=== This is the CP2K Performance-Test === Already up to date. Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 9e23118c5f2bdf14bad76c6d1e1a6705ce46993c ################# 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/01 job id: 50063385 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/02 job id: 50063387 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/03 job id: 50063388 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/04 job id: 50063390 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/05 job id: 50063392 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/06 job id: 50063393 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/07 job id: 50063394 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/08 job id: 50063395 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/09 job id: 50063397 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/10 job id: 50063399 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/11 job id: 50063401 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/12 job id: 50063403 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/13 job id: 50063404 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/14 job id: 50063407 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/15 job id: 50063411 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/16 job id: 50063412 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/17 job id: 50063413 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/18 job id: 50063414 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/19 job id: 50063416 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/20 job id: 50063417 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/21 job id: 50063418 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/22 job id: 50063420 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/23 job id: 50063421 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/24 job id: 50063423 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/25 job id: 50063425 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/26 job id: 50063426 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/27 job id: 50063428 --- 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/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 0.018 0.032 134.010 134.011 farming_run 1 2.0 133.154 133.159 133.983 133.986 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.490960E+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 0.033 0.035 114.875 114.876 qs_energies 1 2.0 0.000 0.000 114.651 114.651 mp2_main 1 3.0 0.000 0.000 112.534 112.535 mp2_gpw_main 1 4.0 0.027 0.033 111.509 111.510 mp2_ri_gpw_compute_in 1 5.0 0.173 0.192 92.308 92.682 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.698 55.071 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.166 41.118 46.169 get_2c_integrals 1 6.0 0.008 0.009 36.869 37.439 integrate_v_rspace 273 8.0 0.439 0.455 24.702 29.577 pw_transfer 6555 10.6 0.377 0.403 26.887 27.338 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 25.512 25.994 grid_integrate_task_list 273 9.0 20.599 25.964 20.599 25.964 fft_wrap_pw1pw2_100 2178 12.4 0.029 0.031 23.064 23.546 compute_2c_integrals 1 7.0 0.002 0.002 19.364 19.367 rpa_ri_compute_en 1 5.0 0.019 0.020 19.096 19.344 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.637 19.076 mp2_eri_2c_integrate_gpw 1 9.0 2.378 2.426 18.635 19.074 cp_fm_cholesky_decompose 12 8.2 17.438 18.019 17.438 18.019 cholesky_decomp 1 7.0 0.000 0.000 16.349 16.926 fft3d_s 5443 13.4 16.161 16.574 16.182 16.594 ao_to_mo_and_store_B_mult_1 272 7.0 10.764 15.327 10.764 15.327 calculate_wavefunction 272 8.0 5.395 5.549 12.321 12.943 rpa_num_int 1 6.0 0.000 0.002 10.644 10.645 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.618 10.644 calc_mat_Q 8 8.0 0.000 0.000 9.485 9.581 contract_S_to_Q 8 9.0 0.000 0.000 8.908 9.006 calc_potential_gpw 544 9.5 0.004 0.005 8.231 8.714 parallel_gemm_fm 14 9.1 0.000 0.000 8.495 8.610 parallel_gemm_fm_cosma 14 10.1 8.495 8.610 8.495 8.610 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.096 8.405 potential_pw2rs 545 10.0 0.106 0.108 7.520 8.138 create_integ_mat 1 6.0 0.007 0.009 7.986 7.986 collocate_single_gaussian 272 10.0 0.039 0.042 7.358 7.609 array2fm 1 7.0 0.000 0.000 6.789 7.433 pw_scatter_s 2720 13.7 4.329 4.502 4.329 4.502 pw_gather_s 2722 13.2 3.459 3.710 3.459 3.710 array2fm_buffer_send 1 8.0 2.853 3.044 2.853 3.044 pw_poisson_solve 545 10.5 1.124 1.202 2.205 2.463 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.509867, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2809.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 0.024 0.038 421.080 421.081 farming_run 1 2.0 420.418 420.427 421.045 421.046 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.242100E+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 15799474. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 216.680 216.681 qs_energies 1 2.0 0.000 0.000 216.455 216.462 scf_env_do_scf 1 3.0 0.000 0.000 114.754 114.754 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.856 113.863 rebuild_ks_matrix 4 6.0 0.000 0.000 113.855 113.862 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 113.855 113.862 hfx_ks_matrix 4 8.0 0.001 0.001 113.507 113.512 integrate_four_center 4 9.0 0.153 0.463 113.506 113.511 integrate_four_center_main 4 10.0 0.145 0.616 101.774 105.309 integrate_four_center_bin 266 11.0 101.629 104.732 101.629 104.732 mp2_main 1 3.0 0.000 0.000 101.413 101.420 mp2_gpw_main 1 4.0 0.051 0.078 100.493 100.505 init_scf_loop 1 4.0 0.000 0.000 96.851 96.851 mp2_ri_gpw_compute_in 1 5.0 0.064 0.071 73.913 75.035 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.757 54.877 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.163 41.405 46.571 integrate_v_rspace 95 8.0 0.399 0.566 27.874 32.851 pw_transfer 2240 10.6 0.146 0.170 29.466 30.025 fft_wrap_pw1pw2 1868 11.4 0.017 0.020 28.489 29.041 grid_integrate_task_list 95 9.0 23.296 28.463 23.296 28.463 mp2_ri_gpw_compute_en 1 5.0 0.055 0.063 26.429 28.172 ao_to_mo_and_store_B_mult_1 91 7.0 10.682 27.595 10.682 27.595 fft_wrap_pw1pw2_100 730 12.4 0.013 0.014 26.253 26.766 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.837 1.891 24.661 24.670 get_2c_integrals 1 6.0 0.000 0.000 20.064 20.093 compute_2c_integrals 1 7.0 0.002 0.003 19.051 19.055 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.557 18.926 mp2_eri_2c_integrate_gpw 1 9.0 1.743 1.842 18.556 18.925 fft3d_s 1823 13.4 18.472 18.891 18.485 18.904 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.901 17.901 calculate_wavefunction 91 8.0 2.011 2.043 9.620 9.800 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.578 8.697 9.071 potential_pw2rs 186 10.0 0.033 0.035 8.405 8.969 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.111 8.519 local_gemm 172 8.0 8.139 8.513 8.139 8.513 mp2_ri_gpw_compute_en_comm 22 7.0 0.504 0.526 7.733 8.248 calc_potential_gpw 182 9.5 0.002 0.002 7.875 8.164 collocate_single_gaussian 91 10.0 0.017 0.021 7.724 7.980 integrate_four_center_load 4 10.0 0.000 0.000 6.749 6.752 hfx_load_balance 1 11.0 0.000 0.000 6.749 6.752 mp2_ri_gpw_compute_en_ener 172 7.0 6.349 6.412 6.349 6.412 mp_sendrecv_dm3 2068 8.0 5.752 6.275 5.752 6.275 mp_sync 37 10.5 3.725 6.129 3.725 6.129 pw_gather_s 912 13.2 4.489 5.019 4.489 5.019 pw_scatter_s 910 13.7 3.876 4.410 3.876 4.410 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.487254, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1501.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 453.451776E+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 257836. 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 0.010 0.032 54.233 54.235 qs_mol_dyn_low 1 2.0 0.003 0.003 53.995 54.004 qs_forces 11 3.9 0.040 0.050 53.869 53.870 qs_energies 11 4.9 0.001 0.002 52.336 52.371 scf_env_do_scf 11 5.9 0.001 0.001 46.381 46.381 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.340 44.340 qs_scf_new_mos 108 7.5 0.000 0.001 34.397 34.697 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.397 34.696 dbcsr_multiply_generic 2286 12.5 0.094 0.098 34.008 34.485 ot_scf_mini 108 9.5 0.002 0.002 32.662 32.864 multiply_cannon 2286 13.5 0.179 0.187 26.376 28.086 multiply_cannon_loop 2286 14.5 1.819 1.920 25.740 27.473 velocity_verlet 10 3.0 0.002 0.002 26.539 26.540 ot_mini 108 10.5 0.001 0.001 19.732 20.001 qs_ot_get_derivative 108 11.5 0.001 0.001 16.670 16.873 mp_waitall_1 245248 16.5 8.664 14.839 8.664 14.839 multiply_cannon_metrocomm3 54864 15.5 0.072 0.077 6.201 13.036 multiply_cannon_multrec 54864 15.5 3.635 5.677 7.603 10.973 qs_ot_get_p 119 10.4 0.001 0.001 8.173 8.466 rebuild_ks_matrix 119 8.3 0.000 0.000 7.788 7.953 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.788 7.953 mp_sum_l 7287 12.8 5.729 7.567 5.729 7.567 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.856 7.018 multiply_cannon_sync_h2d 54864 15.5 5.163 6.387 5.163 6.387 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.754 6.213 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.541 5.654 dbcsr_mm_accdrv_process 76910 16.1 1.840 2.883 3.882 5.546 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.264 5.298 init_scf_run 11 5.9 0.000 0.001 4.601 4.602 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.601 4.601 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.412 4.412 sum_up_and_integrate 119 10.3 0.001 0.001 4.382 4.388 integrate_v_rspace 119 11.3 0.002 0.002 4.371 4.379 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.221 4.222 cp_fm_redistribute_end 50 14.0 2.154 4.198 2.160 4.201 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.055 4.177 calculate_rho_elec 119 8.7 0.011 0.017 4.055 4.177 cp_fm_diag_elpa_base 50 14.0 2.036 4.098 2.039 4.103 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.950 3.173 apply_single 119 13.6 0.000 0.000 2.950 3.173 calculate_dm_sparse 119 9.5 0.000 0.000 2.971 3.135 multiply_cannon_metrocomm1 54864 15.5 0.056 0.060 1.842 2.912 ot_diis_step 108 11.5 0.006 0.006 2.772 2.772 acc_transpose_blocks 54864 15.5 0.220 0.243 2.196 2.706 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.556 2.615 jit_kernel_multiply 13 15.8 1.979 2.582 1.979 2.582 calculate_first_density_matrix 1 7.0 0.000 0.000 2.419 2.430 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.216 2.218 density_rs2pw 119 9.7 0.004 0.004 2.091 2.217 wfi_extrapolate 11 7.9 0.001 0.001 2.107 2.107 grid_integrate_task_list 119 12.3 2.007 2.101 2.007 2.101 mp_sum_d 4137 12.0 1.413 2.039 1.413 2.039 init_scf_loop 11 6.9 0.000 0.000 2.024 2.025 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.898 1.952 potential_pw2rs 119 12.3 0.004 0.004 1.778 1.788 pw_transfer 1439 11.6 0.051 0.058 1.624 1.693 make_m2s 4572 13.5 0.053 0.055 1.586 1.634 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.548 1.621 make_images 4572 14.5 0.133 0.139 1.504 1.553 transfer_rs2pw 487 10.6 0.005 0.006 1.349 1.506 mp_waitany 12084 13.8 1.265 1.467 1.265 1.467 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.403 1.435 acc_transpose_blocks_sync 164592 16.5 1.200 1.428 1.200 1.428 mp_alltoall_d11v 2130 13.8 1.270 1.427 1.270 1.427 grid_collocate_task_list 119 9.7 1.349 1.402 1.349 1.402 fft3d_ps 1201 14.6 0.368 0.477 1.262 1.329 transfer_pw2rs 487 13.2 0.006 0.006 1.303 1.310 fft_wrap_pw1pw2_140 487 13.2 0.107 0.114 1.203 1.274 dbcsr_dot_sd 1205 11.9 0.048 0.060 0.791 1.186 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.235000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.361984E+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 1349368. 