=== 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: e1b64ec72d8c911e7967845d04dd1ccfaacfc398 ################# 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/01 job id: 50107750 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/02 job id: 50107751 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/03 job id: 50107752 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/04 job id: 50107753 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/05 job id: 50107754 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/06 job id: 50107755 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/07 job id: 50107756 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/08 job id: 50107757 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/09 job id: 50107758 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/10 job id: 50107759 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/11 job id: 50107760 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/12 job id: 50107761 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/13 job id: 50107762 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/14 job id: 50107763 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/15 job id: 50107764 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/16 job id: 50107765 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/17 job id: 50107766 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/18 job id: 50107767 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/19 job id: 50107768 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/20 job id: 50107769 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/21 job id: 50107770 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/22 job id: 50107771 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/23 job id: 50107772 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/24 job id: 50107773 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/25 job id: 50107774 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/26 job id: 50107776 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/27 job id: 50107777 --- 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/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 0.034 133.042 133.043 farming_run 1 2.0 132.543 132.544 133.015 133.018 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.509093E+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 0.005 0.026 114.084 114.084 qs_energies 1 2.0 0.000 0.000 113.853 113.856 mp2_main 1 3.0 0.000 0.000 111.794 111.797 mp2_gpw_main 1 4.0 0.027 0.033 110.814 110.816 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 91.932 92.055 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.577 54.698 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.166 40.983 45.797 get_2c_integrals 1 6.0 0.008 0.009 36.466 37.183 integrate_v_rspace 273 8.0 0.440 0.454 24.568 29.050 pw_transfer 6555 10.6 0.382 0.398 26.747 26.958 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 25.355 25.522 grid_integrate_task_list 273 9.0 20.478 25.378 20.478 25.378 fft_wrap_pw1pw2_100 2178 12.4 0.028 0.029 22.893 23.093 compute_2c_integrals 1 7.0 0.002 0.002 18.989 18.991 rpa_ri_compute_en 1 5.0 0.019 0.021 18.778 18.908 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.572 18.710 mp2_eri_2c_integrate_gpw 1 9.0 2.387 2.421 18.570 18.709 cp_fm_cholesky_decompose 12 8.2 17.403 18.094 17.403 18.094 cholesky_decomp 1 7.0 0.000 0.000 16.326 17.016 fft3d_s 5443 13.4 16.073 16.271 16.094 16.291 ao_to_mo_and_store_B_mult_1 272 7.0 10.768 15.316 10.768 15.316 calculate_wavefunction 272 8.0 5.415 5.501 12.314 13.016 rpa_num_int 1 6.0 0.000 0.001 10.590 10.599 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.576 10.578 calc_mat_Q 8 8.0 0.000 0.000 9.450 9.523 contract_S_to_Q 8 9.0 0.000 0.000 8.873 8.948 calc_potential_gpw 544 9.5 0.004 0.005 8.156 8.632 parallel_gemm_fm 14 9.1 0.000 0.000 8.458 8.550 parallel_gemm_fm_cosma 14 10.1 8.458 8.550 8.458 8.550 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.056 8.278 potential_pw2rs 545 10.0 0.107 0.108 7.546 8.124 create_integ_mat 1 6.0 0.015 0.028 7.972 7.982 collocate_single_gaussian 272 10.0 0.039 0.041 7.332 7.607 array2fm 1 7.0 0.000 0.000 6.809 7.411 pw_scatter_s 2720 13.7 4.308 4.529 4.308 4.529 pw_gather_s 2722 13.2 3.455 3.860 3.455 3.860 array2fm_buffer_send 1 8.0 3.059 3.642 3.059 3.642 pw_poisson_solve 545 10.5 1.130 1.195 2.225 2.361 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=110.813367, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2825.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.035 421.226 421.227 farming_run 1 2.0 420.475 420.479 421.191 421.193 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.244189E+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 2391618. 