm_upper_to_full 106 14.8 5.698 7.204 5.698 7.204 dbcsr_complete_redistribute 397 12.7 1.198 1.228 4.588 6.378 calculate_dm_sparse 129 9.5 0.001 0.001 5.885 5.985 fft3d_ps 1301 14.7 2.784 3.010 5.896 5.976 grid_collocate_task_list 129 9.7 5.186 5.629 5.186 5.629 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.493 5.501 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.795 5.470 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.426 5.213 mp_sum_l 8016 12.9 3.233 4.788 3.233 4.788 mp_alltoall_d11v 2429 14.1 4.329 4.767 4.329 4.767 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.561 4.708 potential_pw2rs 129 12.3 0.021 0.023 4.408 4.422 mp_allgather_i34 2529 14.6 1.572 4.384 1.572 4.384 transfer_fm_to_dbcsr 11 9.9 0.019 0.024 2.310 4.066 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.967 4.013 mp_alltoall_i22 720 14.1 1.943 3.888 1.943 3.888 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.724 3.725 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.022000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=887.818182, yerr=5.796836 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 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 4.320341E+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 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 1.170530E+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 660837747920 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3949 58971. MP_Allreduce 10950 1087. MP_Sync 86 MP_Alltoall 1700 12496288. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 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 0.012 0.033 186.022 186.023 qs_mol_dyn_low 1 2.0 0.003 0.003 185.603 185.636 qs_forces 11 3.9 0.004 0.004 185.437 185.449 qs_energies 11 4.9 0.001 0.002 178.369 178.380 scf_env_do_scf 11 5.9 0.001 0.001 161.862 161.875 velocity_verlet 10 3.0 0.002 0.002 123.115 123.121 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.678 115.679 dbcsr_multiply_generic 2485 12.5 0.185 0.192 80.011 80.877 qs_scf_new_mos 116 7.6 0.001 0.001 80.042 80.373 qs_scf_loop_do_ot 116 8.6 0.001 0.001 80.041 80.372 ot_scf_mini 116 9.6 0.003 0.003 75.508 75.827 multiply_cannon 2485 13.5 0.550 0.581 55.603 57.656 multiply_cannon_loop 2485 14.5 1.838 1.939 51.876 53.979 init_scf_loop 11 6.9 0.000 0.000 46.062 46.063 ot_mini 116 10.6 0.001 0.001 42.693 42.993 prepare_preconditioner 11 7.9 0.000 0.000 41.992 42.008 make_preconditioner 11 8.9 0.000 0.000 41.992 42.007 make_full_inverse_cholesky 11 9.9 0.012 0.023 35.653 40.594 multiply_cannon_multrec 29820 15.5 13.556 18.572 26.436 31.061 rebuild_ks_matrix 127 8.3 0.001 0.001 28.864 29.184 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 28.863 29.183 mp_waitall_1 146592 16.7 17.012 27.865 17.012 27.865 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.013 26.303 qs_ot_get_derivative 116 11.6 0.001 0.002 23.211 23.530 make_m2s 4970 13.5 0.096 0.100 20.113 21.155 make_images 4970 14.5 1.940 2.255 19.804 20.849 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.917 19.508 apply_single 127 13.6 0.001 0.001 18.917 19.508 qs_ot_get_p 127 10.4 0.001 0.002 19.053 19.388 ot_diis_step 116 11.6 0.018 0.020 19.356 19.357 cp_fm_upper_to_full 104 14.8 11.375 16.724 11.375 16.724 cp_fm_cholesky_invert 11 10.9 16.044 16.053 16.044 16.053 multiply_cannon_metrocomm3 29820 15.5 0.049 0.053 6.276 15.732 qs_ot_p2m_diag 82 11.4 0.338 0.385 14.908 14.958 sum_up_and_integrate 127 10.3 0.002 0.003 13.750 13.779 integrate_v_rspace 127 11.3 0.003 0.004 13.691 13.721 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.535 13.535 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.458 13.507 calculate_rho_elec 127 8.7 0.171 0.187 13.458 13.506 dbcsr_mm_accdrv_process 61748 16.2 8.404 9.163 12.455 12.954 dbcsr_complete_redistribute 393 12.7 1.501 1.618 9.045 12.725 make_images_data 4970 15.5 0.065 0.072 10.832 12.617 hybrid_alltoall_any 5155 16.4 0.524 2.193 9.732 11.628 multiply_cannon_sync_h2d 29820 15.5 10.367 11.578 10.367 11.578 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.689 11.353 init_scf_run 11 5.9 0.000 0.001 10.632 10.634 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.632 10.634 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.353 10.370 cp_fm_diag_elpa_base 82 14.4 9.408 9.730 10.347 10.363 transfer_fm_to_dbcsr 11 9.9 0.002 0.009 6.316 9.913 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.634 9.856 mp_alltoall_i22 712 14.1 5.526 9.261 5.526 9.261 