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 0.015 0.043 39.236 39.238 qs_mol_dyn_low 1 2.0 0.003 0.003 38.953 38.960 qs_forces 11 3.9 0.020 0.055 38.895 38.896 qs_energies 11 4.9 0.002 0.007 37.195 37.216 scf_env_do_scf 11 5.9 0.001 0.004 31.897 31.898 scf_env_do_scf_inner_loop 108 6.5 0.012 0.082 29.438 29.439 dbcsr_multiply_generic 2286 12.5 0.104 0.115 21.970 22.330 qs_scf_new_mos 108 7.5 0.001 0.001 20.652 20.880 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.651 20.879 ot_scf_mini 108 9.5 0.002 0.003 19.729 19.896 velocity_verlet 10 3.0 0.001 0.002 18.380 18.381 multiply_cannon 2286 13.5 0.208 0.215 16.813 18.298 multiply_cannon_loop 2286 14.5 1.183 1.240 15.656 17.195 ot_mini 108 10.5 0.001 0.001 12.088 12.323 mp_waitall_1 200699 16.5 5.483 10.772 5.483 10.772 qs_ot_get_derivative 108 11.5 0.001 0.001 9.727 9.898 multiply_cannon_metrocomm3 27432 15.5 0.071 0.073 3.989 9.501 multiply_cannon_multrec 27432 15.5 1.820 4.098 6.122 8.895 rebuild_ks_matrix 119 8.3 0.000 0.000 6.726 6.876 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.017 6.725 6.876 dbcsr_mm_accdrv_process 47894 16.0 3.469 5.619 4.222 6.292 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.906 6.039 qs_ot_get_p 119 10.4 0.002 0.012 4.769 4.993 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.630 4.546 mp_sum_l 7287 12.8 2.107 4.093 2.107 4.093 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.064 4.060 apply_single 119 13.6 0.000 0.000 3.064 4.060 init_scf_run 11 5.9 0.000 0.001 4.046 4.047 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.046 4.047 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.638 3.677 calculate_rho_elec 119 8.7 0.021 0.024 3.638 3.676 sum_up_and_integrate 119 10.3 0.001 0.003 3.650 3.657 integrate_v_rspace 119 11.3 0.002 0.003 3.636 3.643 qs_ot_p2m_diag 50 11.0 0.012 0.029 3.030 3.051 make_m2s 4572 13.5 0.052 0.053 2.612 2.938 make_images 4572 14.5 0.206 0.247 2.522 2.851 cp_dbcsr_syevd 50 12.0 0.004 0.009 2.629 2.634 calculate_first_density_matrix 1 7.0 0.001 0.005 2.539 2.541 init_scf_loop 11 6.9 0.001 0.004 2.439 2.440 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.305 2.390 ot_diis_step 108 11.5 0.011 0.011 2.315 2.316 cp_fm_diag_elpa 50 13.0 0.000 0.001 2.265 2.266 multiply_cannon_sync_h2d 27432 15.5 1.713 2.251 1.713 2.251 cp_fm_redistribute_end 50 14.0 1.148 2.240 1.151 2.242 cp_fm_diag_elpa_base 50 14.0 1.059 2.151 1.087 2.189 calculate_dm_sparse 119 9.5 0.000 0.001 2.063 2.146 acc_transpose_blocks 27432 15.5 0.112 0.118 1.610 1.966 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.939 1.942 density_rs2pw 119 9.7 0.004 0.004 1.867 1.937 grid_integrate_task_list 119 12.3 1.841 1.920 1.841 1.920 jit_kernel_multiply 10 16.3 0.695 1.902 0.695 1.902 pw_transfer 1439 11.6 0.065 0.068 1.833 1.870 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.742 1.779 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.719 1.765 make_images_data 4572 15.5 0.047 0.053 1.244 1.609 prepare_preconditioner 11 7.9 0.000 0.000 1.503 1.529 make_preconditioner 11 8.9 0.000 0.002 1.503 1.529 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.403 1.460 wfi_extrapolate 11 7.9 0.001 0.001 1.447 1.447 fft_wrap_pw1pw2_140 487 13.2 0.128 0.132 1.391 1.428 hybrid_alltoall_any 4725 16.4 0.054 0.115 1.103 1.426 potential_pw2rs 119 12.3 0.006 0.006 1.403 1.410 fft3d_ps 1201 14.6 0.520 0.578 1.373 1.403 mp_alltoall_d11v 2130 13.8 1.252 1.365 1.252 1.365 grid_collocate_task_list 119 9.7 1.292 1.362 1.292 1.362 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.248 1.256 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.197 1.245 transfer_rs2pw 487 10.6 0.005 0.005 1.082 1.159 mp_sum_d 4137 12.0 0.559 1.027 0.559 1.027 qs_energies_init_hamiltonians 11 5.9 0.000 0.003 0.980 0.998 mp_allgather_i34 2286 14.5 0.580 0.977 0.580 0.977 acc_transpose_blocks_sync 82296 16.5 0.812 0.932 0.812 0.932 acc_transpose_blocks_kernels 27432 16.5 0.188 0.275 0.660 0.894 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.879 0.893 transfer_pw2rs 487 13.2 0.004 0.006 0.866 0.872 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.238000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.636364, yerr=0.771389 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 519.938048E+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 0.010 0.030 32.523 32.524 qs_mol_dyn_low 1 2.0 0.003 0.003 32.220 32.229 qs_forces 11 3.9 0.002 0.002 32.163 32.164 qs_energies 11 4.9 0.001 0.001 30.598 30.599 scf_env_do_scf 11 5.9 0.001 0.001 25.634 25.634 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 23.232 23.233 dbcsr_multiply_generic 2286 12.5 0.103 0.106 17.321 17.416 qs_scf_new_mos 108 7.5 0.001 0.001 15.522 15.538 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.522 15.538 velocity_verlet 10 3.0 0.001 0.001 15.248 15.249 multiply_cannon 2286 13.5 0.195 0.198 14.038 14.861 ot_scf_mini 108 9.5 0.002 0.002 14.765 14.778 multiply_cannon_loop 2286 14.5 0.855 0.881 13.267 14.094 ot_mini 108 10.5 0.001 0.001 9.165 9.176 qs_ot_get_derivative 108 11.5 0.001 0.001 7.679 7.692 multiply_cannon_multrec 18288 15.5 1.878 2.906 7.165 7.510 rebuild_ks_matrix 119 8.3 0.000 0.000 5.995 6.009 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.994 6.009 dbcsr_mm_accdrv_process 38222 16.0 5.136 5.917 5.193 5.981 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.270 5.283 init_scf_run 11 5.9 0.000 0.001 3.798 3.798 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.797 3.797 mp_waitall_1 158411 16.6 2.702 3.793 2.702 3.793 sum_up_and_integrate 119 10.3 0.001 0.001 3.570 3.576 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.990 3.568 integrate_v_rspace 119 11.3 0.002 0.003 3.557 3.565 qs_ot_get_p 119 10.4 0.001 0.001 3.353 3.378 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.282 3.286 calculate_rho_elec 119 8.7 0.031 0.031 3.282 3.285 calculate_first_density_matrix 1 7.0 0.000 0.000 2.572 2.572 multiply_cannon_metrocomm3 18288 15.5 0.048 0.049 1.497 2.465 init_scf_loop 11 6.9 0.000 0.000 2.385 2.386 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.985 2.361 apply_single 119 13.6 0.000 0.000 1.985 2.360 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.235 2.242 make_m2s 4572 13.5 0.044 0.045 1.844 1.990 density_rs2pw 119 9.7 0.004 0.004 1.849 1.970 calculate_dm_sparse 119 9.5 0.000 0.001 1.943 1.951 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.945 1.946 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.940 1.941 grid_integrate_task_list 119 12.3 1.799 1.919 1.799 1.919 make_images 4572 14.5 0.192 0.204 1.758 1.903 pw_transfer 1439 11.6 0.065 0.069 1.864 1.874 acc_transpose_blocks 18288 15.5 0.077 0.079 1.759 1.823 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.771 1.784 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.761 1.766 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.699 1.707 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.695 1.705 cp_fm_diag_elpa_base 50 14.0 1.672 1.685 1.694 1.704 prepare_preconditioner 11 7.9 0.000 0.000 1.634 1.636 make_preconditioner 11 8.9 0.000 0.000 1.634 1.636 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.497 1.579 ot_diis_step 108 11.5 0.011 0.012 1.471 1.471 fft_wrap_pw1pw2_140 487 13.2 0.178 0.183 1.420 1.431 mp_sum_l 7287 12.8 1.043 1.422 1.043 1.422 potential_pw2rs 119 12.3 0.007 0.008 1.365 1.369 fft3d_ps 1201 14.6 0.545 0.565 1.326 1.339 grid_collocate_task_list 119 9.7 1.234 1.299 1.234 1.299 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.283 1.286 wfi_extrapolate 11 7.9 0.001 0.001 1.172 1.172 multiply_cannon_sync_h2d 18288 15.5 1.048 1.158 1.048 1.158 transfer_rs2pw 487 10.6 0.005 0.005 0.983 1.102 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.943 0.958 acc_transpose_blocks_kernels 18288 16.5 0.218 0.226 0.932 0.955 make_images_data 4572 15.5 0.047 0.050 0.795 0.942 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.941 0.941 hybrid_alltoall_any 4725 16.4 0.058 0.117 0.688 0.864 transfer_pw2rs 487 13.2 0.004 0.004 0.835 0.839 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.797 0.799 acc_transpose_blocks_sync 54864 16.5 0.731 0.788 0.731 0.788 mp_alltoall_d11v 2130 13.8 0.671 0.777 0.671 0.777 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.674 0.742 jit_kernel_transpose 5 15.6 0.713 0.729 0.713 0.729 cp_fm_cholesky_invert 11 10.9 0.696 0.699 0.696 0.699 mp_alltoall_z22v 1201 16.6 0.628 0.697 0.628 0.697 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.524000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.454545, yerr=0.987525 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 558.346240E+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 0.017 0.034 36.873 36.874 qs_mol_dyn_low 1 2.0 0.003 0.004 36.568 36.576 qs_forces 11 3.9 0.002 0.002 36.329 36.330 qs_energies 11 4.9 0.008 0.014 34.623 34.629 scf_env_do_scf 11 5.9 0.001 0.002 29.377 29.378 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.976 25.978 dbcsr_multiply_generic 2286 12.5 0.103 0.107 20.186 20.345 velocity_verlet 10 3.0 0.001 0.002 18.526 18.535 qs_scf_new_mos 108 7.5 0.001 0.001 17.944 17.998 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.944 17.997 ot_scf_mini 108 9.5 0.002 0.003 16.917 16.975 multiply_cannon 2286 13.5 0.219 0.226 16.527 16.960 multiply_cannon_loop 2286 14.5 1.550 1.647 15.584 15.977 ot_mini 108 10.5 0.001 0.001 10.523 10.587 multiply_cannon_multrec 27432 15.5 2.463 3.121 9.225 9.496 qs_ot_get_derivative 108 11.5 0.001 0.001 8.664 8.722 dbcsr_mm_accdrv_process 47916 15.9 6.193 7.930 6.660 7.995 rebuild_ks_matrix 119 8.3 0.000 0.000 6.308 6.364 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.308 6.364 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.595 5.645 init_scf_run 11 5.9 0.000 0.001 3.881 3.882 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.881 3.881 qs_ot_get_p 119 10.4 0.001 0.003 3.568 3.647 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.152 3.544 sum_up_and_integrate 119 10.3 0.001 0.002 3.470 3.479 integrate_v_rspace 119 11.3 0.003 0.005 3.458 3.468 init_scf_loop 11 6.9 0.002 0.013 3.381 3.385 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.340 3.373 calculate_rho_elec 119 8.7 0.040 0.046 3.339 3.373 acc_transpose_blocks 27432 15.5 0.118 0.121 2.472 2.633 prepare_preconditioner 11 7.9 0.000 0.000 2.524 2.532 make_preconditioner 11 8.9 0.000 0.002 2.524 2.532 calculate_first_density_matrix 1 7.0 0.001 0.005 2.504 2.506 make_full_inverse_cholesky 11 9.9 0.000 0.001 2.131 2.463 mp_waitall_1 137007 16.6 1.881 2.421 1.881 2.421 make_m2s 4572 13.5 0.054 0.057 2.270 2.384 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 2.325 2.352 calculate_dm_sparse 119 9.5 0.000 0.001 2.237 2.289 make_images 4572 14.5 0.274 0.335 2.161 2.274 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.089 2.213 apply_single 119 13.6 0.000 0.000 2.089 2.212 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.165 2.173 pw_transfer 1439 11.6 0.065 0.069 1.987 2.020 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.971 1.972 density_rs2pw 119 9.7 0.004 0.004 1.802 1.942 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.895 1.931 grid_integrate_task_list 119 12.3 1.825 1.914 1.825 1.914 ot_diis_step 108 11.5 0.012 0.013 1.818 1.818 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 0.004 1.761 1.761 fft_wrap_pw1pw2_140 487 13.2 0.227 0.232 1.575 1.611 acc_transpose_blocks_sync 82296 16.5 1.460 1.590 1.460 1.590 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.532 1.540 cp_fm_diag_elpa_base 50 14.0 1.497 1.513 1.529 1.538 jit_kernel_multiply 7 16.1 0.401 1.415 0.401 1.415 fft3d_ps 1201 14.6 0.592 0.653 1.384 1.411 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.870 1.409 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.336 1.345 grid_collocate_task_list 119 9.7 1.250 1.327 1.250 1.327 wfi_extrapolate 11 7.9 0.001 0.001 1.325 1.325 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.277 1.296 potential_pw2rs 119 12.3 0.009 0.009 1.277 1.281 mp_sum_l 7287 12.8 0.976 1.251 0.976 1.251 cp_fm_upper_to_full 72 14.2 0.829 1.179 0.829 1.179 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.092 1.093 transfer_rs2pw 487 10.6 0.004 0.005 0.896 1.040 dbcsr_complete_redistribute 329 12.2 0.118 0.148 0.760 1.034 make_images_data 4572 15.5 0.048 0.052 0.850 0.969 hybrid_alltoall_any 4725 16.4 0.066 0.155 0.716 0.918 acc_transpose_blocks_kernels 27432 16.5 0.272 0.279 0.865 0.901 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.879 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.563 0.829 mp_alltoall_d11v 2130 13.8 0.716 0.825 0.716 0.825 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.810 0.816 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.874000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=531.181818, yerr=3.009626 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 630.771712E+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 0.011 0.031 28.803 28.804 qs_mol_dyn_low 1 2.0 0.003 0.003 28.611 28.620 qs_forces 11 3.9 0.002 0.002 28.549 28.550 qs_energies 11 4.9 0.003 0.003 26.838 26.841 scf_env_do_scf 11 5.9 0.001 0.001 21.981 21.981 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.518 19.519 velocity_verlet 10 3.0 0.002 0.002 14.749 14.752 dbcsr_multiply_generic 2286 12.5 0.095 0.098 12.969 13.054 qs_scf_new_mos 108 7.5 0.001 0.001 11.804 11.823 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.803 11.823 ot_scf_mini 108 9.5 0.002 0.002 11.097 11.116 multiply_cannon 2286 13.5 0.225 0.236 10.401 10.866 multiply_cannon_loop 2286 14.5 0.644 0.667 9.502 9.673 ot_mini 108 10.5 0.001 0.001 6.354 6.380 multiply_cannon_multrec 9144 15.5 1.728 1.914 5.971 6.177 rebuild_ks_matrix 119 8.3 0.000 0.000 5.802 5.818 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.802 5.817 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.159 5.173 qs_ot_get_derivative 108 11.5 0.001 0.001 5.067 5.087 dbcsr_mm_accdrv_process 12550 15.8 3.473 4.162 4.134 4.214 sum_up_and_integrate 119 10.3 0.001 0.001 3.457 3.462 integrate_v_rspace 119 11.3 0.003 0.003 3.447 3.452 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.400 3.406 calculate_rho_elec 119 8.7 0.060 0.061 3.400 3.406 init_scf_run 11 5.9 0.000 0.001 3.375 3.375 scf_env_initial_rho_setup 11 6.9 0.002 0.002 3.375 3.375 qs_ot_get_p 119 10.4 0.001 0.001 2.867 2.906 init_scf_loop 11 6.9 0.000 0.000 2.442 2.443 calculate_first_density_matrix 1 7.0 0.000 0.000 2.233 2.234 pw_transfer 1439 11.6 0.065 0.068 2.126 2.135 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.032 2.042 grid_integrate_task_list 119 12.3 1.867 1.938 1.867 1.938 make_m2s 4572 13.5 0.034 0.034 1.783 1.917 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.870 1.873 density_rs2pw 119 9.7 0.003 0.003 1.776 1.870 mp_waitall_1 115863 16.7 1.349 1.853 1.349 1.853 make_images 4572 14.5 0.268 0.299 1.693 1.827 calculate_dm_sparse 119 9.5 0.000 0.000 1.736 1.754 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.726 1.745 prepare_preconditioner 11 7.9 0.000 0.000 1.710 1.713 make_preconditioner 11 8.9 0.000 0.000 1.710 1.713 fft_wrap_pw1pw2_140 487 13.2 0.325 0.333 1.683 1.695 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.628 1.628 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.600 1.626 