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 0.012 0.034 215.868 215.869 qs_energies 1 2.0 0.000 0.000 215.643 215.649 scf_env_do_scf 1 3.0 0.000 0.000 113.910 113.910 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.020 113.027 rebuild_ks_matrix 4 6.0 0.000 0.000 113.019 113.026 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 113.019 113.026 hfx_ks_matrix 4 8.0 0.001 0.001 112.671 112.676 integrate_four_center 4 9.0 0.154 0.457 112.670 112.675 integrate_four_center_main 4 10.0 0.112 0.640 101.721 104.595 integrate_four_center_bin 263 11.0 101.609 104.579 101.609 104.579 mp2_main 1 3.0 0.002 0.012 101.439 101.445 mp2_gpw_main 1 4.0 0.046 0.070 100.461 100.472 init_scf_loop 1 4.0 0.000 0.000 96.170 96.170 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 73.877 74.950 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.726 54.798 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.164 41.405 46.465 integrate_v_rspace 95 8.0 0.398 0.576 27.888 32.805 pw_transfer 2240 10.6 0.143 0.172 29.496 29.828 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.499 28.879 grid_integrate_task_list 95 9.0 23.301 28.426 23.301 28.426 mp2_ri_gpw_compute_en 1 5.0 0.055 0.062 26.438 28.152 ao_to_mo_and_store_B_mult_1 91 7.0 10.655 27.717 10.655 27.717 fft_wrap_pw1pw2_100 730 12.4 0.013 0.014 26.266 26.608 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.850 1.892 24.719 24.730 get_2c_integrals 1 6.0 0.000 0.000 20.063 20.087 compute_2c_integrals 1 7.0 0.002 0.003 19.045 19.047 compute_2c_integrals_loop_lm 1 8.0 0.001 0.003 18.638 18.913 mp2_eri_2c_integrate_gpw 1 9.0 1.738 1.857 18.637 18.912 fft3d_s 1823 13.4 18.490 18.791 18.503 18.804 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.738 17.738 calculate_wavefunction 91 8.0 2.008 2.039 9.611 9.858 mp2_ri_gpw_compute_en_expansio 172 7.0 0.552 0.583 8.767 9.172 potential_pw2rs 186 10.0 0.033 0.035 8.467 9.025 local_gemm 172 8.0 8.215 8.623 8.215 8.623 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.000 0.001 8.172 8.555 mp2_ri_gpw_compute_en_comm 22 7.0 0.503 0.520 7.725 8.127 calc_potential_gpw 182 9.5 0.002 0.002 7.878 8.071 collocate_single_gaussian 91 10.0 0.016 0.021 7.757 7.990 integrate_four_center_load 4 10.0 0.000 0.000 6.768 6.772 hfx_load_balance 1 11.0 0.000 0.000 6.768 6.772 mp2_ri_gpw_compute_en_ener 172 7.0 6.332 6.422 6.332 6.422 mp_sendrecv_dm3 2068 8.0 5.748 6.139 5.748 6.139 mp_sync 37 10.5 2.917 6.068 2.917 6.068 pw_gather_s 912 13.2 4.487 4.998 4.487 4.998 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.469851, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1503.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 454.627328E+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 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.024 54.788 54.789 qs_mol_dyn_low 1 2.0 0.003 0.004 54.590 54.597 qs_forces 11 3.9 0.002 0.002 54.526 54.526 qs_energies 11 4.9 0.002 0.002 53.019 53.038 scf_env_do_scf 11 5.9 0.001 0.001 46.912 46.912 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 44.930 44.930 qs_scf_new_mos 108 7.5 0.000 0.001 35.014 35.302 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.014 35.302 dbcsr_multiply_generic 2286 12.5 0.094 0.097 34.552 35.001 ot_scf_mini 108 9.5 0.002 0.002 33.369 33.577 multiply_cannon 2286 13.5 0.186 0.193 26.716 28.301 multiply_cannon_loop 2286 14.5 1.792 1.897 26.075 27.678 velocity_verlet 10 3.0 0.001 0.002 26.233 26.234 ot_mini 108 10.5 0.001 0.001 20.388 20.637 qs_ot_get_derivative 108 11.5 0.001 0.001 17.349 17.545 mp_waitall_1 245248 16.5 9.056 15.277 9.056 15.277 multiply_cannon_metrocomm3 54864 15.5 0.072 0.077 6.466 13.127 multiply_cannon_multrec 54864 15.5 3.612 5.577 7.551 10.949 qs_ot_get_p 119 10.4 0.001 0.001 8.235 8.543 rebuild_ks_matrix 119 8.3 0.000 0.000 7.816 7.951 qs_ks_build_kohn_sham_matrix 119 9.3 0.019 0.036 7.816 7.951 mp_sum_l 7287 12.8 5.941 7.600 5.941 7.600 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.861 6.985 multiply_cannon_sync_h2d 54864 15.5 5.178 6.498 5.178 6.498 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.832 6.272 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.873 6.027 dbcsr_mm_accdrv_process 76910 16.1 1.840 2.898 3.854 5.516 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.404 5.479 init_scf_run 11 5.9 0.000 0.001 4.855 4.856 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.855 4.855 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.589 4.589 sum_up_and_integrate 119 10.3 0.001 0.002 4.398 4.404 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.402 4.402 integrate_v_rspace 119 11.3 0.002 0.002 4.387 4.395 cp_fm_redistribute_end 50 14.0 2.244 4.379 2.249 4.381 cp_fm_diag_elpa_base 50 14.0 2.126 4.275 2.130 4.281 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.059 