pw_transfer 1535 11.6 0.084 0.098 8.769 8.842 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.545 8.623 grid_integrate_task_list 127 12.3 7.468 7.997 7.468 7.997 fft_wrap_pw1pw2_140 519 13.2 1.433 1.470 7.636 7.729 cp_fm_cholesky_decompose 22 10.9 7.491 7.579 7.491 7.579 wfi_extrapolate 11 7.9 0.001 0.001 7.551 7.551 multiply_cannon_metrocomm4 24850 15.5 0.083 0.093 2.761 7.067 density_rs2pw 127 9.7 0.006 0.006 6.607 6.962 mp_irecv_dv 75445 16.2 2.608 6.792 2.608 6.792 calculate_dm_sparse 127 9.5 0.001 0.001 6.210 6.300 fft3d_ps 1281 14.7 2.829 2.904 5.747 5.805 grid_collocate_task_list 127 9.7 5.247 5.666 5.247 5.666 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.438 5.488 mp_alltoall_d11v 2401 14.1 4.785 5.285 4.785 5.285 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.510 4.617 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.393 4.468 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 4.391 4.392 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.202 4.265 potential_pw2rs 127 12.3 0.023 0.023 4.148 4.158 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.023000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1093.636364, yerr=22.276252 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243750E+12 0.0% 0.0% 100.0% flops max/rank 5.910792E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806546176 0.0% 0.0% 100.0% number of processed stacks 1978768 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 1.535418E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 243264 MPI messages size (bytes): total size 1.342058E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.516877E+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 115488 60548972544 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592388592 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9025 51. MP_Alltoall 9734 793691. MP_ISend 40500 2096701. MP_IRecv 40500 2095807. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57624. MP_Allreduce 11184 1163. MP_Sync 88 MP_Alltoall 1724 18848034. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.032 168.113 168.114 qs_mol_dyn_low 1 2.0 0.003 0.004 167.627 167.640 qs_forces 11 3.9 0.008 0.010 167.520 167.523 qs_energies 11 4.9 0.001 0.002 160.192 160.196 scf_env_do_scf 11 5.9 0.001 0.001 143.079 143.089 velocity_verlet 10 3.0 0.002 0.002 110.915 110.919 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 107.768 107.770 dbcsr_multiply_generic 2534 12.6 0.182 0.193 72.070 72.438 qs_scf_new_mos 118 7.6 0.001 0.001 71.063 71.168 qs_scf_loop_do_ot 118 8.6 0.001 0.001 71.062 71.167 ot_scf_mini 118 9.6 0.003 0.004 66.636 66.714 multiply_cannon 2534 13.6 0.569 0.597 53.565 57.072 multiply_cannon_loop 2534 14.6 0.811 0.842 50.640 51.346 ot_mini 118 10.6 0.001 0.001 37.553 37.621 init_scf_loop 11 6.9 0.000 0.000 35.162 35.164 prepare_preconditioner 11 7.9 0.000 0.000 31.341 31.368 make_preconditioner 11 8.9 0.000 0.000 31.341 31.368 mp_waitall_1 127116 16.7 24.604 30.857 24.604 30.857 make_full_inverse_cholesky 11 9.9 0.016 0.026 29.259 29.549 rebuild_ks_matrix 129 8.3 0.001 0.001 28.965 29.056 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 28.965 29.055 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.464 26.546 multiply_cannon_multrec 10136 15.6 10.492 14.716 18.158 21.427 qs_ot_get_derivative 118 11.6 0.001 0.002 20.447 20.523 multiply_cannon_metrocomm3 10136 15.6 0.026 0.028 12.622 19.801 cp_fm_cholesky_invert 11 10.9 17.868 17.874 17.868 17.874 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.083 17.243 apply_single 129 13.6 0.001 0.001 17.083 17.242 ot_diis_step 118 11.6 0.020 0.020 17.035 17.036 qs_ot_get_p 129 10.4 0.004 0.006 16.338 16.408 make_m2s 5068 13.6 0.067 0.073 14.689 15.537 make_images 5068 14.6 2.186 2.609 14.379 15.219 sum_up_and_integrate 129 10.3 0.002 0.002 14.041 14.086 integrate_v_rspace 129 11.3 0.004 0.004 13.980 14.027 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.941 13.974 calculate_rho_elec 129 8.7 0.258 0.269 13.941 13.974 qs_ot_p2m_diag 84 11.4 0.501 0.506 12.637 12.654 cp_dbcsr_syevd 84 12.4 0.005 0.006 11.473 11.473 multiply_cannon_sync_h2d 10136 15.6 10.980 11.368 10.980 11.368 init_scf_run 11 5.9 0.000 0.001 10.315 10.316 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.315 10.316 hybrid_alltoall_any 5255 16.5 0.847 3.796 8.343 9.732 make_images_data 5068 15.6 0.055 0.066 8.474 9.728 pw_transfer 1559 11.6 0.085 0.094 9.172 9.203 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.948 