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.534 1.535 acc_transpose_blocks 9144 15.5 0.042 0.044 1.430 1.464 fft3d_ps 1201 14.6 0.649 0.662 1.384 1.392 grid_collocate_task_list 119 9.7 1.298 1.384 1.298 1.384 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.368 1.375 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.349 1.361 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.336 1.344 cp_fm_diag_elpa_base 50 14.0 1.310 1.327 1.335 1.343 ot_diis_step 108 11.5 0.013 0.013 1.275 1.275 potential_pw2rs 119 12.3 0.010 0.010 1.263 1.267 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.259 1.260 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.220 1.252 apply_single 119 13.6 0.000 0.000 1.220 1.252 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.128 1.133 wfi_extrapolate 11 7.9 0.001 0.001 1.091 1.091 jit_kernel_multiply 6 15.4 0.622 1.020 0.622 1.020 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.771 0.970 mp_alltoall_d11v 2130 13.8 0.815 0.952 0.815 0.952 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.874 0.925 make_images_data 4572 15.5 0.042 0.046 0.754 0.921 transfer_rs2pw 487 10.6 0.004 0.004 0.804 0.905 cp_fm_cholesky_invert 11 10.9 0.864 0.867 0.864 0.867 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.820 0.827 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.763 0.765 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.707 0.756 acc_transpose_blocks_sync 27432 16.5 0.724 0.754 0.724 0.754 transfer_pw2rs 487 13.2 0.003 0.004 0.658 0.660 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.648 0.653 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.355 0.645 mp_alltoall_z22v 1201 16.6 0.608 0.642 0.608 0.642 mp_allgather_i34 2286 14.5 0.214 0.617 0.214 0.617 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.804000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=592.363636, yerr=11.726216 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 796.028928E+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 0.013 0.037 43.407 43.409 qs_mol_dyn_low 1 2.0 0.003 0.003 43.169 43.178 qs_forces 11 3.9 0.002 0.002 43.049 43.050 qs_energies 11 4.9 0.001 0.001 41.009 41.012 scf_env_do_scf 11 5.9 0.001 0.001 35.210 35.210 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.302 27.304 velocity_verlet 10 3.0 0.002 0.002 24.520 24.533 dbcsr_multiply_generic 2286 12.5 0.104 0.107 19.215 19.457 qs_scf_new_mos 108 7.5 0.001 0.001 17.546 17.645 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.545 17.644 ot_scf_mini 108 9.5 0.002 0.002 16.380 16.484 multiply_cannon 2286 13.5 0.299 0.307 15.171 16.155 multiply_cannon_loop 2286 14.5 0.864 0.879 13.870 14.858 ot_mini 108 10.5 0.001 0.001 9.906 10.024 multiply_cannon_multrec 9144 15.5 3.371 4.620 8.867 8.979 qs_ot_get_derivative 108 11.5 0.001 0.001 7.816 7.920 init_scf_loop 11 6.9 0.000 0.000 7.881 7.885 rebuild_ks_matrix 119 8.3 0.000 0.000 7.332 7.474 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.331 7.474 prepare_preconditioner 11 7.9 0.000 0.000 6.890 6.905 make_preconditioner 11 8.9 0.000 0.000 6.890 6.905 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.471 6.774 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.599 6.728 dbcsr_mm_accdrv_process 12550 15.8 4.765 6.608 5.360 6.667 cp_fm_upper_to_full 72 14.2 3.166 4.521 3.166 4.521 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.231 4.243 calculate_rho_elec 119 8.7 0.117 0.120 4.230 4.242 sum_up_and_integrate 119 10.3 0.001 0.001 3.893 3.900 integrate_v_rspace 119 11.3 0.003 0.003 3.883 3.890 init_scf_run 11 5.9 0.000 0.001 3.746 3.746 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.746 3.746 qs_ot_get_p 119 10.4 0.001 0.001 3.512 3.649 mp_waitall_1 94719 16.7 2.376 3.437 2.376 3.437 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.879 3.285 pw_transfer 1439 11.6 0.069 0.069 3.048 3.052 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.949 2.954 dbcsr_complete_redistribute 329 12.2 0.282 0.287 1.981 2.803 make_m2s 4572 13.5 0.037 0.037 2.448 2.626 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.229 2.537 apply_single 119 13.6 0.000 0.000 2.229 2.537 fft_wrap_pw1pw2_140 487 13.2 0.648 0.650 2.510 2.516 make_images 4572 14.5 0.352 0.386 2.327 2.506 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.652 2.465 calculate_dm_sparse 119 9.5 0.000 0.000 2.386 2.404 density_rs2pw 119 9.7 0.003 0.004 2.254 2.282 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.334 2.260 calculate_first_density_matrix 1 7.0 0.000 0.000 2.237 2.241 mp_alltoall_i22 627 13.8 1.388 2.236 1.388 2.236 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.414 2.221 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.090 2.144 grid_integrate_task_list 119 12.3 2.084 2.113 2.084 2.113 qs_ot_p2m_diag 50 11.0 0.044 0.044 2.063 2.065 ot_diis_step 108 11.5 0.014 0.015 2.061 2.061 fft3d_ps 1201 14.6 0.867 0.885 1.871 1.875 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.846 1.847 mp_sum_l 7287 12.8 1.066 1.843 1.066 1.843 acc_transpose_blocks 9144 15.5 0.044 0.046 1.823 1.840 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.774 1.775 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.702 1.704 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.583 1.625 grid_collocate_task_list 119 9.7 1.525 1.553 1.525 1.553 cp_fm_cholesky_invert 11 10.9 1.482 1.485 1.482 1.485 wfi_extrapolate 11 7.9 0.001 0.001 1.441 1.441 potential_pw2rs 119 12.3 0.014 0.014 1.403 1.405 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.403 1.403 cp_fm_diag_elpa_base 50 14.0 1.257 1.308 1.401 1.402 hybrid_alltoall_any 4725 16.4 0.091 0.152 1.114 1.389 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.381 1.388 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.319 1.340 make_images_data 4572 15.5 0.046 0.049 1.067 1.296 mp_alltoall_d11v 2130 13.8 1.185 1.216 1.185 1.216 acc_transpose_blocks_sync 27432 16.5 1.104 1.121 1.104 1.121 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.104 1.119 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.994 1.036 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.942 0.956 qs_create_task_list 11 7.9 0.000 0.000 0.926 0.937 generate_qs_task_list 11 8.9 0.367 0.386 0.926 0.937 mp_alltoall_z22v 1201 16.6 0.867 0.897 0.867 0.897 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.409000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=744.545455, yerr=11.857975 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.391360E+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 1383689. 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 0.015 0.042 81.502 81.502 qs_mol_dyn_low 1 2.0 0.004 0.006 81.178 81.221 qs_forces 11 3.9 0.003 0.006 81.097 81.098 qs_energies 11 4.9 0.001 0.002 78.254 78.267 scf_env_do_scf 11 5.9 0.000 0.001 69.297 69.299 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 63.985 63.985 dbcsr_multiply_generic 2055 12.4 0.107 0.109 50.891 51.289 qs_scf_new_mos 99 7.5 0.000 0.001 47.038 47.220 qs_scf_loop_do_ot 99 8.5 0.000 0.001 47.038 47.220 ot_scf_mini 99 9.5 0.007 0.030 44.633 44.789 multiply_cannon 2055 13.4 0.176 0.182 42.540 43.363 velocity_verlet 10 3.0 0.001 0.002 42.822 42.823 multiply_cannon_loop 2055 14.4 1.777 1.812 41.610 42.432 ot_mini 99 10.5 0.001 0.001 26.622 26.739 qs_ot_get_derivative 99 11.5 0.001 0.001 19.777 19.912 multiply_cannon_multrec 49320 15.4 11.399 12.120 17.383 18.014 rebuild_ks_matrix 110 8.3 0.000 0.000 14.377 14.547 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.015 14.376 14.547 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.604 12.752 mp_waitall_1 220248 16.4 10.796 11.609 10.796 11.609 multiply_cannon_sync_h2d 49320 15.4 9.615 10.343 9.615 10.343 qs_ot_get_p 110 10.4 0.001 0.001 9.543 9.700 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.528 8.084 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 7.219 7.842 apply_single 110 13.6 0.000 0.000 7.219 7.842 multiply_cannon_metrocomm3 49320 15.4 0.081 0.084 6.602 7.509 sum_up_and_integrate 110 10.3 0.002 0.004 6.907 6.920 init_scf_run 11 5.9 0.000 0.001 6.913 6.913 scf_env_initial_rho_setup 11 6.9 0.001 0.003 6.912 6.913 integrate_v_rspace 110 11.3 0.003 0.003 6.881 6.901 ot_diis_step 99 11.5 0.006 0.008 6.598 6.598 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.401 6.425 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.219 6.375 calculate_rho_elec 110 8.6 0.021 0.025 6.218 6.374 dbcsr_mm_accdrv_process 87628 16.1 3.045 3.139 5.854 6.126 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.580 5.580 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.386 5.451 init_scf_loop 11 6.9 0.000 0.000 5.286 5.286 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.025 5.047 cp_fm_diag_elpa_base 48 14.0 5.013 5.035 5.024 5.046 mp_sum_l 6594 12.7 4.029 4.707 4.029 4.707 wfi_extrapolate 11 7.9 0.001 0.003 4.005 4.005 make_m2s 4110 13.4 0.060 0.066 3.861 3.963 calculate_dm_sparse 110 9.5 0.001 0.001 3.767 3.905 make_images 4110 14.4 0.178 0.191 3.766 3.873 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.517 3.522 multiply_cannon_metrocomm1 49320 15.4 0.064 0.066 2.391 3.449 density_rs2pw 110 9.6 0.004 0.004 3.254 3.440 grid_integrate_task_list 110 12.3 3.253 3.418 3.253 3.418 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.172 3.251 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.182 3.227 prepare_preconditioner 11 7.9 0.000 0.000 3.138 3.172 make_preconditioner 11 8.9 0.000 0.000 3.138 3.172 pw_transfer 1331 11.6 0.054 0.068 2.969 3.046 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.936 2.991 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.881 2.962 calculate_first_density_matrix 1 7.0 0.000 0.001 2.811 2.815 acc_transpose_blocks 49320 15.4 0.206 0.214 2.532 2.597 fft_wrap_pw1pw2_140 451 13.1 0.309 0.329 2.480 2.565 mp_alltoall_d11v 2046 13.8 2.024 2.546 2.024 2.546 jit_kernel_multiply 13 15.9 2.532 2.543 2.532 2.543 potential_pw2rs 110 12.3 0.005 0.006 2.520 2.535 grid_collocate_task_list 110 9.6 2.157 2.271 2.157 2.271 fft3d_ps 1111 14.6 0.796 0.887 2.175 2.238 transfer_rs2pw 451 10.6 0.005 0.006 1.860 2.024 mp_waitany 14300 13.8 1.810 1.959 1.810 1.959 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.902 1.934 make_images_data 4110 15.4 0.043 0.048 1.754 1.888 mp_sum_d 3891 11.9 1.383 1.871 1.383 1.871 hybrid_alltoall_any 4261 16.3 0.084 0.484 1.527 1.787 cp_fm_cholesky_invert 11 10.9 1.729 1.733 1.729 1.733 transfer_pw2rs 451 13.1 0.006 0.007 1.716 1.726 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.640 1.667 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.502000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.818182, yerr=2.442733 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 592.994304E+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 2022902. 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 0.064 0.235 68.868 68.870 qs_mol_dyn_low 1 2.0 0.004 0.007 68.253 68.262 qs_forces 11 3.9 0.003 0.003 68.183 68.184 qs_energies 11 4.9 0.001 0.003 64.874 64.878 scf_env_do_scf 11 5.9 0.001 0.002 56.147 56.151 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.688 48.689 dbcsr_multiply_generic 2055 12.4 0.115 0.119 38.098 38.250 velocity_verlet 10 3.0 0.001 0.002 35.748 35.749 qs_scf_new_mos 99 7.5 0.001 0.001 32.949 33.110 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.948 33.109 multiply_cannon 2055 13.4 0.225 0.243 31.523 32.628 ot_scf_mini 99 9.5 0.003 0.003 31.289 31.425 multiply_cannon_loop 2055 14.4 1.157 1.189 30.256 31.256 ot_mini 99 10.5 0.001 0.001 18.285 18.433 multiply_cannon_multrec 24660 15.4 6.997 8.888 14.194 15.835 rebuild_ks_matrix 110 8.3 0.000 0.000 13.487 13.630 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.027 13.486 13.629 qs_ot_get_derivative 99 11.5 0.001 0.001 12.525 12.665 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.867 12.002 mp_waitall_1 176588 16.5 7.521 10.132 7.521 10.132 multiply_cannon_metrocomm3 24660 15.4 0.073 0.075 5.135 7.726 multiply_cannon_sync_h2d 24660 15.4 6.431 7.450 6.431 7.450 init_scf_loop 11 6.9 0.000 0.000 7.424 7.425 dbcsr_mm_accdrv_process 52282 16.1 5.698 6.483 7.027 7.413 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.467 7.108 apply_single 110 13.6 0.000 0.001 6.466 7.108 qs_ot_get_p 110 10.4 0.001 0.002 6.352 6.501 init_scf_run 11 5.9 0.000 0.001 6.350 6.350 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.350 6.350 sum_up_and_integrate 110 10.3 0.002 0.003 6.305 6.318 integrate_v_rspace 110 11.3 0.002 0.003 6.277 6.291 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.812 5.820 calculate_rho_elec 110 8.6 0.039 0.047 5.812 5.820 ot_diis_step 99 11.5 0.010 0.010 5.711 5.711 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.701 5.437 prepare_preconditioner 11 7.9 0.000 0.000 5.408 5.431 make_preconditioner 11 8.9 0.000 0.000 5.407 5.431 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.989 5.142 make_m2s 4110 13.4 0.057 0.059 4.182 4.676 make_images 4110 14.4 0.407 0.471 4.070 4.560 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.460 4.479 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.000 4.000 pw_transfer 1331 11.6 0.066 0.071 3.516 3.656 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.410 3.552 wfi_extrapolate 11 7.9 0.001 0.001 3.500 3.500 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.442 3.456 cp_fm_diag_elpa_base 48 14.0 3.396 3.419 3.440 3.453 grid_integrate_task_list 110 12.3 3.164 3.376 3.164 3.376 