4.175 calculate_rho_elec 119 8.7 0.011 0.016 4.058 4.174 multiply_cannon_metrocomm1 54864 15.5 0.055 0.059 1.998 3.376 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.926 3.131 apply_single 119 13.6 0.000 0.000 2.925 3.131 calculate_dm_sparse 119 9.5 0.000 0.000 2.866 2.998 ot_diis_step 108 11.5 0.006 0.006 2.735 2.735 acc_transpose_blocks 54864 15.5 0.220 0.242 2.189 2.705 calculate_first_density_matrix 1 7.0 0.000 0.000 2.689 2.697 jit_kernel_multiply 13 15.8 1.947 2.653 1.947 2.653 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.506 2.592 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.334 2.336 grid_integrate_task_list 119 12.3 2.057 2.149 2.057 2.149 density_rs2pw 119 9.7 0.004 0.004 2.055 2.141 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.066 2.130 wfi_extrapolate 11 7.9 0.001 0.001 2.094 2.094 mp_sum_d 4137 12.0 1.390 1.966 1.390 1.966 init_scf_loop 11 6.9 0.000 0.001 1.965 1.965 potential_pw2rs 119 12.3 0.004 0.004 1.740 1.751 pw_transfer 1439 11.6 0.051 0.056 1.591 1.660 make_m2s 4572 13.5 0.053 0.056 1.576 1.628 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.516 1.585 make_images 4572 14.5 0.132 0.139 1.494 1.545 mp_alltoall_d11v 2130 13.8 1.306 1.452 1.306 1.452 acc_transpose_blocks_sync 164592 16.5 1.185 1.426 1.185 1.426 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.383 1.413 grid_collocate_task_list 119 9.7 1.355 1.408 1.355 1.408 mp_waitany 12084 13.8 1.253 1.395 1.253 1.395 transfer_rs2pw 487 10.6 0.005 0.006 1.302 1.377 transfer_pw2rs 487 13.2 0.006 0.007 1.288 1.303 fft3d_ps 1201 14.6 0.368 0.473 1.229 1.294 fft_wrap_pw1pw2_140 487 13.2 0.108 0.114 1.175 1.247 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.785 1.188 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.789000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=433.727273, yerr=0.445362 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 490.422272E+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 1031941. 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 0.007 0.027 39.744 39.745 qs_mol_dyn_low 1 2.0 0.003 0.003 39.577 39.584 qs_forces 11 3.9 0.002 0.003 39.516 39.517 qs_energies 11 4.9 0.002 0.002 37.844 37.849 scf_env_do_scf 11 5.9 0.000 0.001 32.298 32.299 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.811 29.811 dbcsr_multiply_generic 2286 12.5 0.102 0.105 22.398 22.759 qs_scf_new_mos 108 7.5 0.001 0.001 21.073 21.297 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.072 21.296 ot_scf_mini 108 9.5 0.002 0.003 20.144 20.306 velocity_verlet 10 3.0 0.002 0.003 18.616 18.617 multiply_cannon 2286 13.5 0.214 0.222 17.056 18.547 multiply_cannon_loop 2286 14.5 1.178 1.246 15.722 17.375 ot_mini 108 10.5 0.001 0.001 12.342 12.571 mp_waitall_1 200699 16.5 5.690 10.756 5.690 10.756 qs_ot_get_derivative 108 11.5 0.001 0.001 9.966 10.129 multiply_cannon_metrocomm3 27432 15.5 0.071 0.073 4.087 9.468 multiply_cannon_multrec 27432 15.5 1.810 4.109 6.126 8.874 rebuild_ks_matrix 119 8.3 0.000 0.000 6.825 6.978 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.824 6.978 dbcsr_mm_accdrv_process 47894 16.0 3.557 5.839 4.236 6.347 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.008 6.146 qs_ot_get_p 119 10.4 0.001 0.001 4.869 5.089 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.721 4.635 init_scf_run 11 5.9 0.000 0.001 4.274 4.274 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.274 4.274 mp_sum_l 7287 12.8 2.154 4.140 2.154 4.140 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.083 4.069 apply_single 119 13.6 0.000 0.000 3.083 4.069 sum_up_and_integrate 119 10.3 0.001 0.001 3.742 3.747 integrate_v_rspace 119 11.3 0.002 0.003 3.728 3.733 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.689 3.722 calculate_rho_elec 119 8.7 0.021 0.024 3.689 3.721 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.166 3.184 make_m2s 4572 13.5 0.051 0.053 2.731 3.168 make_images 4572 14.5 0.205 0.244 2.641 3.080 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.760 2.761 calculate_first_density_matrix 1 7.0 0.000 0.000 2.587 2.591 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.400 2.485 init_scf_loop 11 6.9 0.000 0.000 2.467 2.467 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.402 2.402 cp_fm_redistribute_end 50 14.0 1.214 2.369 1.219 2.373 ot_diis_step 108 11.5 0.011 0.011 2.327 2.328 cp_fm_diag_elpa_base 50 14.0 1.120 2.267 1.149 2.313 multiply_cannon_sync_h2d 27432 15.5 1.691 2.197 1.691 2.197 calculate_dm_sparse 119 9.5 0.000 0.001 2.092 2.169 density_rs2pw 119 9.7 0.004 0.004 1.981 2.072 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.054 2.056 grid_integrate_task_list 119 12.3 1.841 1.938 1.841 1.938 acc_transpose_blocks 27432 15.5 0.113 0.119 1.601 1.935 pw_transfer 1439 11.6 0.064 0.069 1.897 1.931 