8.985 cp_fm_diag_elpa 84 13.4 0.000 0.000 8.617 8.630 cp_fm_diag_elpa_base 84 14.4 8.381 8.463 8.614 8.627 grid_integrate_task_list 129 12.3 7.813 8.122 7.813 8.122 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.015 8.072 cp_fm_cholesky_decompose 22 10.9 7.837 8.050 7.837 8.050 dbcsr_mm_accdrv_process 20954 16.1 3.339 4.212 7.310 8.010 fft_wrap_pw1pw2_140 527 13.2 1.818 1.847 7.962 8.001 wfi_extrapolate 11 7.9 0.001 0.001 7.284 7.284 multiply_cannon_metrocomm1 10136 15.6 0.030 0.031 4.323 7.092 density_rs2pw 129 9.7 0.006 0.006 6.607 6.813 calculate_dm_sparse 129 9.5 0.001 0.001 6.101 6.183 grid_collocate_task_list 129 9.7 5.617 5.884 5.617 5.884 fft3d_ps 1301 14.7 2.784 2.861 5.651 5.687 dbcsr_complete_redistribute 397 12.7 2.101 2.149 5.092 5.469 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.243 5.245 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.116 5.124 mp_alltoall_d11v 2429 14.1 4.456 4.977 4.456 4.977 mp_allgather_i34 2534 14.6 1.107 4.590 1.107 4.590 potential_pw2rs 129 12.3 0.027 0.028 4.157 4.165 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.054 4.087 multiply_cannon_metrocomm4 7602 15.6 0.027 0.029 1.727 4.060 mp_irecv_dv 29142 15.9 1.687 3.980 1.687 3.980 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.572 3.873 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.611 3.675 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.307 3.636 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.585 3.622 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.410 3.418 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.346 3.408 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=168.114000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1463.363636, yerr=1.493789 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/20/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 11.606413E+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 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.112661E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108580288 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 59065. MP_Allreduce 11095 1504. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.074 0.122 287.471 287.488 qs_mol_dyn_low 1 2.0 0.003 0.003 286.846 286.880 qs_forces 11 3.9 0.007 0.024 286.713 286.718 qs_energies 11 4.9 0.002 0.002 277.887 277.897 scf_env_do_scf 11 5.9 0.001 0.001 255.330 255.336 velocity_verlet 10 3.0 0.002 0.002 207.924 207.936 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 131.667 131.668 init_scf_loop 11 6.9 0.000 0.000 123.387 123.389 prepare_preconditioner 11 7.9 0.000 0.000 118.594 118.618 make_preconditioner 11 8.9 0.000 0.000 118.594 118.618 make_full_inverse_cholesky 11 9.9 0.038 0.039 94.887 115.733 qs_scf_new_mos 116 7.6 0.001 0.001 88.855 88.924 qs_scf_loop_do_ot 116 8.6 0.001 0.001 88.854 88.923 ot_scf_mini 116 9.6 0.004 0.004 84.052 84.085 dbcsr_multiply_generic 2485 12.5 0.210 0.222 81.280 81.765 cp_fm_upper_to_full 104 14.8 52.527 75.166 52.527 75.166 multiply_cannon 2485 13.5 0.677 0.722 58.283 59.676 multiply_cannon_loop 2485 14.5 1.023 1.049 54.286 55.617 ot_mini 116 10.6 0.001 0.001 44.144 44.187 dbcsr_complete_redistribute 393 12.7 3.999 4.040 29.693 42.686 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.182 39.130 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.664 36.519 mp_alltoall_i22 712 14.1 21.417 34.285 21.417 34.285 rebuild_ks_matrix 127 8.3 0.001 0.001 33.510 33.541 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 33.509 33.540 cp_fm_cholesky_invert 11 10.9 33.307 33.315 33.307 33.315 mp_waitall_1 102768 16.8 27.682 31.829 27.682 31.829 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.065 31.109 qs_ot_get_p 127 10.4 0.009 0.014 25.243 25.318 qs_ot_get_derivative 116 11.6 0.002 0.002 24.382 24.416 qs_ot_p2m_diag 82 11.4 0.868 0.873 21.251 21.279 make_m2s 4970 13.5 0.074 0.076 18.500 19.974 multiply_cannon_metrocomm3 9940 15.5 0.025 0.027 18.474 19.759 ot_diis_step 116 11.6 0.022 0.022 19.729 19.729 make_images 4970 14.5 3.039 3.226 18.023 19.494 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.482 19.488 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.094 19.234 apply_single 127 13.6 0.001 0.001 19.094 19.234 multiply_cannon_multrec 9940 15.5 10.185 11.989 18.027 18.109 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.331 16.332 cp_fm_diag_elpa_base 82 14.4 11.960 13.540 16.328 16.328 