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.272 3.344 density_rs2pw 110 9.6 0.004 0.004 3.070 3.291 calculate_dm_sparse 110 9.5 0.001 0.001 3.215 3.243 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.236 3.238 fft_wrap_pw1pw2_140 451 13.1 0.357 0.380 2.933 3.072 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.898 2.947 calculate_first_density_matrix 1 7.0 0.000 0.001 2.755 2.757 make_images_data 4110 15.4 0.048 0.053 2.284 2.756 hybrid_alltoall_any 4261 16.3 0.105 0.448 1.971 2.720 fft3d_ps 1111 14.6 1.117 1.349 2.502 2.656 cp_fm_cholesky_invert 11 10.9 2.510 2.517 2.510 2.517 mp_sum_l 6594 12.7 1.774 2.437 1.774 2.437 grid_collocate_task_list 110 9.6 2.169 2.303 2.169 2.303 potential_pw2rs 110 12.3 0.008 0.009 2.190 2.198 acc_transpose_blocks 24660 15.4 0.114 0.118 1.960 1.995 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.951 1.975 mp_alltoall_d11v 2046 13.8 1.704 1.880 1.704 1.880 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.868 1.869 jit_kernel_multiply 9 16.3 0.964 1.862 0.964 1.862 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.790 1.801 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.774 1.785 transfer_rs2pw 451 10.6 0.006 0.007 1.396 1.688 mp_allgather_i34 2055 14.4 0.575 1.634 0.575 1.634 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.756 1.494 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.388 1.490 mp_waitany 10164 13.8 1.149 1.459 1.149 1.459 mp_irecv_dv 57340 16.2 0.632 1.379 0.632 1.379 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.870000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.000000, yerr=7.083913 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 668.323840E+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 0.075 0.103 60.056 60.058 qs_mol_dyn_low 1 2.0 0.003 0.003 59.272 59.321 qs_forces 11 3.9 0.052 0.058 59.178 59.179 qs_energies 11 4.9 0.003 0.008 55.969 56.008 scf_env_do_scf 11 5.9 0.001 0.001 48.108 48.109 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 39.728 39.728 velocity_verlet 10 3.0 0.001 0.002 32.069 32.072 dbcsr_multiply_generic 2055 12.4 0.109 0.115 28.865 29.143 qs_scf_new_mos 99 7.5 0.001 0.001 25.264 25.370 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.264 25.370 ot_scf_mini 99 9.5 0.003 0.003 24.024 24.149 multiply_cannon 2055 13.4 0.212 0.222 22.568 23.778 multiply_cannon_loop 2055 14.4 0.809 0.835 21.542 22.793 ot_mini 99 10.5 0.001 0.001 13.880 14.007 rebuild_ks_matrix 110 8.3 0.000 0.000 12.024 12.180 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.016 12.023 12.180 multiply_cannon_multrec 16440 15.4 3.677 4.764 9.886 11.073 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.576 10.722 mp_waitall_1 139946 16.5 6.976 9.970 6.976 9.970 qs_ot_get_derivative 99 11.5 0.001 0.001 9.464 9.590 init_scf_loop 11 6.9 0.001 0.004 8.347 8.348 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.469 7.487 prepare_preconditioner 11 7.9 0.000 0.000 6.603 6.619 make_preconditioner 11 8.9 0.000 0.002 6.603 6.619 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.949 6.301 sum_up_and_integrate 110 10.3 0.002 0.003 6.203 6.218 integrate_v_rspace 110 11.3 0.003 0.003 6.177 6.192 dbcsr_mm_accdrv_process 34862 16.1 5.366 5.741 6.053 6.173 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.674 5.685 calculate_rho_elec 110 8.6 0.059 0.059 5.673 5.685 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.930 5.511 apply_single 110 13.6 0.000 0.000 4.929 5.511 qs_ot_get_p 110 10.4 0.001 0.002 5.249 5.459 init_scf_run 11 5.9 0.000 0.001 5.448 5.448 scf_env_initial_rho_setup 11 6.9 0.002 0.002 5.448 5.448 ot_diis_step 99 11.5 0.011 0.011 4.386 4.386 make_m2s 4110 13.4 0.050 0.051 4.086 4.334 make_images 4110 14.4 0.400 0.519 3.971 4.220 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.264 3.929 multiply_cannon_sync_h2d 16440 15.4 3.261 3.893 3.261 3.893 qs_ot_p2m_diag 48 11.0 0.049 0.167 3.642 3.756 pw_transfer 1331 11.6 0.065 0.072 3.558 3.569 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.451 3.465 grid_integrate_task_list 110 12.3 3.181 3.355 3.181 3.355 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.263 3.263 density_rs2pw 110 9.6 0.004 0.004 2.919 3.105 fft_wrap_pw1pw2_140 451 13.1 0.460 0.469 3.005 3.016 wfi_extrapolate 11 7.9 0.001 0.001 2.911 2.911 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.769 2.770 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.728 2.737 cp_fm_diag_elpa_base 48 14.0 2.664 2.691 2.727 2.735 make_images_data 4110 15.4 0.046 0.050 2.281 2.608 hybrid_alltoall_any 4261 16.3 0.108 0.380 2.053 2.577 cp_fm_cholesky_invert 11 10.9 2.545 2.551 2.545 2.551 calculate_dm_sparse 110 9.5 0.001 0.001 2.512 2.539 mp_sum_l 6594 12.7 1.679 2.499 1.679 2.499 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.422 2.492 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.434 2.484 multiply_cannon_metrocomm4 14385 15.4 0.046 0.050 0.887 2.479 calculate_first_density_matrix 1 7.0 0.004 0.018 2.444 2.446 fft3d_ps 1111 14.6 1.117 1.130 2.369 2.382 grid_collocate_task_list 110 9.6 2.222 2.358 2.222 2.358 mp_irecv_dv 48980 15.7 0.813 2.344 0.813 2.344 potential_pw2rs 110 12.3 0.011 0.011 2.071 2.075 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.977 2.014 mp_alltoall_d11v 2046 13.8 1.753 1.994 1.753 1.994 dbcsr_complete_redistribute 325 12.2 0.330 0.392 1.422 1.869 cp_fm_upper_to_full 70 14.2 1.395 1.755 1.395 1.755 acc_transpose_blocks 16440 15.4 0.077 0.079 1.597 1.739 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.640 1.656 cp_fm_cholesky_decompose 22 10.9 1.531 1.550 1.531 1.550 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.370 1.470 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.455 1.466 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.966 1.411 transfer_rs2pw 451 10.6 0.005 0.006 1.242 1.405 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.382 1.396 mp_allgather_i34 2055 14.4 0.341 1.356 0.341 1.356 mp_waitany 17072 13.8 1.130 1.302 1.130 1.302 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.058000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.818182, yerr=8.558018 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 741.474304E+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 0.010 0.031 65.186 65.188 qs_mol_dyn_low 1 2.0 0.003 0.003 64.969 64.982 qs_forces 11 3.9 0.003 0.003 64.855 64.855 qs_energies 11 4.9 0.001 0.001 61.486 61.491 scf_env_do_scf 11 5.9 0.000 0.001 52.971 52.974 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.521 41.522 velocity_verlet 10 3.0 0.002 0.002 36.886 36.890 dbcsr_multiply_generic 2055 12.4 0.118 0.122 30.569 30.731 qs_scf_new_mos 99 7.5 0.001 0.001 27.063 27.167 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.062 27.166 ot_scf_mini 99 9.5 0.002 0.003 25.379 25.487 multiply_cannon 2055 13.4 0.238 0.259 23.672 24.753 multiply_cannon_loop 2055 14.4 1.437 1.478 22.350 22.902 ot_mini 99 10.5 0.001 0.001 14.690 14.823 multiply_cannon_multrec 24660 15.4 4.086 6.677 13.037 14.243 rebuild_ks_matrix 110 8.3 0.000 0.000 11.841 11.959 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.012 11.841 11.958 init_scf_loop 11 6.9 0.000 0.000 11.412 11.413 qs_ot_get_derivative 99 11.5 0.001 0.001 10.525 10.635 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.458 10.561 prepare_preconditioner 11 7.9 0.000 0.000 9.707 9.723 make_preconditioner 11 8.9 0.000 0.000 9.707 9.722 dbcsr_mm_accdrv_process 52304 16.0 7.750 9.211 8.792 9.704 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.927 9.411 sum_up_and_integrate 110 10.3 0.001 0.002 6.164 6.178 integrate_v_rspace 110 11.3 0.003 0.003 6.138 6.152 mp_waitall_1 121746 16.5 4.081 5.885 4.081 5.885 init_scf_run 11 5.9 0.000 0.001 5.806 5.806 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.805 5.806 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.770 5.781 calculate_rho_elec 110 8.6 0.078 0.081 5.769 5.781 qs_ot_get_p 110 10.4 0.001 0.002 5.616 5.759 make_m2s 4110 13.4 0.059 0.061 5.328 5.582 make_images 4110 14.4 0.576 0.700 5.188 5.440 cp_fm_upper_to_full 70 14.2 3.357 4.878 3.357 4.878 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.050 4.147 apply_single 110 13.6 0.000 0.000 4.050 4.147 ot_diis_step 99 11.5 0.011 0.012 4.126 4.126 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.796 3.810 dbcsr_complete_redistribute 325 12.2 0.412 0.458 2.620 3.781 pw_transfer 1331 11.6 0.065 0.073 3.705 3.739 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.599 3.636 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.504 3.548 grid_integrate_task_list 110 12.3 3.288 3.430 3.288 3.430 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.324 3.386 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.109 3.258 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.236 3.236 fft_wrap_pw1pw2_140 451 13.1 0.531 0.542 3.120 3.158 multiply_cannon_metrocomm3 24660 15.4 0.037 0.037 1.403 3.090 calculate_dm_sparse 110 9.5 0.001 0.001 3.024 3.058 density_rs2pw 110 9.6 0.004 0.004 2.902 3.017 wfi_extrapolate 11 7.9 0.001 0.001 2.946 2.946 make_images_data 4110 15.4 0.048 0.052 2.627 2.944 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.913 2.914 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.771 2.905 mp_alltoall_i22 605 13.7 1.658 2.841 1.658 2.841 hybrid_alltoall_any 4261 16.3 0.123 0.460 2.275 2.834 calculate_first_density_matrix 1 7.0 0.000 0.000 2.764 2.768 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.722 2.731 cp_fm_diag_elpa_base 48 14.0 2.577 2.636 2.719 2.730 acc_transpose_blocks 24660 15.4 0.111 0.114 2.612 2.724 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.570 2.618 multiply_cannon_sync_h2d 24660 15.4 2.381 2.572 2.381 2.572 cp_fm_cholesky_invert 11 10.9 2.561 2.570 2.561 2.570 fft3d_ps 1111 14.6 1.153 1.192 2.425 2.439 grid_collocate_task_list 110 9.6 2.264 2.438 2.264 2.438 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.223 2.223 potential_pw2rs 110 12.3 0.012 0.013 1.991 1.997 mp_alltoall_d11v 2046 13.8 1.702 1.905 1.702 1.905 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.799 1.833 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.611 1.707 jit_kernel_multiply 9 15.8 0.699 1.691 0.699 1.691 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.622 1.634 cp_fm_cholesky_decompose 22 10.9 1.595 1.631 1.595 1.631 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.582 1.599 mp_sum_l 6594 12.7 0.878 1.556 0.878 1.556 acc_transpose_blocks_sync 73980 16.4 1.448 1.541 1.448 1.541 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.843 1.505 mp_irecv_dv 62702 16.1 0.741 1.428 0.741 1.428 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.188000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=701.636364, yerr=11.145914 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 875.552768E+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 0.010 0.029 57.896 57.898 qs_mol_dyn_low 1 2.0 0.003 0.003 57.666 57.675 qs_forces 11 3.9 0.003 0.003 57.603 57.604 qs_energies 11 4.9 0.001 0.002 53.916 53.922 scf_env_do_scf 11 5.9 0.000 0.001 45.105 45.105 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 36.848 36.848 velocity_verlet 10 3.0 0.002 0.002 32.371 32.374 dbcsr_multiply_generic 2055 12.4 0.108 0.110 24.256 24.429 qs_scf_new_mos 99 7.5 0.001 0.001 21.848 21.914 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.847 21.913 ot_scf_mini 99 9.5 0.002 0.002 20.564 20.621 multiply_cannon 2055 13.4 0.238 0.249 18.273 19.655 multiply_cannon_loop 2055 14.4 0.602 0.623 16.911 17.264 rebuild_ks_matrix 110 8.3 0.000 0.000 11.945 12.006 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.945 12.006 ot_mini 99 10.5 0.001 0.001 11.333 11.393 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.635 10.687 multiply_cannon_multrec 8220 15.4 3.188 4.812 7.609 8.891 init_scf_loop 11 6.9 0.000 0.000 8.211 8.212 mp_waitall_1 103326 16.6 6.418 8.206 6.418 8.206 qs_ot_get_derivative 99 11.5 0.001 0.001 7.542 7.599 prepare_preconditioner 11 7.9 0.000 0.000 6.488 6.496 make_preconditioner 11 8.9 0.000 0.000 6.488 6.496 sum_up_and_integrate 110 10.3 0.001 0.002 6.429 6.442 integrate_v_rspace 110 11.3 0.003 0.003 6.402 6.415 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.224 6.234 calculate_rho_elec 110 8.6 0.113 0.113 6.224 6.233 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.060 6.148 init_scf_run 11 5.9 0.000 0.001 5.658 5.658 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.658 5.658 qs_ot_get_p 110 10.4 0.001 0.002 5.170 5.241 dbcsr_mm_accdrv_process 17442 15.9 3.102 4.152 4.282 5.197 make_m2s 4110 13.4 0.038 0.039 4.481 4.752 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 3.141 4.638 make_images 4110 14.4 0.646 0.705 4.350 4.622 pw_transfer 1331 11.6 0.066 0.070 4.242 4.254 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.135 4.150 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.808 3.853 apply_single 110 13.6 0.000 0.000 3.807 3.853 ot_diis_step 99 11.5 0.012 0.012 3.755 3.756 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.699 3.703 fft_wrap_pw1pw2_140 451 13.1 0.719 0.733 3.545 3.560 grid_integrate_task_list 110 12.3 3.364 3.535 3.364 3.535 density_rs2pw 110 9.6 0.004 0.004 3.241 3.415 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.329 3.329 cp_fm_cholesky_invert 11 10.9 3.242 3.246 3.242 3.246 wfi_extrapolate 11 7.9 0.001 0.001 2.822 2.822 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.756 2.763 