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.861 1.900 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.806 1.842 make_images_data 4572 15.5 0.047 0.053 1.293 1.682 wfi_extrapolate 11 7.9 0.001 0.001 1.624 1.624 prepare_preconditioner 11 7.9 0.000 0.000 1.573 1.600 make_preconditioner 11 8.9 0.000 0.000 1.573 1.600 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.473 1.529 potential_pw2rs 119 12.3 0.006 0.006 1.484 1.493 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.134 1.477 fft3d_ps 1201 14.6 0.521 0.580 1.437 1.468 fft_wrap_pw1pw2_140 487 13.2 0.128 0.131 1.416 1.450 mp_alltoall_d11v 2130 13.8 1.242 1.438 1.242 1.438 jit_kernel_multiply 7 16.3 0.619 1.357 0.619 1.357 grid_collocate_task_list 119 9.7 1.291 1.341 1.291 1.341 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.275 1.285 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.218 1.266 transfer_rs2pw 487 10.6 0.005 0.005 1.118 1.190 mp_allgather_i34 2286 14.5 0.739 1.183 0.739 1.183 mp_sum_d 4137 12.0 0.613 1.108 0.613 1.108 qs_energies_init_hamiltonians 11 5.9 0.015 0.024 1.001 1.002 transfer_pw2rs 487 13.2 0.004 0.006 0.931 0.939 acc_transpose_blocks_sync 82296 16.5 0.807 0.937 0.807 0.937 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.875 0.890 acc_transpose_blocks_kernels 27432 16.5 0.186 0.274 0.657 0.867 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.745000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=466.181818, yerr=1.991718 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 524.218368E+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 0.034 0.083 33.225 33.229 qs_mol_dyn_low 1 2.0 0.003 0.003 32.742 32.797 qs_forces 11 3.9 0.003 0.011 32.675 32.676 qs_energies 11 4.9 0.002 0.006 30.990 30.992 scf_env_do_scf 11 5.9 0.001 0.002 26.071 26.072 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 23.597 23.599 dbcsr_multiply_generic 2286 12.5 0.097 0.100 17.026 17.091 qs_scf_new_mos 108 7.5 0.001 0.001 15.776 15.789 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.775 15.788 velocity_verlet 10 3.0 0.001 0.002 15.401 15.402 ot_scf_mini 108 9.5 0.002 0.003 15.002 15.019 multiply_cannon 2286 13.5 0.194 0.201 13.615 14.393 multiply_cannon_loop 2286 14.5 0.855 0.928 12.837 13.583 ot_mini 108 10.5 0.001 0.001 9.212 9.232 qs_ot_get_derivative 108 11.5 0.001 0.001 7.699 7.718 multiply_cannon_multrec 18288 15.5 1.867 2.855 6.945 7.331 rebuild_ks_matrix 119 8.3 0.000 0.000 6.128 6.145 qs_ks_build_kohn_sham_matrix 119 9.3 0.015 0.035 6.128 6.145 dbcsr_mm_accdrv_process 38222 16.0 4.890 5.706 4.985 5.773 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.347 5.363 mp_waitall_1 158411 16.6 2.692 3.928 2.692 3.928 init_scf_run 11 5.9 0.000 0.001 3.676 3.677 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.676 3.676 sum_up_and_integrate 119 10.3 0.001 0.003 3.647 3.653 integrate_v_rspace 119 11.3 0.003 0.003 3.634 3.642 qs_ot_get_p 119 10.4 0.001 0.002 3.541 3.568 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.867 3.484 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.338 3.345 calculate_rho_elec 119 8.7 0.031 0.032 3.337 3.345 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.443 2.562 init_scf_loop 11 6.9 0.001 0.003 2.456 2.460 calculate_first_density_matrix 1 7.0 0.001 0.005 2.436 2.437 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.373 2.380 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.035 2.353 apply_single 119 13.6 0.000 0.000 2.035 2.352 cp_dbcsr_syevd 50 12.0 0.003 0.005 2.068 2.068 make_m2s 4572 13.5 0.044 0.045 1.863 2.002 density_rs2pw 119 9.7 0.004 0.004 1.886 1.968 pw_transfer 1439 11.6 0.065 0.068 1.912 1.929 make_images 4572 14.5 0.191 0.205 1.777 1.917 grid_integrate_task_list 119 12.3 1.803 1.904 1.803 1.904 calculate_dm_sparse 119 9.5 0.000 0.001 1.864 1.876 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.819 1.832 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.798 1.809 cp_fm_diag_elpa_base 50 14.0 1.773 1.785 1.795 1.805 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.758 1.765 acc_transpose_blocks 18288 15.5 0.078 0.080 1.626 1.755 prepare_preconditioner 11 7.9 0.000 0.000 1.674 1.678 make_preconditioner 11 8.9 0.000 0.001 1.674 1.678 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.651 1.653 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.534 1.618 mp_sum_l 7287 12.8 1.189 1.572 1.189 1.572 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.502 1.507 ot_diis_step 108 11.5 0.011 0.012 1.491 1.491 fft_wrap_pw1pw2_140 487 13.2 0.178 0.182 1.450 1.460 potential_pw2rs 119 12.3 0.007 0.008 1.434 1.445 fft3d_ps 1201 14.6 0.547 0.569 1.375 1.388 grid_collocate_task_list 119 9.7 1.234 1.329 1.234 1.329 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.188 1.192 wfi_extrapolate 11 7.9 0.001 0.001 1.188 1.188 