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.268 16.280 calculate_rho_elec 127 8.7 0.477 0.478 16.268 16.279 sum_up_and_integrate 127 10.3 0.002 0.002 15.845 15.930 integrate_v_rspace 127 11.3 0.004 0.004 15.782 15.868 multiply_cannon_sync_h2d 9940 15.5 14.221 14.287 14.221 14.287 hybrid_alltoall_any 5155 16.4 1.303 3.065 10.716 12.830 make_images_data 4970 15.5 0.063 0.067 10.505 12.674 init_scf_run 11 5.9 0.000 0.001 12.198 12.199 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.198 12.199 pw_transfer 1535 11.6 0.094 0.095 11.195 11.209 fft_wrap_pw1pw2 1281 12.7 0.011 0.012 10.958 10.971 fft_wrap_pw1pw2_140 519 13.2 3.025 3.062 9.785 9.795 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.520 9.548 dbcsr_mm_accdrv_process 20590 16.0 4.181 5.996 7.596 9.478 wfi_extrapolate 11 7.9 0.001 0.001 8.934 8.934 cp_fm_cholesky_decompose 22 10.9 8.845 8.870 8.845 8.870 grid_integrate_task_list 127 12.3 8.498 8.698 8.498 8.698 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 8.196 8.201 density_rs2pw 127 9.7 0.005 0.006 7.373 7.420 mp_alltoall_d11v 2401 14.1 6.867 6.972 6.867 6.972 calculate_dm_sparse 127 9.5 0.001 0.001 6.818 6.868 grid_collocate_task_list 127 9.7 6.364 6.435 6.364 6.435 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.349 6.375 fft3d_ps 1281 14.7 2.832 2.848 6.241 6.302 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.162 6.229 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=287.488000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2854.727273, yerr=141.249676 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 1.259803E+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 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 122780. 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 0.021 0.049 85.000 85.002 qs_energies 1 2.0 0.000 0.000 84.372 84.378 ls_scf 1 3.0 0.000 0.000 83.433 83.439 dbcsr_multiply_generic 111 6.7 0.015 0.019 72.438 72.572 multiply_cannon 111 7.7 0.018 0.023 55.891 57.027 multiply_cannon_loop 111 8.7 0.225 0.240 52.471 53.602 ls_scf_main 1 4.0 0.000 0.000 52.142 52.142 density_matrix_trs4 2 5.0 0.002 0.003 46.618 46.700 ls_scf_init_scf 1 4.0 0.000 0.001 28.267 28.269 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.130 27.198 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 24.988 25.009 mp_waitall_1 11031 10.9 22.124 24.640 22.124 24.640 multiply_cannon_multrec 2664 9.7 8.174 8.947 15.565 17.332 multiply_cannon_sync_h2d 2664 9.7 13.761 15.887 13.761 15.887 make_m2s 222 7.7 0.009 0.012 12.989 13.492 make_images 222 8.7 0.098 0.108 12.966 13.472 multiply_cannon_metrocomm1 2664 9.7 0.010 0.010 9.505 11.784 make_images_data 222 9.7 0.004 0.005 7.590 8.198 dbcsr_mm_accdrv_process 4760 10.4 0.588 0.698 7.004 7.956 hybrid_alltoall_any 227 10.6 0.216 1.845 6.498 7.636 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.364 7.517 dbcsr_mm_accdrv_process_sort 4760 11.4 6.205 7.083 6.205 7.083 calculate_norms 4752 9.8 5.540 6.189 5.540 6.189 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.029 5.169 mp_sum_l 887 5.1 3.022 4.348 3.022 4.348 make_images_sizes 222 9.7 0.000 0.000 0.703 3.392 mp_alltoall_i44 222 10.7 0.703 3.391 0.703 3.391 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.056 3.300 mp_irecv_dv 6231 10.9 2.040 3.271 2.040 3.271 arnoldi_extremal 4 6.8 0.000 0.000 3.218 3.254 arnoldi_normal_ev 4 7.8 0.001 0.003 3.218 3.254 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.268 3.176 build_subspace 16 8.4 0.009 0.012 3.120 3.122 ls_scf_post 1 4.0 0.000 0.000 3.024 3.030 ls_scf_store_result 1 5.0 0.000 0.000 2.841 2.882 dbcsr_special_finalize 555 9.7 0.005 0.006 2.344 2.730 dbcsr_merge_single_wm 555 10.7 0.465 0.592 2.335 2.721 make_images_pack 222 9.7 2.205 2.621 2.207 2.622 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.335 2.561 dbcsr_sort_data 658 11.4 2.131 2.469 2.131 2.469 dbcsr_matrix_vector_mult_local 304 10.0 2.066 2.456 2.068 2.458 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.302 2.384 buffer_matrices_ensure_size 222 8.7 1.767 2.117 1.767 2.117 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.759 1.760 rebuild_ks_matrix 3 7.3 0.000 0.000 1.750 1.751 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.750 1.751 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.002000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 2.143199E+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 266673. MP_Allreduce 3138 10075. 