cp_fm_diag_elpa_base 48 14.0 2.697 2.724 2.755 2.762 make_images_data 4110 15.4 0.040 0.046 2.378 2.752 calculate_first_density_matrix 1 7.0 0.000 0.000 2.736 2.739 hybrid_alltoall_any 4261 16.3 0.199 0.850 2.313 2.732 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.722 2.723 fft3d_ps 1111 14.6 1.293 1.300 2.680 2.695 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.666 2.666 calculate_dm_sparse 110 9.5 0.001 0.001 2.553 2.583 multiply_cannon_sync_h2d 8220 15.4 2.331 2.573 2.331 2.573 grid_collocate_task_list 110 9.6 2.360 2.511 2.360 2.511 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.348 2.376 potential_pw2rs 110 12.3 0.015 0.015 2.200 2.208 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.147 2.189 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.085 2.115 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.784 2.003 mp_alltoall_d11v 2046 13.8 1.724 1.978 1.724 1.978 cp_fm_cholesky_decompose 22 10.9 1.787 1.803 1.787 1.803 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.640 1.648 mp_allgather_i34 2055 14.4 0.531 1.645 0.531 1.645 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.497 1.609 acc_transpose_blocks 8220 15.4 0.040 0.042 1.575 1.601 dbcsr_complete_redistribute 325 12.2 0.549 0.570 1.474 1.550 jit_kernel_multiply 9 15.6 0.865 1.530 0.865 1.530 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.478 1.486 multiply_cannon_metrocomm1 8220 15.4 0.022 0.022 0.963 1.476 transfer_rs2pw 451 10.6 0.005 0.005 1.240 1.421 multiply_cannon_metrocomm4 6165 15.4 0.018 0.020 0.478 1.399 mp_irecv_dv 24056 15.7 0.452 1.355 0.452 1.355 qs_create_task_list 11 7.9 0.000 0.000 1.204 1.304 generate_qs_task_list 11 8.9 0.373 0.440 1.204 1.304 mp_waitany 9240 13.8 1.105 1.302 1.105 1.302 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.172 1.219 mp_alltoall_z22v 1111 16.6 1.154 1.176 1.154 1.176 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.898000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=818.181818, yerr=15.367132 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 1.383014E+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 0.077 0.102 87.980 87.993 qs_mol_dyn_low 1 2.0 0.003 0.003 87.234 87.259 qs_forces 11 3.9 0.030 0.058 87.170 87.170 qs_energies 11 4.9 0.001 0.001 82.975 83.003 scf_env_do_scf 11 5.9 0.000 0.001 72.877 72.878 velocity_verlet 10 3.0 0.002 0.002 55.766 55.771 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.621 44.622 dbcsr_multiply_generic 2055 12.4 0.122 0.126 29.424 29.521 init_scf_loop 11 6.9 0.000 0.000 28.187 28.189 qs_scf_new_mos 99 7.5 0.001 0.001 26.736 26.768 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.735 26.768 prepare_preconditioner 11 7.9 0.000 0.000 26.177 26.185 make_preconditioner 11 8.9 0.000 0.000 26.177 26.185 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.343 25.655 ot_scf_mini 99 9.5 0.002 0.002 24.943 24.968 multiply_cannon 2055 13.4 0.331 0.356 22.438 23.197 multiply_cannon_loop 2055 14.4 0.827 0.850 20.666 20.942 cp_fm_upper_to_full 70 14.2 12.731 18.294 12.731 18.294 ot_mini 99 10.5 0.001 0.001 14.087 14.113 rebuild_ks_matrix 110 8.3 0.000 0.001 13.919 13.954 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.919 13.953 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.627 12.657 dbcsr_complete_redistribute 325 12.2 1.002 1.043 7.362 10.658 multiply_cannon_multrec 8220 15.4 4.149 4.356 9.825 9.995 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.375 9.667 qs_ot_get_derivative 99 11.5 0.001 0.001 9.528 9.554 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.819 9.087 mp_alltoall_i22 605 13.7 5.450 8.720 5.450 8.720 mp_waitall_1 84994 16.7 7.392 8.237 7.392 8.237 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.460 7.503 calculate_rho_elec 110 8.6 0.222 0.223 7.459 7.502 sum_up_and_integrate 110 10.3 0.002 0.002 6.901 6.918 integrate_v_rspace 110 11.3 0.004 0.004 6.873 6.890 pw_transfer 1331 11.6 0.075 0.075 5.799 5.807 init_scf_run 11 5.9 0.000 0.001 5.802 5.802 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.801 5.801 make_m2s 4110 13.4 0.043 0.044 5.272 5.759 dbcsr_mm_accdrv_process 11614 15.7 3.756 4.137 5.527 5.753 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.682 5.689 qs_ot_get_p 110 10.4 0.002 0.002 5.595 5.626 make_images 4110 14.4 0.885 0.938 5.084 5.570 cp_fm_cholesky_invert 11 10.9 5.164 5.168 5.164 5.168 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.701 5.166 apply_single 110 13.6 0.000 0.000 4.701 5.166 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.798 5.080 fft_wrap_pw1pw2_140 451 13.1 1.368 1.375 4.989 4.998 ot_diis_step 99 11.5 0.015 0.016 4.543 4.544 qs_ot_p2m_diag 48 11.0 0.151 0.157 4.038 4.048 density_rs2pw 110 9.6 0.004 0.004 3.922 3.932 grid_integrate_task_list 110 12.3 3.680 3.736 3.680 3.736 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.691 3.719 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.519 3.520 hybrid_alltoall_any 4261 16.3 0.264 0.563 2.776 3.500 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.988 3.441 fft3d_ps 1111 14.6 1.882 1.889 3.386 3.396 make_images_data 4110 15.4 0.045 0.048 2.758 3.355 wfi_extrapolate 11 7.9 0.001 0.001 3.318 3.318 calculate_dm_sparse 110 9.5 0.001 0.001 3.222 3.250 multiply_cannon_sync_h2d 8220 15.4 3.121 3.163 3.121 3.163 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.940 2.940 cp_fm_diag_elpa_base 48 14.0 2.395 2.597 2.939 2.939 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.861 2.862 grid_collocate_task_list 110 9.6 2.669 2.687 2.669 2.687 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.634 2.651 potential_pw2rs 110 12.3 0.021 0.021 2.427 2.429 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.372 2.391 calculate_first_density_matrix 1 7.0 0.000 0.000 2.360 2.368 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.218 2.270 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.102 2.237 mp_alltoall_d11v 2046 13.8 2.013 2.074 2.013 2.074 cp_fm_cholesky_decompose 22 10.9 1.956 1.968 1.956 1.968 qs_create_task_list 11 7.9 0.000 0.000 1.871 1.918 generate_qs_task_list 11 8.9 0.729 0.781 1.871 1.917 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.891 1.895 acc_transpose_blocks 8220 15.4 0.041 0.041 1.728 1.777 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.993000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1242.818182, yerr=60.834826 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/15/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 1.086553E+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 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 632.307712E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 57688. MP_Allreduce 11097 796. MP_Sync 86 MP_Alltoall 2210 3331812. MP_SendRecv 24130 18752. MP_ISendRecv 24130 18752. MP_Wait 42150 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.031 206.328 206.329 qs_mol_dyn_low 1 2.0 0.003 0.003 205.888 205.901 qs_forces 11 3.9 0.005 0.005 205.801 205.801 qs_energies 11 4.9 0.001 0.002 200.221 200.245 scf_env_do_scf 11 5.9 0.001 0.002 183.693 183.698 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 163.341 163.342 dbcsr_multiply_generic 2485 12.5 0.176 0.180 126.013 127.013 qs_scf_new_mos 116 7.6 0.001 0.001 124.448 124.781 qs_scf_loop_do_ot 116 8.6 0.001 0.001 124.447 124.780 velocity_verlet 10 3.0 0.001 0.002 122.258 122.259 ot_scf_mini 116 9.6 0.003 0.003 117.762 118.103 multiply_cannon 2485 13.5 0.234 0.242 101.587 104.475 multiply_cannon_loop 2485 14.5 2.366 2.420 99.468 101.880 ot_mini 116 10.6 0.001 0.001 66.601 66.921 multiply_cannon_multrec 59640 15.5 31.608 33.954 41.354 43.681 qs_ot_get_derivative 116 11.6 0.001 0.001 41.880 42.176 rebuild_ks_matrix 127 8.3 0.001 0.001 33.262 33.745 qs_ks_build_kohn_sham_matrix 127 9.3 0.019 0.033 33.261 33.745 mp_waitall_1 264810 16.5 29.168 32.056 29.168 32.056 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.832 30.282 qs_ot_get_p 127 10.4 0.002 0.005 29.031 29.380 multiply_cannon_sync_h2d 59640 15.5 26.179 28.388 26.179 28.388 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.186 24.799 apply_single 127 13.6 0.001 0.001 24.186 24.799 ot_diis_step 116 11.6 0.008 0.008 24.436 24.437 qs_ot_p2m_diag 82 11.4 0.078 0.090 22.414 22.527 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 20.155 20.407 init_scf_loop 11 6.9 0.000 0.000 20.278 20.279 cp_dbcsr_syevd 82 12.4 0.005 0.005 19.501 19.502 multiply_cannon_metrocomm3 59640 15.5 0.113 0.119 16.392 18.362 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.617 16.655 cp_fm_diag_elpa_base 82 14.4 16.552 16.590 16.613 16.653 prepare_preconditioner 11 7.9 0.000 0.000 15.634 15.686 make_preconditioner 11 8.9 0.000 0.000 15.634 15.686 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.890 15.073 make_m2s 4970 13.5 0.103 0.112 13.669 14.071 make_images 4970 14.5 0.395 0.412 13.486 13.896 sum_up_and_integrate 127 10.3 0.002 0.004 13.807 13.821 integrate_v_rspace 127 11.3 0.003 0.004 13.749 13.765 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.911 13.034 calculate_rho_elec 127 8.7 0.045 0.063 12.910 13.033 init_scf_run 11 5.9 0.000 0.001 12.421 12.422 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.421 12.422 mp_sum_l 7884 12.9 9.505 10.876 9.505 10.876 dbcsr_mm_accdrv_process 123452 16.2 4.701 4.846 9.314 9.784 wfi_extrapolate 11 7.9 0.001 0.001 9.040 9.040 cp_fm_cholesky_invert 11 10.9 8.890 8.898 8.890 8.898 calculate_dm_sparse 127 9.5 0.001 0.001 8.630 8.737 multiply_cannon_metrocomm1 59640 15.5 0.092 0.097 6.417 8.402 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.173 8.342 qs_ot_get_orbitals 116 10.6 0.001 0.001 8.165 8.272 pw_transfer 1535 11.6 0.073 0.090 7.503 7.673 make_images_data 4970 15.5 0.066 0.072 6.670 7.632 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.303 7.474 grid_integrate_task_list 127 12.3 6.995 7.378 6.995 7.378 density_rs2pw 127 9.7 0.006 0.006 6.617 7.253 hybrid_alltoall_any 5155 16.4 0.290 2.234 5.858 7.038 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.692 6.703 fft_wrap_pw1pw2_140 519 13.2 0.851 0.892 6.417 6.590 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.823 5.956 fft3d_ps 1281 14.7 2.179 2.830 5.326 5.648 mp_alltoall_d11v 2401 14.1 4.251 5.456 4.251 5.456 grid_collocate_task_list 127 9.7 4.806 5.217 4.806 5.217 cp_fm_cholesky_decompose 22 10.9 4.583 4.597 4.583 4.597 potential_pw2rs 127 12.3 0.009 0.010 4.386 4.406 mp_sum_d 4456 12.1 3.458 4.151 3.458 4.151 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.329000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=599.727273, yerr=7.287485 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+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 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 844.021760E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 5224120. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.034 189.983 189.985 qs_mol_dyn_low 1 2.0 0.003 0.003 189.578 189.597 qs_forces 11 3.9 0.004 0.004 187.944 187.945 qs_energies 11 4.9 0.001 0.002 181.191 181.203 scf_env_do_scf 11 5.9 0.001 0.002 164.977 164.987 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 132.225 132.227 velocity_verlet 10 3.0 0.001 0.002 120.362 120.396 dbcsr_multiply_generic 2507 12.6 0.188 0.195 97.292 98.621 qs_scf_new_mos 117 7.6 0.001 0.001 93.910 94.630 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.909 94.630 ot_scf_mini 117 9.6 0.004 0.004 89.091 89.736 multiply_cannon 2507 13.6 0.504 0.575 77.327 81.930 multiply_cannon_loop 2507 14.6 1.561 1.631 74.001 76.731 ot_mini 117 10.6 0.001 0.001 50.278 50.975 mp_waitall_1 214728 16.6 24.055 39.393 24.055 39.393 multiply_cannon_multrec 30084 15.6 21.036 25.374 31.596 36.127 rebuild_ks_matrix 128 8.3 0.001 0.001 32.143 32.746 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.143 32.745 init_scf_loop 11 6.9 0.000 0.000 32.666 32.667 multiply_cannon_metrocomm3 30084 15.6 0.098 0.104 15.713 29.456 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.891 29.437 qs_ot_get_derivative 117 11.6 0.001 0.002 28.106 28.748 prepare_preconditioner 11 7.9 0.000 0.000 28.359 28.440 make_preconditioner 11 8.9 0.000 0.000 28.359 28.440 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.016 27.537 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.303 23.443 apply_single 128 13.6 0.001 0.001 22.303 23.443 ot_diis_step 117 11.6 0.014 0.015 22.000 22.002 qs_ot_get_p 128 10.4 0.001 0.002 20.798 21.434 multiply_cannon_sync_h2d 30084 15.6 18.143 21.041 18.143 21.041 cp_fm_cholesky_invert 11 10.9 16.497 16.510 16.497 16.510 qs_ot_p2m_diag 83 11.4 0.188 0.216 16.024 16.058 make_m2s 5014 13.6 0.087 0.092 14.137 15.779 make_images 5014 14.6 1.173 1.371 13.919 15.560 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.822 14.823 sum_up_and_integrate 128 10.3 0.002 0.004 13.887 13.921 integrate_v_rspace 128 11.3 0.003 0.003 13.827 13.864 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.267 13.301 calculate_rho_elec 128 8.7 0.087 0.103 13.266 13.300 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.684 11.718 cp_fm_diag_elpa_base 83 14.4 11.433 11.522 11.679 11.707 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.105 11.566 init_scf_run 11 5.9 0.000 0.001 11.432 11.433 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.431 11.433 multiply_cannon_metrocomm4 27577 15.6 0.097 0.110 3.798 10.743 dbcsr_mm_accdrv_process 62242 16.2 5.515 6.321 10.005 10.579 mp_irecv_dv 69486 16.3 3.601 10.357 3.601 10.357 make_images_data 5014 15.6 0.066 0.075 8.362 10.191 hybrid_alltoall_any 5200 16.5 0.349 1.564 7.191 9.699 pw_transfer 1547 11.6 0.085 0.102 8.536 8.596 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.313 8.376 wfi_extrapolate 11 7.9 0.001 0.001 8.344 8.344 density_rs2pw 128 9.7 0.006 0.006 6.994 7.484 grid_integrate_task_list 128 12.3 7.132 7.458 7.132 7.458 fft_wrap_pw1pw2_140 523 13.2 0.933 0.952 7.344 7.425 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.273 7.032 cp_fm_cholesky_decompose 22 10.9 6.935 7.024 6.935 7.024 calculate_dm_sparse 128 9.5 0.001 0.001 6.484 6.602 mp_sum_l 7950 12.9 4.030 6.093 4.030 6.093 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.054 6.063 fft3d_ps 1291 14.7 2.825 3.009 5.873 5.921 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.396 5.463 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.174 5.328 grid_collocate_task_list 128 9.7 5.001 5.314 5.001 5.314 mp_allgather_i34 2507 14.6 1.844 4.958 1.844 4.958 mp_alltoall_d11v 2415 14.1 4.152 4.937 4.152 4.937 potential_pw2rs 128 12.3 0.016 0.018 4.435 4.451 dbcsr_complete_redistribute 395 12.7 0.764 0.866 3.167 3.997 mp_sum_d 4472 12.1 2.608 3.973 2.608 3.973 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.985000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=804.272727, yerr=1.542778 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 946.536448E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57952. MP_Allreduce 11127 998. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.037 175.667 175.669 qs_mol_dyn_low 1 2.0 0.003 0.003 175.197 175.250 qs_forces 11 3.9 0.004 0.005 172.739 172.742 qs_energies 11 4.9 0.003 0.014 166.170 166.188 scf_env_do_scf 11 5.9 0.001 0.002 150.753 150.755 scf_env_do_scf_inner_loop 117 6.6 0.004 0.013 116.155 116.156 velocity_verlet 10 3.0 0.001 0.002 113.576 113.652 dbcsr_multiply_generic 2507 12.6 0.188 0.194 81.501 82.700 qs_scf_new_mos 117 7.6 0.001 0.001 79.283 79.634 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.282 79.633 ot_scf_mini 117 9.6 0.003 0.004 75.049 75.448 multiply_cannon 2507 13.6 0.511 0.528 62.317 67.370 multiply_cannon_loop 2507 14.6 1.119 1.184 59.197 61.837 ot_mini 117 10.6 0.001 0.001 42.423 42.817 init_scf_loop 11 6.9 0.001 0.005 34.503 34.504 mp_waitall_1 170520 16.6 24.053 32.851 24.053 32.851 rebuild_ks_matrix 128 8.3 0.001 0.001 30.091 30.596 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.042 30.090 30.595 prepare_preconditioner 11 7.9 0.000 0.000 30.520 30.570 make_preconditioner 11 8.9 0.000 0.001 30.520 30.570 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.135 29.562 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.070 27.519 multiply_cannon_multrec 20056 15.6 13.160 16.108 22.471 25.546 multiply_cannon_metrocomm3 20056 15.6 0.061 0.064 15.216 24.342 qs_ot_get_derivative 117 11.6 0.001 0.002 22.955 23.345 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.546 20.594 apply_single 128 13.6 0.001 0.001 19.545 20.594 ot_diis_step 117 11.6 0.018 0.018 19.356 19.356 qs_ot_get_p 128 10.4 0.001 0.002 18.099 18.573 make_m2s 5014 13.6 0.078 0.085 14.110 15.526 multiply_cannon_sync_h2d 20056 15.6 13.675 15.349 13.675 15.349 make_images 5014 14.6 1.153 1.232 13.873 15.282 cp_fm_cholesky_invert 11 10.9 14.126 14.135 14.126 14.135 sum_up_and_integrate 128 10.3 0.002 0.003 13.901 13.932 integrate_v_rspace 128 11.3 0.003 0.004 13.841 13.871 qs_ot_p2m_diag 83 11.4 0.265 0.272 13.832 13.839 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.492 13.524 calculate_rho_elec 128 8.7 0.129 0.143 13.492 13.524 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.790 12.791 init_scf_run 11 5.9 0.000 0.001 10.374 10.374 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.374 10.374 make_images_data 5014 15.6 0.063 0.074 8.271 10.129 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.675 9.696 cp_fm_diag_elpa_base 83 14.4 9.271 9.426 9.672 9.692 hybrid_alltoall_any 5200 16.5 0.448 2.031 7.351 9.577 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.069 9.353 multiply_cannon_metrocomm4 17549 15.6 0.064 0.074 3.465 9.253 mp_irecv_dv 50230 16.2 3.338 9.008 3.338 9.008 dbcsr_mm_accdrv_process 41502 16.2 5.644 5.935 8.764 8.928 pw_transfer 1547 11.6 0.084 0.103 8.776 8.888 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.553 8.671 fft_wrap_pw1pw2_140 523 13.2 1.063 1.102 7.619 7.744 grid_integrate_task_list 128 12.3 7.306 7.617 7.306 7.617 wfi_extrapolate 11 7.9 0.001 0.001 7.374 7.374 cp_fm_upper_to_full 105 14.8 5.755 7.305 5.755 7.305 cp_fm_cholesky_decompose 22 10.9 7.192 7.210 7.192 7.210 density_rs2pw 128 9.7 0.006 0.006 6.837 7.188 dbcsr_complete_redistribute 395 12.7 1.160 1.183 4.543 6.268 fft3d_ps 1291 14.7 2.822 3.073 5.865 5.980 calculate_dm_sparse 128 9.5 0.001 0.001 5.819 5.919 grid_collocate_task_list 128 9.7 5.153 5.550 5.153 5.550 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.401 5.407 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.619 5.309 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.409 5.135 mp_alltoall_d11v 2415 14.1 4.262 4.779 4.262 4.779 mp_sum_l 7950 12.9 3.249 4.757 3.249 4.757 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.534 4.663 mp_allgather_i34 2507 14.6 1.613 4.405 1.613 4.405 potential_pw2rs 128 12.3 0.020 0.022 4.332 4.352 transfer_fm_to_dbcsr 11 9.9 0.019 0.023 2.366 4.061 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.896 3.923 mp_alltoall_i22 716 14.1 1.989 3.896 1.989 3.896 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.746 3.747 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.517 3.571 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.669000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=899.454545, yerr=11.211712 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 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 1986252263424 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 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239167967232 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239167967232 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 94.233007E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383584 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.138004E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+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 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922014800 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57626. MP_Allreduce 11184 1079. MP_Sync 88 MP_Alltoall 1724 12509628. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.033 189.428 189.429 qs_mol_dyn_low 1 2.0 0.003 0.003 189.051 189.065 qs_forces 11 3.9 0.004 0.004 188.948 188.958 qs_energies 11 4.9 0.001 0.002 181.828 181.838 scf_env_do_scf 11 5.9 0.001 0.001 165.030 165.042 velocity_verlet 10 3.0 0.002 0.002 126.474 126.477 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 118.473 118.478 dbcsr_multiply_generic 2529 12.6 0.194 0.201 82.334 83.026 qs_scf_new_mos 118 7.6 0.001 0.001 81.800 82.140 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.800 82.140 ot_scf_mini 118 9.6 0.003 0.003 76.920 77.262 multiply_cannon 2529 13.6 0.559 0.602 56.898 60.065 multiply_cannon_loop 2529 14.6 1.869 1.956 52.554 54.190 init_scf_loop 11 6.9 0.000 0.000 46.438 46.452 ot_mini 118 10.6 0.001 0.001 44.050 44.372 prepare_preconditioner 11 7.9 0.000 0.000 42.322 42.362 make_preconditioner 11 8.9 0.000 0.000 42.322 42.362 make_full_inverse_cholesky 11 9.9 0.013 0.022 35.795 40.974 multiply_cannon_multrec 30348 15.6 13.783 18.978 26.825 31.715 rebuild_ks_matrix 129 8.3 0.001 0.001 29.545 29.849 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.018 29.544 29.848 mp_waitall_1 149172 16.7 17.573 28.310 17.573 28.310 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.667 26.940 qs_ot_get_derivative 118 11.6 0.001 0.002 23.746 24.092 make_m2s 5058 13.6 0.093 0.097 21.017 22.801 make_images 5058 14.6 1.998 2.317 20.712 22.501 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.757 20.231 apply_single 129 13.6 0.001 0.001 19.757 20.231 ot_diis_step 118 11.6 0.018 0.019 20.175 20.177 qs_ot_get_p 129 10.4 0.001 0.002 18.681 19.034 cp_fm_upper_to_full 106 14.8 11.712 17.289 11.712 17.289 cp_fm_cholesky_invert 11 10.9 15.879 15.888 15.879 15.888 multiply_cannon_metrocomm3 30348 15.6 0.049 0.052 6.394 14.637 qs_ot_p2m_diag 84 11.4 0.346 0.394 14.537 14.590 sum_up_and_integrate 129 10.3 0.002 0.003 14.100 14.126 integrate_v_rspace 129 11.3 0.003 0.004 14.040 14.067 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.931 13.987 calculate_rho_elec 129 8.7 0.171 0.187 13.930 13.986 make_images_data 5058 15.6 0.066 0.072 11.448 13.951 hybrid_alltoall_any 5245 16.5 0.537 2.235 10.273 13.510 dbcsr_mm_accdrv_process 62780 16.2 8.549 9.369 12.602 13.133 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.122 13.123 dbcsr_complete_redistribute 397 12.7 1.488 1.594 9.083 12.995 copy_fm_to_dbcsr 210 11.7 0.001 0.002 7.742 11.630 multiply_cannon_sync_h2d 30348 15.6 10.493 11.133 10.493 11.133 init_scf_run 11 5.9 0.000 0.001 10.900 10.902 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.900 10.901 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 10.156 10.399 transfer_fm_to_dbcsr 11 9.9 0.002 0.009 6.504 10.311 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.062 10.076 cp_fm_diag_elpa_base 84 14.4 9.081 9.421 10.056 10.068 mp_alltoall_i22 720 14.1 5.772 9.631 5.772 9.631 pw_transfer 1559 11.6 0.086 0.104 9.228 9.300 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.000 9.077 fft_wrap_pw1pw2_140 527 13.2 1.254 1.289 8.026 8.121 grid_integrate_task_list 129 12.3 7.588 7.883 7.588 7.883 multiply_cannon_metrocomm4 25290 15.6 0.080 0.089 2.828 7.767 wfi_extrapolate 11 7.9 0.001 0.001 7.732 7.732 cp_fm_cholesky_decompose 22 10.9 7.409 7.508 7.409 7.508 mp_irecv_dv 76751 16.2 2.676 7.485 2.676 7.485 density_rs2pw 129 9.7 0.005 0.006 6.853 7.224 calculate_dm_sparse 129 9.5 0.001 0.001 6.611 6.742 fft3d_ps 1301 14.7 3.018 3.073 6.031 6.072 grid_collocate_task_list 129 9.7 5.322 5.719 5.322 5.719 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.560 5.619 mp_alltoall_d11v 2429 14.1 5.064 5.507 5.064 5.507 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.537 4.658 mp_allgather_i34 2529 14.6 2.413 4.548 2.413 4.548 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.447 4.540 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.405 4.405 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.291 4.344 potential_pw2rs 129 12.3 0.023 0.023 4.313 4.325 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.429000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1077.636364, yerr=13.282064 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/19/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 5.910120E+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 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.496015E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526748E+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 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592466352 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4045 57596. MP_Allreduce 11189 1162. 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 0.026 0.070 169.757 169.763 qs_mol_dyn_low 1 2.0 0.003 0.003 169.334 169.358 qs_forces 11 3.9 0.004 0.004 169.240 169.243 qs_energies 11 4.9 0.002 0.002 161.745 161.751 scf_env_do_scf 11 5.9 0.001 0.001 144.413 144.423 velocity_verlet 10 3.0 0.002 0.002 112.197 112.200 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 108.794 108.795 dbcsr_multiply_generic 2529 12.6 0.192 0.201 72.065 72.409 qs_scf_new_mos 118 7.6 0.001 0.001 71.364 71.431 qs_scf_loop_do_ot 118 8.6 0.001 0.001 71.363 71.430 ot_scf_mini 118 9.6 0.003 0.004 66.913 66.962 multiply_cannon 2529 13.6 0.566 0.599 53.436 56.715 multiply_cannon_loop 2529 14.6 0.807 0.848 50.443 51.144 ot_mini 118 10.6 0.001 0.001 37.322 37.374 init_scf_loop 11 6.9 0.000 0.000 35.470 35.472 prepare_preconditioner 11 7.9 0.000 0.000 31.629 31.650 make_preconditioner 11 8.9 0.000 0.000 31.629 31.650 mp_waitall_1 126876 16.7 24.510 30.796 24.510 30.796 make_full_inverse_cholesky 11 9.9 0.020 0.028 29.568 29.816 rebuild_ks_matrix 129 8.3 0.001 0.001 29.494 29.562 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.494 29.561 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.885 26.954 multiply_cannon_multrec 10116 15.6 10.464 14.541 18.186 21.153 qs_ot_get_derivative 118 11.6 0.002 0.002 20.312 20.361 multiply_cannon_metrocomm3 10116 15.6 0.024 0.025 12.542 19.844 cp_fm_cholesky_invert 11 10.9 18.065 18.071 18.065 18.071 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.016 17.098 apply_single 129 13.6 0.001 0.001 17.016 17.097 ot_diis_step 118 11.6 0.020 0.021 16.942 16.942 qs_ot_get_p 129 10.4 0.001 0.002 16.816 16.886 make_m2s 5058 13.6 0.065 0.069 14.763 15.644 make_images 5058 14.6 2.179 2.601 14.451 15.325 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.331 14.356 calculate_rho_elec 129 8.7 0.255 0.266 14.330 14.355 sum_up_and_integrate 129 10.3 0.002 0.002 14.279 14.325 integrate_v_rspace 129 11.3 0.004 0.004 14.218 14.264 qs_ot_p2m_diag 84 11.4 0.502 0.507 13.097 13.112 cp_dbcsr_syevd 84 12.4 0.005 0.006 11.925 11.925 multiply_cannon_sync_h2d 10116 15.6 11.014 11.247 11.014 11.247 init_scf_run 11 5.9 0.000 0.001 10.505 10.505 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.505 