multiply_cannon_sync_h2d 18288 15.5 1.026 1.187 1.026 1.187 transfer_rs2pw 487 10.6 0.005 0.005 1.023 1.107 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.968 0.984 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.966 0.967 make_images_data 4572 15.5 0.046 0.050 0.792 0.945 transfer_pw2rs 487 13.2 0.004 0.004 0.896 0.904 hybrid_alltoall_any 4725 16.4 0.058 0.117 0.690 0.877 acc_transpose_blocks_sync 54864 16.5 0.738 0.870 0.738 0.870 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.853 0.855 mp_alltoall_d11v 2130 13.8 0.713 0.834 0.713 0.834 acc_transpose_blocks_kernels 18288 16.5 0.216 0.221 0.791 0.818 jit_kernel_multiply 4 15.7 0.036 0.765 0.036 0.765 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.676 0.763 mp_alltoall_z22v 1201 16.6 0.672 0.745 0.672 0.745 cp_fm_cholesky_invert 11 10.9 0.723 0.727 0.723 0.727 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.229000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=498.363636, yerr=2.672171 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 559.435776E+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 0.026 36.793 36.796 qs_mol_dyn_low 1 2.0 0.003 0.003 36.628 36.636 qs_forces 11 3.9 0.002 0.002 36.541 36.542 qs_energies 11 4.9 0.002 0.003 34.828 34.833 scf_env_do_scf 11 5.9 0.000 0.001 29.688 29.689 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 26.297 26.297 dbcsr_multiply_generic 2286 12.5 0.103 0.105 20.228 20.402 velocity_verlet 10 3.0 0.002 0.002 18.510 18.513 qs_scf_new_mos 108 7.5 0.001 0.001 18.186 18.243 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.185 18.242 ot_scf_mini 108 9.5 0.002 0.003 17.151 17.210 multiply_cannon 2286 13.5 0.218 0.230 16.423 16.860 multiply_cannon_loop 2286 14.5 1.555 1.627 15.459 15.765 ot_mini 108 10.5 0.001 0.001 10.668 10.734 multiply_cannon_multrec 27432 15.5 2.453 3.107 9.055 9.326 qs_ot_get_derivative 108 11.5 0.001 0.001 8.806 8.865 dbcsr_mm_accdrv_process 47916 15.9 5.857 7.625 6.500 7.809 rebuild_ks_matrix 119 8.3 0.000 0.000 6.356 6.402 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.355 6.401 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.627 5.668 init_scf_run 11 5.9 0.000 0.001 3.805 3.805 scf_env_initial_rho_setup 11 6.9 0.003 0.004 3.805 3.805 qs_ot_get_p 119 10.4 0.001 0.001 3.624 3.697 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.040 3.536 sum_up_and_integrate 119 10.3 0.001 0.001 3.519 3.528 integrate_v_rspace 119 11.3 0.003 0.003 3.508 3.516 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.368 3.406 calculate_rho_elec 119 8.7 0.040 0.046 3.368 3.405 init_scf_loop 11 6.9 0.000 0.000 3.371 3.372 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.197 2.673 apply_single 119 13.6 0.000 0.000 2.197 2.673 prepare_preconditioner 11 7.9 0.000 0.000 2.540 2.547 make_preconditioner 11 8.9 0.000 0.000 2.540 2.547 acc_transpose_blocks 27432 15.5 0.119 0.124 2.453 2.527 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.147 2.478 mp_waitall_1 137007 16.6 1.971 2.438 1.971 2.438 calculate_first_density_matrix 1 7.0 0.000 0.001 2.413 2.417 make_m2s 4572 13.5 0.053 0.055 2.282 2.408 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.341 2.369 make_images 4572 14.5 0.273 0.334 2.174 2.299 calculate_dm_sparse 119 9.5 0.000 0.000 2.172 2.227 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.206 2.214 pw_transfer 1439 11.6 0.064 0.068 1.995 2.030 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.903 1.941 density_rs2pw 119 9.7 0.004 0.004 1.817 1.924 grid_integrate_task_list 119 12.3 1.834 1.908 1.834 1.908 jit_kernel_multiply 10 15.7 0.575 1.890 0.575 1.890 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.889 1.890 ot_diis_step 108 11.5 0.012 0.012 1.821 1.821 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.799 1.800 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.717 1.729 mp_sum_l 7287 12.8 1.089 1.627 1.089 1.627 fft_wrap_pw1pw2_140 487 13.2 0.227 0.234 1.582 1.621 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.543 1.552 cp_fm_diag_elpa_base 50 14.0 1.510 1.527 1.541 1.550 acc_transpose_blocks_sync 82296 16.5 1.465 1.524 1.465 1.524 fft3d_ps 1201 14.6 0.591 0.650 1.392 1.415 multiply_cannon_metrocomm3 27432 15.5 0.039 0.041 0.923 1.370 grid_collocate_task_list 119 9.7 1.249 1.340 1.249 1.340 wfi_extrapolate 11 7.9 0.001 0.001 1.335 1.335 potential_pw2rs 119 12.3 0.009 0.009 1.315 1.317 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.286 1.304 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.266 1.276 cp_fm_upper_to_full 72 14.2 0.817 1.167 0.817 1.167 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.072 1.073 dbcsr_complete_redistribute 329 12.2 0.119 0.148 0.750 1.021 transfer_rs2pw 487 10.6 0.004 0.005 0.893 1.007 make_images_data 4572 15.5 0.047 0.052 0.851 0.976 hybrid_alltoall_any 