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 0.022 0.040 90.557 90.558 qs_energies 1 2.0 0.000 0.000 90.113 90.118 ls_scf 1 3.0 0.000 0.000 88.630 88.663 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.505 74.820 multiply_cannon 111 7.7 0.028 0.042 52.782 57.744 ls_scf_main 1 4.0 0.000 0.000 54.932 54.938 multiply_cannon_loop 111 8.7 0.135 0.144 49.902 53.484 density_matrix_trs4 2 5.0 0.002 0.003 49.256 49.502 mp_waitall_1 9105 10.9 20.700 30.139 20.700 30.139 ls_scf_init_scf 1 4.0 0.000 0.000 30.042 30.074 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.859 28.986 multiply_cannon_multrec 1332 9.7 13.397 16.941 22.677 27.297 matrix_sqrt_Newton_Schulz 2 6.5 0.062 0.106 26.361 26.375 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.436 20.550 make_m2s 222 7.7 0.006 0.007 15.149 15.785 make_images 222 8.7 1.370 1.689 15.119 15.755 dbcsr_mm_accdrv_process 4041 10.4 0.370 0.543 8.872 10.435 dbcsr_mm_accdrv_process_sort 4041 11.4 8.382 9.892 8.382 9.892 make_images_data 222 9.7 0.004 0.004 8.767 9.710 hybrid_alltoall_any 227 10.6 0.542 2.551 8.094 9.544 mp_sum_l 887 5.1 5.099 8.368 5.099 8.368 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.194 7.674 mp_irecv_dv 3311 11.0 3.173 7.620 3.173 7.620 calculate_norms 2376 9.8 6.053 6.681 6.053 6.681 multiply_cannon_sync_h2d 1332 9.7 4.889 6.646 4.889 6.646 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.844 6.562 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.916 5.135 arnoldi_extremal 4 6.8 0.000 0.000 4.758 4.834 arnoldi_normal_ev 4 7.8 0.054 0.093 4.758 4.834 build_subspace 16 8.4 0.014 0.021 4.390 4.393 ls_scf_post 1 4.0 0.000 0.000 3.655 3.660 ls_scf_store_result 1 5.0 0.000 0.000 3.373 3.476 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.163 3.402 dbcsr_matrix_vector_mult_local 304 10.0 2.775 3.267 2.778 3.269 mp_allgather_i34 111 8.7 0.867 3.123 0.867 3.123 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.139 2.786 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.594 2.679 dbcsr_data_new 4174 10.1 2.113 2.396 2.113 2.396 make_images_pack 222 9.7 1.822 2.120 1.825 2.122 dbcsr_sort_data 436 11.2 1.849 2.092 1.849 2.092 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.849 1.851 rebuild_ks_matrix 3 7.3 0.000 0.000 1.836 1.838 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.836 1.838 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.558000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1774.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 2.901213E+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 265448. MP_Allreduce 3138 10896. 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 0.065 0.084 92.967 92.969 qs_energies 1 2.0 0.000 0.000 92.355 92.358 ls_scf 1 3.0 0.000 0.000 90.608 90.613 dbcsr_multiply_generic 111 6.7 0.017 0.030 74.860 75.217 ls_scf_main 1 4.0 0.000 0.000 56.188 56.193 multiply_cannon 111 7.7 0.035 0.093 51.884 55.484 multiply_cannon_loop 111 8.7 0.117 0.130 49.134 53.446 density_matrix_trs4 2 5.0 0.002 0.003 50.266 50.435 mp_waitall_1 7281 11.0 23.103 33.575 23.103 33.575 ls_scf_init_scf 1 4.0 0.000 0.000 30.703 30.706 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.506 29.594 matrix_sqrt_Newton_Schulz 2 6.5 0.073 0.164 26.923 26.932 multiply_cannon_multrec 888 9.7 12.684 15.348 21.247 24.661 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.884 23.441 make_m2s 222 7.7 0.006 0.007 16.141 16.910 make_images 222 8.7 1.580 1.857 16.103 16.870 make_images_data 222 9.7 0.004 0.005 9.328 10.481 hybrid_alltoall_any 227 10.6 0.641 2.948 8.969 10.309 dbcsr_mm_accdrv_process 3754 10.4 0.318 0.501 8.078 9.362 dbcsr_mm_accdrv_process_sort 3754 11.4 7.641 8.860 7.641 8.860 mp_sum_l 887 5.1 5.142 8.421 5.142 8.421 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.488 7.560 mp_irecv_dv 2335 11.1 2.473 7.494 2.473 7.494 multiply_cannon_sync_h2d 888 9.7 6.135 7.350 6.135 7.350 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.965 6.701 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.534 6.440 arnoldi_extremal 4 6.8 0.000 0.000 5.211 5.294 arnoldi_normal_ev 4 7.8 0.086 0.150 5.211 5.294 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.751 5.004 build_subspace 16 8.4 0.014 0.020 4.809 4.816 calculate_norms 1584 9.8 4.346 4.716 4.346 4.716 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.473 3.803 ls_scf_post 1 4.0 0.000 0.000 3.717 3.722 dbcsr_matrix_vector_mult_local 304 10.0 3.057 3.640 3.059 3.642 mp_allgather_i34 111 8.7 0.867 3.616 0.867 3.616 ls_scf_store_result 1 5.0 0.000 0.000 3.450 3.539 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.858 2.975 dbcsr_data_new 4116 9.9 2.108 2.451 2.108 2.451 dbcsr_sort_data 325 11.1 1.870 2.123 1.870 2.123 make_images_sizes 222 9.7 0.000 0.000 0.980 2.047 mp_alltoall_i44 222 10.7 0.980 2.047 0.980 2.047 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.886 1.888 make_images_pack 222 9.7 1.625 1.876 1.628 1.879 rebuild_ks_matrix 3 7.3 0.000 0.000 1.868 1.870 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.868 1.870 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.969000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2216.