10.505 pw_transfer 1559 11.6 0.086 0.094 9.975 10.014 hybrid_alltoall_any 5245 16.5 0.846 3.805 8.356 9.898 make_images_data 5058 15.6 0.055 0.065 8.528 9.833 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 9.750 9.795 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.050 9.062 cp_fm_diag_elpa_base 84 14.4 8.812 8.880 9.046 9.058 fft_wrap_pw1pw2_140 527 13.2 1.691 1.734 8.653 8.700 dbcsr_mm_accdrv_process 20934 16.1 3.225 4.581 7.367 8.077 grid_integrate_task_list 129 12.3 7.836 8.066 7.836 8.066 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.012 8.051 cp_fm_cholesky_decompose 22 10.9 7.902 8.029 7.902 8.029 wfi_extrapolate 11 7.9 0.001 0.001 7.415 7.415 density_rs2pw 129 9.7 0.005 0.005 6.991 7.266 multiply_cannon_metrocomm1 10116 15.6 0.030 0.031 4.218 6.990 calculate_dm_sparse 129 9.5 0.001 0.001 6.168 6.230 fft3d_ps 1301 14.7 3.163 3.268 6.176 6.199 grid_collocate_task_list 129 9.7 5.618 5.807 5.618 5.807 dbcsr_complete_redistribute 397 12.7 2.105 2.157 5.136 5.458 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.254 5.255 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.204 5.242 mp_alltoall_d11v 2429 14.1 4.530 4.910 4.530 4.910 mp_allgather_i34 2529 14.6 1.157 4.598 1.157 4.598 potential_pw2rs 129 12.3 0.026 0.027 4.394 4.401 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.088 4.149 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.584 3.922 multiply_cannon_metrocomm4 7587 15.6 0.026 0.028 1.720 3.680 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.330 3.661 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.548 3.622 mp_irecv_dv 29102 15.9 1.681 3.609 1.681 3.609 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.579 3.604 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.517 3.527 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.405 3.520 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=169.763000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1405.545455, yerr=41.083869 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 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 11.787404E+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 1982128 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.0 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 3.195404E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101360 MPI messages size (bytes): total size 1.145137E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.297725E+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 45848 35601252352 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592388592 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58558. MP_Allreduce 11184 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.067 291.909 291.913 qs_mol_dyn_low 1 2.0 0.003 0.003 291.344 291.372 qs_forces 11 3.9 0.005 0.005 291.127 291.129 qs_energies 11 4.9 0.002 0.002 282.210 282.215 scf_env_do_scf 11 5.9 0.001 0.001 259.937 259.944 velocity_verlet 10 3.0 0.002 0.002 211.541 211.564 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 135.976 135.978 init_scf_loop 11 6.9 0.000 0.000 123.685 123.688 prepare_preconditioner 11 7.9 0.000 0.000 118.837 118.860 make_preconditioner 11 8.9 0.000 0.000 118.837 118.860 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.432 115.967 qs_scf_new_mos 118 7.6 0.001 0.001 90.546 90.617 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.545 90.616 ot_scf_mini 118 9.6 0.004 0.004 85.695 85.722 dbcsr_multiply_generic 2534 12.6 0.218 0.229 82.448 82.980 cp_fm_upper_to_full 106 14.8 52.731 75.060 52.731 75.060 multiply_cannon 2534 13.6 0.693 0.730 59.153 60.465 multiply_cannon_loop 2534 14.6 1.055 1.074 55.078 56.461 ot_mini 118 10.6 0.001 0.001 45.059 45.100 dbcsr_complete_redistribute 397 12.7 3.975 4.030 29.319 41.666 copy_fm_to_dbcsr 210 11.7 0.001 0.002 25.848 38.216 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.363 35.606 rebuild_ks_matrix 129 8.3 0.001 0.001 34.977 35.028 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 34.977 35.027 mp_alltoall_i22 720 14.1 21.121 33.624 21.121 33.624 cp_fm_cholesky_invert 11 10.9 33.597 33.602 33.597 33.602 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.469 32.525 mp_waitall_1 104780 16.8 28.107 32.320 28.107 32.320 qs_ot_get_p 129 10.4 0.002 0.002 25.721 25.809 qs_ot_get_derivative 118 11.6 0.002 0.002 24.887 24.915 qs_ot_p2m_diag 84 11.4 0.890 0.897 21.621 21.651 multiply_cannon_metrocomm3 10136 15.6 0.025 0.026 18.815 20.230 make_m2s 5068 13.6 0.075 0.079 18.697 20.191 ot_diis_step 118 11.6 0.022 0.023 20.139 20.139 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.807 19.810 make_images 5068 14.6 3.091 3.283 18.211 19.706 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.434 19.573 apply_single 129 13.6 0.001 0.001 19.434 19.573 multiply_cannon_multrec 10136 15.6 10.379 12.246 18.210 18.274 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.461 17.482 calculate_rho_elec 129 8.7 0.482 0.482 17.460 17.481 sum_up_and_integrate 129 10.3 0.002 0.002 16.592 16.679 integrate_v_rspace 129 11.3 0.004 0.005 16.529 16.616 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.585 16.586 cp_fm_diag_elpa_base 84 14.4 12.158 13.721 16.580 16.581 multiply_cannon_sync_h2d 10136 15.6 14.422 14.446 14.422 14.446 pw_transfer 1559 11.6 0.096 0.097 13.257 13.262 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 13.017 13.022 hybrid_alltoall_any 5255 16.5 1.322 3.066 10.786 12.946 make_images_data 5068 15.6 0.063 0.068 10.592 12.788 init_scf_run 11 5.9 0.000 0.001 11.983 11.985 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.983 11.984 fft_wrap_pw1pw2_140 527 13.2 3.244 3.328 11.588 11.597 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.824 9.845 dbcsr_mm_accdrv_process 20954 16.1 4.199 6.256 7.576 9.503 cp_fm_cholesky_decompose 22 10.9 8.974 8.994 8.974 8.994 wfi_extrapolate 11 7.9 0.001 0.001 8.891 8.891 grid_integrate_task_list 129 12.3 8.619 8.821 8.619 8.821 density_rs2pw 129 9.7 0.005 0.005 8.423 8.470 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.139 8.140 fft3d_ps 1301 14.7 4.002 4.013 7.440 7.494 mp_alltoall_d11v 2429 14.1 6.932 7.066 6.932 7.066 calculate_dm_sparse 129 9.5 0.001 0.001 6.720 6.754 grid_collocate_task_list 129 9.7 6.465 6.547 6.465 6.547 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.219 6.262 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.123 6.230 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=291.913000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2900.454545, yerr=164.298157 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 1.260667E+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 13113986. 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 0.015 0.031 84.854 84.855 qs_energies 1 2.0 0.000 0.000 84.403 84.410 ls_scf 1 3.0 0.000 0.000 83.513 83.520 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.539 72.728 multiply_cannon 111 7.7 0.017 0.020 55.780 56.984 multiply_cannon_loop 111 8.7 0.227 0.241 52.350 53.797 ls_scf_main 1 4.0 0.000 0.000 52.310 52.311 density_matrix_trs4 2 5.0 0.002 0.003 46.785 46.863 ls_scf_init_scf 1 4.0 0.000 0.000 28.179 28.180 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.057 27.097 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 24.963 24.984 mp_waitall_1 11031 10.9 22.426 24.546 22.426 24.546 multiply_cannon_multrec 2664 9.7 8.158 8.957 15.524 17.309 multiply_cannon_sync_h2d 2664 9.7 13.484 14.853 13.484 14.853 make_m2s 222 7.7 0.009 0.012 13.100 13.552 make_images 222 8.7 0.098 0.109 13.077 13.531 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.670 12.352 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.472 9.643 make_images_data 222 9.7 0.004 0.005 7.618 8.191 hybrid_alltoall_any 227 10.6 0.215 1.835 6.549 8.050 dbcsr_mm_accdrv_process 4760 10.4 0.593 0.701 6.982 7.943 dbcsr_mm_accdrv_process_sort 4760 11.4 6.190 7.084 6.190 7.084 calculate_norms 4752 9.8 5.501 6.089 5.501 6.089 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.024 5.156 mp_sum_l 887 5.1 3.116 4.562 3.116 4.562 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.050 4.211 mp_irecv_dv 6231 10.9 2.033 4.176 2.033 4.176 make_images_sizes 222 9.7 0.000 0.000 0.753 3.644 mp_alltoall_i44 222 10.7 0.753 3.643 0.753 3.643 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.342 3.600 arnoldi_extremal 4 6.8 0.000 0.000 3.248 3.272 arnoldi_normal_ev 4 7.8 0.001 0.004 3.248 3.272 build_subspace 16 8.4 0.009 0.012 3.156 3.158 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 2.848 2.887 dbcsr_special_finalize 555 9.7 0.005 0.006 2.375 2.841 dbcsr_merge_single_wm 555 10.7 0.459 0.594 2.367 2.832 make_images_pack 222 9.7 2.206 2.622 2.208 2.623 dbcsr_sort_data 658 11.4 2.167 2.567 2.167 2.567 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.333 2.566 dbcsr_matrix_vector_mult_local 304 10.0 2.068 2.454 2.070 2.456 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.291 2.371 buffer_matrices_ensure_size 222 8.7 1.756 2.119 1.756 2.119 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.763 1.764 rebuild_ks_matrix 3 7.3 0.000 0.000 1.753 1.755 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.753 1.755 mp_allgather_i34 111 8.7 0.738 1.734 0.738 1.734 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.855000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 2.173850E+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 0.025 0.040 90.091 90.092 qs_energies 1 2.0 0.000 0.000 89.613 89.618 ls_scf 1 3.0 0.000 0.000 88.275 88.278 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.286 74.618 multiply_cannon 111 7.7 0.026 0.036 52.657 56.755 ls_scf_main 1 4.0 0.000 0.000 54.588 54.592 multiply_cannon_loop 111 8.7 0.135 0.147 49.929 53.090 density_matrix_trs4 2 5.0 0.002 0.003 48.845 49.058 ls_scf_init_scf 1 4.0 0.000 0.000 29.917 29.919 mp_waitall_1 9105 10.9 20.883 29.845 20.883 29.845 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.731 28.826 multiply_cannon_multrec 1332 9.7 13.258 17.099 22.522 27.449 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.366 26.379 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.537 20.385 make_m2s 222 7.7 0.006 0.008 15.159 15.700 make_images 222 8.7 1.369 1.699 15.129 15.670 dbcsr_mm_accdrv_process 4041 10.4 0.326 0.544 8.862 10.436 dbcsr_mm_accdrv_process_sort 4041 11.4 8.380 9.892 8.380 9.892 make_images_data 222 9.7 0.004 0.005 8.793 9.652 hybrid_alltoall_any 227 10.6 0.540 2.529 8.187 9.387 mp_sum_l 887 5.1 5.018 8.180 5.018 8.180 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.235 7.881 mp_irecv_dv 3311 11.0 3.216 7.830 3.216 7.830 calculate_norms 2376 9.8 6.064 6.786 6.064 6.786 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.732 6.509 multiply_cannon_sync_h2d 1332 9.7 4.857 6.169 4.857 6.169 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.992 5.205 arnoldi_extremal 4 6.8 0.000 0.000 4.757 4.774 arnoldi_normal_ev 4 7.8 0.001 0.005 4.757 4.774 build_subspace 16 8.4 0.014 0.021 4.505 4.508 ls_scf_post 1 4.0 0.000 0.000 3.769 3.774 ls_scf_store_result 1 5.0 0.000 0.000 3.475 3.594 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.206 3.452 dbcsr_matrix_vector_mult_local 304 10.0 2.786 3.256 2.788 3.258 mp_allgather_i34 111 8.7 0.716 2.878 0.716 2.878 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.193 2.744 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.648 2.741 dbcsr_data_new 4174 10.1 2.113 2.397 2.113 2.397 make_images_pack 222 9.7 1.817 2.138 1.819 2.140 dbcsr_sort_data 436 11.2 1.850 2.123 1.850 2.123 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.865 1.867 rebuild_ks_matrix 3 7.3 0.000 0.000 1.852 1.854 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.852 1.854 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.092000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1780.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 2.852073E+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 0.027 0.042 92.123 92.124 qs_energies 1 2.0 0.000 0.000 91.504 91.507 ls_scf 1 3.0 0.000 0.000 90.076 90.081 dbcsr_multiply_generic 111 6.7 0.016 0.016 74.722 75.125 multiply_cannon 111 7.7 0.032 0.064 51.968 56.693 ls_scf_main 1 4.0 0.000 0.000 55.763 55.768 multiply_cannon_loop 111 8.7 0.116 0.130 49.248 52.343 density_matrix_trs4 2 5.0 0.002 0.003 49.861 50.031 mp_waitall_1 7281 11.0 22.843 32.225 22.843 32.225 ls_scf_init_scf 1 4.0 0.000 0.000 30.590 30.593 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.414 29.478 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.067 27.074 multiply_cannon_multrec 888 9.7 12.714 15.314 21.265 24.598 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.548 21.615 make_m2s 222 7.7 0.006 0.007 16.382 17.078 make_images 222 8.7 1.584 1.843 16.344 17.037 make_images_data 222 9.7 0.004 0.004 9.459 10.517 hybrid_alltoall_any 227 10.6 0.639 2.942 8.918 10.210 dbcsr_mm_accdrv_process 3754 10.4 0.303 0.493 8.074 9.341 dbcsr_mm_accdrv_process_sort 3754 11.4 7.641 8.848 7.641 8.848 mp_sum_l 887 5.1 4.667 7.855 4.667 7.855 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.477 7.367 multiply_cannon_sync_h2d 888 9.7 6.095 7.094 6.095 7.094 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.469 6.996 mp_irecv_dv 2335 11.1 2.453 6.940 2.453 6.940 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.515 5.995 arnoldi_extremal 4 6.8 0.000 0.000 5.209 5.221 arnoldi_normal_ev 4 7.8 0.001 0.005 5.209 5.221 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.776 4.992 build_subspace 16 8.4 0.014 0.020 4.909 4.915 calculate_norms 1584 9.8 4.377 4.756 4.377 4.756 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.545 3.868 ls_scf_post 1 4.0 0.000 0.000 3.723 3.726 mp_allgather_i34 111 8.7 0.829 3.693 0.829 3.693 dbcsr_matrix_vector_mult_local 304 10.0 3.078 3.637 3.081 3.640 ls_scf_store_result 1 5.0 0.000 0.000 3.438 3.549 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.853 2.971 dbcsr_data_new 4116 9.9 2.111 2.445 2.111 2.445 make_images_sizes 222 9.7 0.000 0.000 1.101 2.410 mp_alltoall_i44 222 10.7 1.101 2.409 1.101 2.409 dbcsr_sort_data 325 11.1 1.870 2.080 1.870 2.080 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.885 1.887 make_images_pack 222 9.7 1.622 1.879 1.625 1.882 rebuild_ks_matrix 3 7.3 0.000 0.000 1.867 1.869 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.867 1.869 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.124000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2241.