4725 16.4 0.066 0.158 0.755 0.936 acc_transpose_blocks_kernels 27432 16.5 0.269 0.275 0.841 0.917 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.876 mp_alltoall_d11v 2130 13.8 0.767 0.873 0.767 0.873 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.563 0.829 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.817 0.822 transfer_pw2rs 487 13.2 0.004 0.004 0.744 0.746 cp_fm_cholesky_invert 11 10.9 0.738 0.742 0.738 0.742 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.796000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.090909, yerr=3.449757 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 634.052608E+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 0.014 0.046 28.834 28.849 qs_mol_dyn_low 1 2.0 0.003 0.003 28.626 28.634 qs_forces 11 3.9 0.004 0.014 28.568 28.569 qs_energies 11 4.9 0.002 0.002 26.861 26.865 scf_env_do_scf 11 5.9 0.000 0.001 22.026 22.026 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.552 19.553 velocity_verlet 10 3.0 0.002 0.002 14.766 14.769 dbcsr_multiply_generic 2286 12.5 0.096 0.099 13.202 13.260 qs_scf_new_mos 108 7.5 0.001 0.001 11.841 11.868 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.841 11.867 ot_scf_mini 108 9.5 0.002 0.002 11.130 11.153 multiply_cannon 2286 13.5 0.224 0.234 10.612 11.106 multiply_cannon_loop 2286 14.5 0.639 0.662 9.692 9.836 ot_mini 108 10.5 0.001 0.001 6.464 6.493 multiply_cannon_multrec 9144 15.5 1.697 1.924 6.001 6.146 rebuild_ks_matrix 119 8.3 0.000 0.000 5.805 5.828 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 5.805 5.827 qs_ot_get_derivative 108 11.5 0.001 0.001 5.169 5.196 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.164 5.185 dbcsr_mm_accdrv_process 12550 15.8 3.624 4.240 4.195 4.284 sum_up_and_integrate 119 10.3 0.001 0.001 3.443 3.448 integrate_v_rspace 119 11.3 0.003 0.003 3.434 3.438 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.388 3.393 calculate_rho_elec 119 8.7 0.060 0.061 3.387 3.392 init_scf_run 11 5.9 0.000 0.001 3.324 3.324 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.324 3.324 qs_ot_get_p 119 10.4 0.001 0.001 2.783 2.815 init_scf_loop 11 6.9 0.000 0.000 2.454 2.454 calculate_first_density_matrix 1 7.0 0.000 0.001 2.174 2.175 pw_transfer 1439 11.6 0.065 0.067 2.107 2.116 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.015 2.024 mp_waitall_1 115863 16.7 1.494 2.021 1.494 2.021 make_m2s 4572 13.5 0.033 0.034 1.791 1.951 grid_integrate_task_list 119 12.3 1.855 1.948 1.855 1.948 make_images 4572 14.5 0.268 0.299 1.702 1.860 density_rs2pw 119 9.7 0.003 0.003 1.773 1.847 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.806 1.809 calculate_dm_sparse 119 9.5 0.000 0.000 1.761 1.781 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.721 1.740 prepare_preconditioner 11 7.9 0.000 0.000 1.720 1.724 make_preconditioner 11 8.9 0.000 0.000 1.720 1.724 fft_wrap_pw1pw2_140 487 13.2 0.324 0.332 1.671 1.681 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.608 1.635 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.587 1.588 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.567 1.567 acc_transpose_blocks 9144 15.5 0.042 0.044 1.450 1.494 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.427 1.436 fft3d_ps 1201 14.6 0.650 0.665 1.367 1.377 grid_collocate_task_list 119 9.7 1.297 1.372 1.297 1.372 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.347 1.360 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.289 1.297 cp_fm_diag_elpa_base 50 14.0 1.262 1.278 1.287 1.295 qs_energies_init_hamiltonians 11 5.9 0.012 0.012 1.293 1.294 ot_diis_step 108 11.5 0.013 0.014 1.277 1.277 potential_pw2rs 119 12.3 0.010 0.010 1.252 1.254 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.214 1.246 apply_single 119 13.6 0.000 0.000 1.214 1.245 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.147 1.153 wfi_extrapolate 11 7.9 0.001 0.001 1.095 1.095 jit_kernel_multiply 6 15.8 0.531 1.056 0.531 1.056 hybrid_alltoall_any 4725 16.4 0.065 0.175 0.762 1.001 make_images_data 4572 15.5 0.041 0.046 0.773 0.959 mp_alltoall_d11v 2130 13.8 0.806 0.930 0.806 0.930 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.926 transfer_rs2pw 487 10.6 0.004 0.004 0.795 0.875 cp_fm_cholesky_invert 11 10.9 0.848 0.850 0.848 0.850 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.818 0.823 qs_env_update_s_mstruct 11 6.9 0.001 0.002 0.732 0.781 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.762 0.765 acc_transpose_blocks_sync 27432 16.5 0.715 0.759 0.715 0.759 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 0.677 0.685 mp_allgather_i34 2286 14.5 0.234 0.656 0.234 0.656 transfer_pw2rs 487 13.2 0.003 0.004 0.652 0.653 mp_alltoall_z22v 1201 16.6 0.591 0.624 0.591 0.624 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.330 0.623 