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 3.359568E+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 266673. MP_Allreduce 3138 13030. 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 0.164 0.226 97.111 97.112 qs_energies 1 2.0 0.000 0.000 96.116 96.122 ls_scf 1 3.0 0.000 0.000 94.456 94.462 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.230 78.480 ls_scf_main 1 4.0 0.000 0.000 58.615 58.616 multiply_cannon 111 7.7 0.049 0.121 51.389 56.471 density_matrix_trs4 2 5.0 0.002 0.003 52.526 52.647 multiply_cannon_loop 111 8.7 0.154 0.169 46.289 49.791 ls_scf_init_scf 1 4.0 0.000 0.000 32.633 32.635 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.434 31.503 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.924 28.938 mp_waitall_1 6369 11.0 22.242 28.900 22.242 28.900 multiply_cannon_multrec 1332 9.7 14.275 17.498 22.182 25.852 make_m2s 222 7.7 0.006 0.008 21.115 22.570 make_images 222 8.7 3.142 3.618 21.065 22.523 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 8.769 17.549 make_images_data 222 9.7 0.004 0.004 11.774 13.431 hybrid_alltoall_any 227 10.6 0.798 3.779 11.066 12.775 dbcsr_mm_accdrv_process 3641 10.4 0.297 0.491 7.546 9.044 dbcsr_mm_accdrv_process_sort 3641 11.4 7.098 8.542 7.098 8.542 mp_sum_l 887 5.1 4.146 7.276 4.146 7.276 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.095 5.962 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.531 5.946 mp_irecv_dv 3229 10.9 2.070 5.876 2.070 5.876 multiply_cannon_sync_h2d 1332 9.7 5.410 5.864 5.410 5.864 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.249 5.746 arnoldi_extremal 4 6.8 0.000 0.000 5.229 5.250 arnoldi_normal_ev 4 7.8 0.001 0.005 5.229 5.249 build_subspace 16 8.4 0.014 0.021 4.894 4.903 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.585 4.753 mp_allgather_i34 111 8.7 2.221 4.723 2.221 4.723 calculate_norms 2376 9.8 4.199 4.611 4.199 4.611 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.604 3.902 dbcsr_matrix_vector_mult_local 304 10.0 3.225 3.726 3.227 3.728 dbcsr_sort_data 658 11.4 3.118 3.331 3.118 3.331 dbcsr_special_finalize 555 9.7 0.006 0.007 2.870 3.215 ls_scf_post 1 4.0 0.000 0.000 3.208 3.213 dbcsr_merge_single_wm 555 10.7 0.538 0.661 2.861 3.207 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.994 3.048 ls_scf_store_result 1 5.0 0.000 0.000 2.957 3.013 dbcsr_data_release 10477 10.7 1.579 2.415 1.579 2.415 dbcsr_finalize 304 7.8 0.049 0.061 1.793 1.963 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.112000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2708.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 4.740817E+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 265536. MP_Allreduce 3129 15263. 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 0.036 0.050 92.828 92.829 qs_energies 1 2.0 0.000 0.000 92.096 92.101 ls_scf 1 3.0 0.000 0.000 90.151 90.155 dbcsr_multiply_generic 111 6.7 0.018 0.019 71.745 72.015 ls_scf_main 1 4.0 0.000 0.000 57.068 57.068 multiply_cannon 111 7.7 0.093 0.183 52.961 55.747 multiply_cannon_loop 111 8.7 0.088 0.094 50.368 52.323 density_matrix_trs4 2 5.0 0.002 0.003 50.104 50.214 ls_scf_init_scf 1 4.0 0.000 0.000 29.638 29.640 mp_waitall_1 5436 11.0 24.930 29.014 24.930 29.014 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.401 28.433 matrix_sqrt_Newton_Schulz 2 6.5 0.015 0.026 26.132 26.146 multiply_cannon_multrec 444 9.7 13.681 15.999 20.791 22.047 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.134 15.997 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.145 14.846 make_m2s 222 7.7 0.005 0.005 13.700 14.669 make_images 222 8.7 2.041 2.477 13.633 14.601 hybrid_alltoall_any 227 10.6 0.801 3.822 8.248 9.971 make_images_data 222 9.7 0.003 0.004 8.469 9.745 multiply_cannon_sync_h2d 444 9.7 6.735 8.291 6.735 8.291 dbcsr_mm_accdrv_process 3003 10.4 0.409 0.518 6.793 