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 3.374440E+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 0.027 0.047 97.432 97.433 qs_energies 1 2.0 0.000 0.000 96.863 96.870 ls_scf 1 3.0 0.000 0.000 95.197 95.204 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.909 79.091 ls_scf_main 1 4.0 0.000 0.000 58.743 58.745 multiply_cannon 111 7.7 0.046 0.114 52.225 56.654 density_matrix_trs4 2 5.0 0.002 0.003 52.648 52.779 multiply_cannon_loop 111 8.7 0.154 0.167 47.205 49.909 ls_scf_init_scf 1 4.0 0.000 0.000 33.216 33.218 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.007 32.072 mp_waitall_1 6369 11.0 22.971 30.063 22.971 30.063 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.478 29.490 multiply_cannon_multrec 1332 9.7 14.198 17.173 22.150 25.272 make_m2s 222 7.7 0.007 0.008 21.164 22.562 make_images 222 8.7 3.145 3.605 21.113 22.514 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.272 16.645 make_images_data 222 9.7 0.004 0.004 11.799 13.403 hybrid_alltoall_any 227 10.6 0.799 3.794 11.192 12.867 dbcsr_mm_accdrv_process 3641 10.4 0.354 0.736 7.574 9.030 dbcsr_mm_accdrv_process_sort 3641 11.4 7.091 8.535 7.091 8.535 mp_sum_l 887 5.1 3.926 7.077 3.926 7.077 multiply_cannon_sync_h2d 1332 9.7 5.452 5.957 5.452 5.957 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.072 5.949 mp_irecv_dv 3229 10.9 2.048 5.886 2.048 5.886 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.733 5.668 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.991 5.514 arnoldi_extremal 4 6.8 0.000 0.000 5.334 5.347 arnoldi_normal_ev 4 7.8 0.001 0.005 5.334 5.347 build_subspace 16 8.4 0.014 0.021 4.997 5.003 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.572 4.782 calculate_norms 2376 9.8 4.183 4.562 4.183 4.562 mp_allgather_i34 111 8.7 2.131 4.539 2.131 4.539 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.616 3.921 dbcsr_matrix_vector_mult_local 304 10.0 3.239 3.740 3.242 3.742 dbcsr_sort_data 658 11.4 3.067 3.493 3.067 3.493 ls_scf_post 1 4.0 0.000 0.000 3.238 3.246 dbcsr_special_finalize 555 9.7 0.006 0.007 2.810 3.213 dbcsr_merge_single_wm 555 10.7 0.538 0.661 2.801 3.205 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.003 3.059 ls_scf_store_result 1 5.0 0.000 0.000 2.996 3.046 dbcsr_data_release 10477 10.7 1.608 2.468 1.608 2.468 dbcsr_finalize 304 7.8 0.049 0.061 1.805 2.027 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.433000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2697.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 4.737180E+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 0.051 91.940 91.941 qs_energies 1 2.0 0.000 0.000 91.193 91.198 ls_scf 1 3.0 0.000 0.000 89.240 89.244 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.774 70.981 ls_scf_main 1 4.0 0.000 0.000 56.685 56.685 multiply_cannon 111 7.7 0.081 0.172 52.851 56.299 multiply_cannon_loop 111 8.7 0.089 0.099 50.274 51.677 density_matrix_trs4 2 5.0 0.002 0.003 49.739 49.807 ls_scf_init_scf 1 4.0 0.000 0.000 29.201 29.202 mp_waitall_1 5436 11.0 24.590 28.883 24.590 28.883 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.950 27.983 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.846 25.857 multiply_cannon_multrec 444 9.7 13.652 16.438 20.718 21.979 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.682 15.905 make_m2s 222 7.7 0.005 0.005 13.392 14.325 make_images 222 8.7 2.039 2.473 13.324 14.257 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.549 13.772 hybrid_alltoall_any 227 10.6 0.801 3.797 8.117 9.609 make_images_data 222 9.7 0.003 0.004 8.176 9.608 dbcsr_mm_accdrv_process 3003 10.4 0.350 0.404 6.751 7.888 multiply_cannon_sync_h2d 444 9.7 6.782 7.834 6.782 7.834 dbcsr_mm_accdrv_process_sort 3003 11.4 6.390 7.488 6.390 7.488 arnoldi_extremal 4 6.8 0.000 0.000 5.870 5.883 arnoldi_normal_ev 4 7.8 0.002 0.006 5.870 5.883 build_subspace 16 8.4 0.015 0.019 5.471 5.488 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.362 4.564 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.193 4.393 dbcsr_matrix_vector_mult_local 304 10.0 3.729 4.194 3.731 4.197 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.579 4.036 mp_sum_l 887 5.1 2.530 4.014 2.530 4.014 mp_irecv_dv 1241 11.2 1.564 4.003 1.564 4.003 mp_allgather_i34 111 8.7 1.160 3.739 1.160 3.739 calculate_norms 792 9.8 3.617 3.704 3.617 3.704 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.535 3.593 ls_scf_post 1 4.0 0.000 0.000 3.354 3.359 ls_scf_store_result 1 5.0 0.000 0.000 3.148 3.182 make_images_sizes 222 9.7 0.000 0.000 0.834 3.095 mp_alltoall_i44 222 10.7 0.834 3.094 0.834 3.094 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.783 3.059 dbcsr_finalize 304 7.8 0.062 0.077 2.197 2.307 dbcsr_merge_all 275 8.9 0.479 0.526 2.055 2.144 dbcsr_data_new 4608 9.7 1.785 2.113 1.785 2.113 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.028 2.029 rebuild_ks_matrix 3 7.3 0.000 0.000 1.996 1.996 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.996 1.996 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 1.937 1.937 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.941000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3745.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 8.832631E+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 0.073 0.090 105.391 105.392 qs_energies 1 2.0 0.000 0.000 104.052 104.062 ls_scf 1 3.0 0.000 0.000 101.074 101.083 dbcsr_multiply_generic 111 6.7 0.024 0.027 74.682 74.810 ls_scf_main 1 4.0 0.000 0.000 63.719 63.719 density_matrix_trs4 2 5.0 0.002 0.003 54.662 54.721 multiply_cannon 111 7.7 0.151 0.243 48.565 50.967 multiply_cannon_loop 111 8.7 0.098 0.101 45.441 46.001 ls_scf_init_scf 1 4.0 0.000 0.000 33.575 33.575 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.021 32.046 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.309 29.314 mp_waitall_1 4527 11.1 21.717 25.685 21.717 25.685 make_m2s 222 7.7 0.005 0.005 22.516 23.708 make_images 222 8.7 3.579 3.876 22.408 23.598 multiply_cannon_multrec 444 9.7 17.895 18.448 22.528 23.100 hybrid_alltoall_any 227 10.6 1.656 3.636 12.742 15.537 make_images_data 222 9.7 0.003 0.004 12.939 15.155 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.187 10.475 multiply_cannon_sync_h2d 444 9.7 8.790 8.850 8.790 8.850 arnoldi_extremal 4 6.8 0.000 0.000 7.391 7.399 arnoldi_normal_ev 4 7.8 0.003 0.009 7.391 7.399 build_subspace 16 8.4 0.026 0.036 6.841 6.849 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.491 5.633 dbcsr_matrix_vector_mult_local 304 10.0 5.092 5.377 5.095 5.380 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.195 5.291 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.859 5.128 dbcsr_mm_accdrv_process 1814 10.4 0.220 0.360 4.445 4.581 dbcsr_mm_accdrv_process_sort 1814 11.4 4.112 4.241 4.112 4.241 mp_allgather_i34 111 8.7 1.227 3.898 1.227 3.898 ls_scf_post 1 4.0 0.000 0.000 3.780 3.790 ls_scf_store_result 1 5.0 0.000 0.000 3.503 3.541 make_images_sizes 222 9.7 0.000 0.000 1.448 3.519 mp_alltoall_i44 222 10.7 1.447 3.518 1.447 3.518 calculate_norms 792 9.8 3.230 3.277 3.230 3.277 dbcsr_finalize 304 7.8 0.082 0.089 3.080 3.162 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.948 2.948 dbcsr_merge_all 275 8.9 0.884 0.913 2.865 2.942 dbcsr_complete_redistribute 5 7.6 1.439 1.482 2.745 2.867 matrix_ls_to_qs 2 6.0 0.000 0.000 2.406 2.542 dbcsr_sort_data 325 11.1 2.443 2.495 2.443 2.495 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.457 2.459 rebuild_ks_matrix 3 7.3 0.000 0.000 2.391 2.394 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.000 2.391 2.394 dbcsr_data_new 6591 9.6 1.830 2.326 1.830 2.326 dbcsr_new_transposed 4 7.5 0.242 0.251 2.285 2.297 dbcsr_frobenius_norm 74 6.6 2.055 2.138 2.198 2.229 dbcsr_add_d 103 6.2 0.000 0.000 2.125 2.200 dbcsr_add_anytype 103 7.2 0.859 0.890 2.124 2.200 dbcsr_data_release 12724 10.6 1.980 2.195 1.980 2.195 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.392000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7054.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9e23118c5f2bdf14bad76c6d1e1a6705ce46993c_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 594.132992E+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 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 0.022 0.117 226.222 226.223 qs_mol_dyn_low 1 2.0 0.004 0.028 225.028 225.082 qs_forces 5 3.8 0.005 0.046 224.868 224.870 qs_energies 5 4.8 0.048 0.103 221.910 221.944 scf_env_do_scf 5 5.8 0.000 0.002 206.421 206.424 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 180.097 180.098 qs_scf_new_mos 105 7.6 0.000 0.001 140.799 140.985 qs_scf_loop_do_ot 105 8.6 0.001 0.001 140.798 140.984 dbcsr_multiply_generic 1445 12.2 0.134 0.166 133.592 134.051 ot_scf_mini 105 9.6 0.003 0.004 130.928 131.104 multiply_cannon 1445 13.2 0.273 0.283 114.825 117.123 multiply_cannon_loop 1445 14.2 2.825 2.989 113.075 114.467 velocity_verlet 4 3.0 0.001 0.032 105.733 105.735 ot_mini 105 10.6 0.001 0.001 59.846 59.963 multiply_cannon_multrec 69360 15.2 29.613 34.824 39.620 44.438 mp_waitall_1 488190 16.1 34.849 42.303 34.849 42.303 qs_ot_get_p 112 10.4 0.001 0.002 41.098 41.417 qs_ot_get_derivative 55 11.6 0.001 0.001 38.070 38.201 multiply_cannon_sync_h2d 69360 15.2 29.110 32.969 29.110 32.969 multiply_cannon_metrocomm3 69360 15.2 0.198 0.209 25.667 32.898 qs_ot_p2m_diag 40 11.0 0.020 0.031 30.153 30.306 rebuild_ks_matrix 110 8.4 0.000 0.000 28.676 28.853 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.042 28.675 28.853 cp_dbcsr_syevd 40 12.0 0.002 0.002 27.021 27.022 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.338 26.495 init_scf_loop 7 6.6 0.000 0.015 26.295 26.296 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.070 23.284 apply_single 62 13.6 0.000 0.000 23.070 23.284 cp_fm_syevd 40 13.0 0.000 0.001 21.976 22.127 prepare_preconditioner 7 7.6 0.000 0.000 21.490 21.521 make_preconditioner 7 8.6 0.000 0.003 21.490 21.521 ot_new_cg_direction 55 11.6 0.001 0.001 21.079 21.080 cp_fm_redistribute_end 40 14.0 8.665 17.289 8.669 17.291 cp_fm_syevd_base 40 14.0 8.614 17.239 8.614 17.239 qs_rho_update_rho_low 110 7.6 0.001 0.002 16.291 16.584 calculate_rho_elec 110 8.6 0.030 0.033 16.290 16.584 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.772 14.978 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.299 14.368 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.778 13.868 mp_sum_l 4764 12.2 11.956 12.754 11.956 12.754 init_scf_run 5 5.8 0.000 0.001 12.674 12.675 scf_env_initial_rho_setup 5 6.8 0.002 0.003 12.674 12.675 pw_transfer 1645 12.4 0.080 0.106 11.546 11.806 calculate_dm_sparse 110 9.5 0.000 0.001 11.500 11.727 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.406 11.669 density_rs2pw 110 9.6 0.005 0.006 10.750 11.111 dbcsr_mm_accdrv_process 154766 15.8 6.232 6.411 9.875 10.673 qs_vxc_create 110 10.4 0.002 0.005 10.363 10.413 fft_wrap_pw1pw2_240 915 15.0 0.829 0.924 10.040 10.273 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.179 10.252 cp_fm_cholesky_invert 7 10.6 9.982 9.990 9.982 9.990 check_diag 80 13.5 8.593 8.885 9.376 9.527 acc_transpose_blocks 69360 15.2 0.354 0.371 7.946 8.680 fft3d_pb 915 16.0 2.403 2.696 8.228 8.514 multiply_cannon_metrocomm1 69360 15.2 0.096 0.102 4.788 8.351 sum_up_and_integrate 60 10.3 0.001 0.003 8.314 8.332 integrate_v_rspace 60 11.3 0.002 0.024 8.297 8.315 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 8.085 8.101 transfer_rs2pw 445 10.6 0.007 0.008 7.503 7.875 calculate_first_density_matrix 1 7.0 0.001 0.020 7.814 7.829 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.638 7.694 xc_rho_set_and_dset_create 110 12.4 0.075 0.120 7.266 7.551 make_m2s 2890 13.2 0.078 0.086 6.332 7.006 xc_vxc_pw_create 60 11.3 0.039 0.051 6.933 6.983 make_full_single_inverse 7 9.6 0.001 0.001 6.922 6.953 make_images 2890 14.2 0.237 0.257 6.225 6.897 xc_pw_derive 510 13.4 0.005 0.007 6.146 6.230 acc_transpose_blocks_kernels 69360 16.2 0.851 0.895 5.202 5.835 mp_alltoall_z22v 2340 17.7 5.246 5.496 5.246 5.496 mp_waitany 7680 13.5 4.480 5.072 4.480 5.072 jit_kernel_transpose 5 15.0 4.352 4.947 4.352 4.947 potential_pw2rs 60 12.3 0.002 0.003 4.556 4.575 wfi_extrapolate 5 7.8 0.000 0.001 4.555 4.555 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=226.223000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=563.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 9e23118c5f2bdf14bad76c6d1e1a6705ce46993c Summary: empty Status: OK