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 0.418 0.613 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.849000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=597.000000, yerr=10.162319 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 768.049152E+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 0.025 0.078 43.043 43.044 qs_mol_dyn_low 1 2.0 0.003 0.004 42.732 42.741 qs_forces 11 3.9 0.002 0.003 42.670 42.670 qs_energies 11 4.9 0.002 0.002 40.659 40.662 scf_env_do_scf 11 5.9 0.001 0.001 34.992 34.992 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.187 27.189 velocity_verlet 10 3.0 0.002 0.002 24.364 24.370 dbcsr_multiply_generic 2286 12.5 0.105 0.107 19.015 19.190 qs_scf_new_mos 108 7.5 0.001 0.001 17.415 17.515 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.414 17.514 ot_scf_mini 108 9.5 0.002 0.002 16.261 16.364 multiply_cannon 2286 13.5 0.295 0.298 15.002 15.919 multiply_cannon_loop 2286 14.5 0.874 0.932 13.737 14.655 ot_mini 108 10.5 0.001 0.001 9.926 10.043 multiply_cannon_multrec 9144 15.5 3.372 4.730 8.748 8.866 qs_ot_get_derivative 108 11.5 0.001 0.001 7.726 7.829 init_scf_loop 11 6.9 0.000 0.000 7.779 7.782 rebuild_ks_matrix 119 8.3 0.000 0.001 7.326 7.469 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.325 7.468 prepare_preconditioner 11 7.9 0.000 0.000 6.807 6.821 make_preconditioner 11 8.9 0.000 0.000 6.807 6.821 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.588 6.718 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.368 6.700 dbcsr_mm_accdrv_process 12550 15.8 4.647 6.501 5.238 6.535 cp_fm_upper_to_full 72 14.2 3.179 4.562 3.179 4.562 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.250 4.263 calculate_rho_elec 119 8.7 0.117 0.120 4.250 4.263 sum_up_and_integrate 119 10.3 0.001 0.001 3.899 3.905 integrate_v_rspace 119 11.3 0.003 0.003 3.888 3.895 init_scf_run 11 5.9 0.000 0.001 3.615 3.615 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.614 3.614 qs_ot_get_p 119 10.4 0.001 0.001 3.399 3.541 mp_waitall_1 94719 16.7 2.298 3.300 2.298 3.300 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.809 3.218 pw_transfer 1439 11.6 0.068 0.069 3.034 3.038 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.936 2.940 dbcsr_complete_redistribute 329 12.2 0.281 0.285 1.970 2.793 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.292 2.631 apply_single 119 13.6 0.000 0.000 2.292 2.631 make_m2s 4572 13.5 0.037 0.037 2.423 2.579 fft_wrap_pw1pw2_140 487 13.2 0.649 0.653 2.509 2.515 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.667 2.476 make_images 4572 14.5 0.353 0.387 2.301 2.458 density_rs2pw 119 9.7 0.003 0.003 2.242 2.298 calculate_dm_sparse 119 9.5 0.000 0.000 2.229 2.247 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.284 2.237 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.434 2.237 mp_alltoall_i22 627 13.8 1.411 2.235 1.411 2.235 ot_diis_step 108 11.5 0.014 0.015 2.176 2.176 grid_integrate_task_list 119 12.3 2.119 2.161 2.119 2.161 calculate_first_density_matrix 1 7.0 0.000 0.000 2.151 2.153 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.057 2.109 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.008 2.010 acc_transpose_blocks 9144 15.5 0.044 0.045 1.854 1.906 fft3d_ps 1201 14.6 0.867 0.885 1.856 1.859 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.833 1.834 mp_sum_l 7287 12.8 1.037 1.795 1.037 1.795 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.767 1.767 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.651 1.652 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.593 1.631 grid_collocate_task_list 119 9.7 1.550 1.585 1.550 1.585 cp_fm_cholesky_invert 11 10.9 1.438 1.441 1.438 1.441 wfi_extrapolate 11 7.9 0.001 0.001 1.401 1.401 hybrid_alltoall_any 4725 16.4 0.091 0.150 1.134 1.381 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.363 1.364 cp_fm_diag_elpa_base 50 14.0 1.218 1.270 1.362 1.362 potential_pw2rs 119 12.3 0.014 0.014 1.354 1.355 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.308 1.326 make_images_data 4572 15.5 0.046 0.049 1.081 1.279 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.241 1.249 mp_alltoall_d11v 2130 13.8 1.194 1.234 1.194 1.234 acc_transpose_blocks_sync 27432 16.5 1.107 1.152 1.107 1.152 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.098 1.118 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.994 1.023 jit_kernel_multiply 6 15.9 0.561 1.013 0.561 1.013 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.948 0.961 qs_create_task_list 11 7.9 0.001 0.001 0.929 0.940 generate_qs_task_list 11 8.9 0.366 0.388 0.928 0.939 mp_alltoall_z22v 1201 16.6 0.852 0.877 0.852 0.877 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.044000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=720.090909, yerr=14.177155 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 502.484992E+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 1150386. 