7.890 dbcsr_mm_accdrv_process_sort 3003 11.4 6.384 7.479 6.384 7.479 arnoldi_extremal 4 6.8 0.000 0.000 5.861 5.888 arnoldi_normal_ev 4 7.8 0.009 0.021 5.861 5.888 build_subspace 16 8.4 0.015 0.020 5.457 5.463 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.625 4.851 mp_sum_l 887 5.1 3.076 4.445 3.076 4.445 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.195 4.399 dbcsr_matrix_vector_mult_local 304 10.0 3.733 4.206 3.735 4.208 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.543 3.830 mp_irecv_dv 1241 11.2 1.529 3.787 1.529 3.787 mp_allgather_i34 111 8.7 1.174 3.784 1.174 3.784 calculate_norms 792 9.8 3.630 3.737 3.630 3.737 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.505 3.580 ls_scf_post 1 4.0 0.000 0.000 3.444 3.449 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.105 3.391 make_images_sizes 222 9.7 0.000 0.000 0.846 3.347 mp_alltoall_i44 222 10.7 0.846 3.347 0.846 3.347 ls_scf_store_result 1 5.0 0.000 0.000 3.238 3.270 dbcsr_finalize 304 7.8 0.062 0.076 2.198 2.321 dbcsr_data_new 4608 9.7 1.783 2.232 1.783 2.232 dbcsr_merge_all 275 8.9 0.478 0.533 2.057 2.159 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.013 2.014 rebuild_ks_matrix 3 7.3 0.000 0.000 1.979 1.980 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.979 1.980 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 1.930 1.930 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.829000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3778.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_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 8.854213E+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 284089. MP_Allreduce 3123 21388. 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 0.070 0.088 104.949 104.950 qs_energies 1 2.0 0.000 0.000 103.562 103.570 ls_scf 1 3.0 0.000 0.000 100.636 100.644 dbcsr_multiply_generic 111 6.7 0.024 0.028 74.395 74.545 ls_scf_main 1 4.0 0.000 0.000 63.431 63.431 density_matrix_trs4 2 5.0 0.002 0.003 54.523 54.587 multiply_cannon 111 7.7 0.129 0.199 48.111 50.131 multiply_cannon_loop 111 8.7 0.097 0.100 45.124 45.730 ls_scf_init_scf 1 4.0 0.000 0.000 33.413 33.413 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.921 31.942 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.232 29.240 mp_waitall_1 4527 11.1 21.645 25.250 21.645 25.250 make_m2s 222 7.7 0.005 0.005 22.707 23.785 make_images 222 8.7 3.577 3.870 22.600 23.676 multiply_cannon_multrec 444 9.7 17.813 18.505 22.439 23.138 hybrid_alltoall_any 227 10.6 1.653 3.616 12.823 15.826 make_images_data 222 9.7 0.003 0.004 13.095 15.208 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.961 10.388 multiply_cannon_sync_h2d 444 9.7 8.774 8.833 8.774 8.833 arnoldi_extremal 4 6.8 0.000 0.000 7.403 7.414 arnoldi_normal_ev 4 7.8 0.003 0.009 7.403 7.414 build_subspace 16 8.4 0.026 0.036 6.858 6.869 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.510 5.668 dbcsr_matrix_vector_mult_local 304 10.0 5.083 5.406 5.086 5.409 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.063 5.159 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.831 5.094 dbcsr_mm_accdrv_process 1814 10.4 0.268 0.358 4.447 4.567 dbcsr_mm_accdrv_process_sort 1814 11.4 4.109 4.237 4.109 4.237 ls_scf_post 1 4.0 0.000 0.000 3.793 3.802 make_images_sizes 222 9.7 0.000 0.000 1.462 3.641 mp_alltoall_i44 222 10.7 1.462 3.641 1.462 3.641 ls_scf_store_result 1 5.0 0.000 0.000 3.516 3.553 mp_allgather_i34 111 8.7 1.087 3.472 1.087 3.472 calculate_norms 792 9.8 3.248 3.285 3.248 3.285 dbcsr_finalize 304 7.8 0.082 0.089 3.076 3.158 dbcsr_merge_all 275 8.9 0.884 0.915 2.862 2.937 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.895 2.895 dbcsr_complete_redistribute 5 7.6 1.429 1.490 2.746 2.867 matrix_ls_to_qs 2 6.0 0.000 0.000 2.389 2.512 dbcsr_sort_data 325 11.1 2.438 2.494 2.438 2.494 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.382 2.384 dbcsr_data_new 6591 9.6 1.893 2.370 1.893 2.370 rebuild_ks_matrix 3 7.3 0.000 0.000 2.315 2.317 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.002 2.315 2.317 dbcsr_new_transposed 4 7.5 0.241 0.253 2.280 2.289 dbcsr_frobenius_norm 74 6.6 2.054 2.135 2.212 2.250 dbcsr_add_d 103 6.2 0.000 0.000 2.124 2.205 dbcsr_add_anytype 103 7.2 0.858 0.893 2.124 2.204 dbcsr_data_release 12724 10.6 1.972 2.163 1.972 2.163 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=104.950000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6990.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6b94574d37ca4648fbb0f33b2f35874362a23d3a_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 588.726272E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 1130802756. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.081 223.180 223.182 qs_mol_dyn_low 1 2.0 0.003 0.011 222.182 222.204 qs_forces 5 3.8 0.005 0.018 222.036 222.038 qs_energies 5 4.8 0.001 0.018 219.121 219.144 scf_env_do_scf 5 5.8 0.000 0.001 204.196 204.199 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 178.223 178.224 qs_scf_new_mos 105 7.6 0.000 0.001 139.186 139.372 qs_scf_loop_do_ot 105 8.6 0.001 0.001 139.186 139.372 dbcsr_multiply_generic 1445 12.2 0.125 0.133 132.492 132.856 ot_scf_mini 105 9.6 0.003 0.004 129.284 129.419 multiply_cannon 1445 13.2 0.275 0.285 114.128 116.032 multiply_cannon_loop 1445 14.2 2.829 3.004 112.468 113.523 velocity_verlet 4 3.0 0.003 0.009 104.539 104.540 ot_mini 105 10.6 0.001 0.002 59.082 59.200 multiply_cannon_multrec 69360 15.2 29.791 34.607 39.605 44.224 mp_waitall_1 488190 16.1 34.063 41.260 34.063 41.260 qs_ot_get_p 112 10.4 0.001 0.001 40.383 40.688 qs_ot_get_derivative 55 11.6 0.001 0.001 37.340 37.461 multiply_cannon_sync_h2d 69360 15.2 29.214 33.196 29.214 33.196 multiply_cannon_metrocomm3 69360 15.2 0.200 0.213 25.187 32.444 qs_ot_p2m_diag 40 11.0 0.020 0.030 29.573 29.637 rebuild_ks_matrix 110 8.4 0.000 0.000 28.437 28.614 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.016 28.436 28.613 cp_dbcsr_syevd 40 12.0 0.002 0.002 26.378 26.379 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.106 26.271 init_scf_loop 7 6.6 0.000 0.000 25.944 25.945 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.017 23.211 apply_single 62 13.6 0.000 0.000 23.017 23.211 cp_fm_syevd 40 13.0 0.000 0.000 21.303 21.446 prepare_preconditioner 7 7.6 0.000 0.000 21.188 21.226 make_preconditioner 7 8.6 0.000 0.000 21.188 21.226 ot_new_cg_direction 55 11.6 0.001 0.003 21.016 21.016 qs_rho_update_rho_low 110 7.6 0.001 0.001 16.240 16.621 calculate_rho_elec 110 8.6 0.030 0.032 16.239 16.620 cp_fm_redistribute_end 40 14.0 8.326 16.611 8.330 16.612 cp_fm_syevd_base 40 14.0 8.275 16.564 8.275 16.564 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.642 14.834 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.165 14.229 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.335 13.413 mp_sum_l 4764 12.2 11.627 12.409 11.627 12.409 init_scf_run 5 5.8 0.000 0.001 12.331 12.332 scf_env_initial_rho_setup 5 6.8 0.002 0.003 12.331 12.331 calculate_dm_sparse 110 9.5 0.000 0.001 11.482 11.662 pw_transfer 1645 12.4 0.079 0.102 11.394 11.628 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.255 11.496 density_rs2pw 110 9.6 0.006 0.007 10.608 10.964 dbcsr_mm_accdrv_process 154766 15.8 6.177 6.382 9.682 10.493 qs_vxc_create 110 10.4 0.002 0.004 10.282 10.331 fft_wrap_pw1pw2_240 915 15.0 1.143 1.252 9.879 10.121 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 9.971 10.076 cp_fm_cholesky_invert 7 10.6 9.960 9.967 9.960 9.967 check_diag 80 13.5 8.688 8.943 9.406 9.564 acc_transpose_blocks 69360 15.2 0.353 0.369 7.978 8.616 fft3d_pb 915 16.0 2.400 2.685 8.080 8.330 sum_up_and_integrate 60 10.3 0.001 0.002 8.197 8.208 integrate_v_rspace 60 11.3 0.001 0.012 8.180 8.191 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 7.930 7.947 transfer_rs2pw 445 10.6 0.007 0.008 7.385 7.753 multiply_cannon_metrocomm1 69360 15.2 0.097 0.105 4.563 7.659 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.526 7.581 calculate_first_density_matrix 1 7.0 0.000 0.001 7.511 7.529 xc_rho_set_and_dset_create 110 12.4 0.076 0.098 7.243 7.497 xc_vxc_pw_create 60 11.3 0.039 0.050 6.888 6.937 make_full_single_inverse 7 9.6 0.001 0.002 6.767 6.804 make_m2s 2890 13.2 0.079 0.089 6.250 6.801 make_images 2890 14.2 0.239 0.260 6.142 6.692 xc_pw_derive 510 13.4 0.005 0.007 6.057 6.152 acc_transpose_blocks_kernels 69360 16.2 0.858 0.906 5.212 5.763 mp_alltoall_z22v 2340 17.7 5.094 5.351 5.094 5.351 mp_waitany 7680 13.5 4.518 5.015 4.518 5.015 multiply_cannon_metrocomm4 67915 15.2 0.184 0.202 2.015 4.891 jit_kernel_transpose 5 15.0 4.353 4.887 4.353 4.887 wfi_extrapolate 5 7.8 0.000 0.002 4.496 4.496 potential_pw2rs 60 12.3 0.003 0.003 4.476 4.486 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=223.182000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=561.000000, yerr=2.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 6b94574d37ca4648fbb0f33b2f35874362a23d3a Summary: empty Status: OK