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 0.012 0.041 82.045 82.047 qs_mol_dyn_low 1 2.0 0.003 0.003 81.701 81.737 qs_forces 11 3.9 0.004 0.010 81.628 81.629 qs_energies 11 4.9 0.002 0.003 78.761 78.777 scf_env_do_scf 11 5.9 0.004 0.025 69.724 69.727 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 64.116 64.116 dbcsr_multiply_generic 2055 12.4 0.106 0.109 51.184 51.555 qs_scf_new_mos 99 7.5 0.000 0.001 47.184 47.315 qs_scf_loop_do_ot 99 8.5 0.000 0.001 47.184 47.315 ot_scf_mini 99 9.5 0.002 0.002 44.855 44.972 multiply_cannon 2055 13.4 0.182 0.187 42.727 43.969 multiply_cannon_loop 2055 14.4 1.763 1.808 41.760 43.061 velocity_verlet 10 3.0 0.001 0.001 42.893 42.893 ot_mini 99 10.5 0.001 0.001 26.516 26.639 qs_ot_get_derivative 99 11.5 0.001 0.001 19.689 19.782 multiply_cannon_multrec 49320 15.4 11.377 12.054 17.592 18.394 rebuild_ks_matrix 110 8.3 0.000 0.001 14.489 14.684 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.028 14.488 14.684 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.709 12.885 mp_waitall_1 220248 16.4 10.613 11.452 10.613 11.452 multiply_cannon_sync_h2d 49320 15.4 9.647 10.549 9.647 10.549 qs_ot_get_p 110 10.4 0.001 0.001 9.811 9.973 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.631 8.152 multiply_cannon_metrocomm3 49320 15.4 0.082 0.085 6.488 7.782 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.258 7.733 apply_single 110 13.6 0.000 0.001 7.258 7.733 sum_up_and_integrate 110 10.3 0.003 0.015 6.992 7.007 integrate_v_rspace 110 11.3 0.003 0.004 6.965 6.988 init_scf_run 11 5.9 0.000 0.001 6.970 6.970 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.969 6.970 qs_ot_p2m_diag 48 11.0 0.013 0.019 6.602 6.626 ot_diis_step 99 11.5 0.006 0.006 6.608 6.608 dbcsr_mm_accdrv_process 87628 16.1 3.083 3.183 6.084 6.400 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.247 6.375 calculate_rho_elec 110 8.6 0.021 0.025 6.247 6.375 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.743 5.743 init_scf_loop 11 6.9 0.000 0.001 5.578 5.582 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.372 5.449 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.173 5.199 cp_fm_diag_elpa_base 48 14.0 5.160 5.186 5.171 5.197 mp_sum_l 6594 12.7 4.124 4.833 4.124 4.833 wfi_extrapolate 11 7.9 0.001 0.001 4.021 4.021 make_m2s 4110 13.4 0.061 0.067 3.873 3.998 make_images 4110 14.4 0.178 0.190 3.777 3.907 calculate_dm_sparse 110 9.5 0.001 0.001 3.704 3.795 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.779 3.783 density_rs2pw 110 9.6 0.004 0.004 3.296 3.550 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.437 3.493 grid_integrate_task_list 110 12.3 3.264 3.433 3.264 3.433 prepare_preconditioner 11 7.9 0.000 0.000 3.364 3.388 make_preconditioner 11 8.9 0.000 0.000 3.364 3.388 multiply_cannon_metrocomm1 49320 15.4 0.064 0.067 2.316 3.347 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.228 3.264 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.132 3.194 pw_transfer 1331 11.6 0.055 0.065 3.066 3.143 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.978 3.058 calculate_first_density_matrix 1 7.0 0.000 0.001 2.850 2.853 jit_kernel_multiply 13 15.9 2.715 2.791 2.715 2.791 acc_transpose_blocks 49320 15.4 0.207 0.214 2.664 2.756 fft_wrap_pw1pw2_140 451 13.1 0.308 0.331 2.543 2.633 potential_pw2rs 110 12.3 0.005 0.006 2.580 2.599 mp_alltoall_d11v 2046 13.8 2.071 2.509 2.071 2.509 fft3d_ps 1111 14.6 0.799 0.889 2.272 2.322 grid_collocate_task_list 110 9.6 2.157 2.290 2.157 2.290 transfer_rs2pw 451 10.6 0.005 0.006 1.865 2.065 mp_waitany 14300 13.8 1.820 2.026 1.820 2.026 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.914 1.936 mp_sum_d 3891 11.9 1.377 1.901 1.377 1.901 make_images_data 4110 15.4 0.043 0.047 1.737 1.855 cp_fm_cholesky_invert 11 10.9 1.828 1.832 1.828 1.832 hybrid_alltoall_any 4261 16.3 0.084 0.486 1.521 1.800 transfer_pw2rs 451 13.1 0.006 0.007 1.737 1.748 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.660 1.682 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.047000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.181818, yerr=2.690663 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 595.943424E+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 3067248. 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 0.019 0.041 68.716 68.718 qs_mol_dyn_low 1 2.0 0.003 0.003 68.418 68.428 qs_forces 11 3.9 0.003 0.003 68.334 68.335 qs_energies 11 4.9 0.001 0.002 64.982 64.986 scf_env_do_scf 11 5.9 0.000 0.001 56.370 56.374 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.891 48.891 dbcsr_multiply_generic 2055 12.4 0.115 0.119 37.893 38.110 velocity_verlet 10 3.0 0.001 0.002 35.877 35.878 qs_scf_new_mos 99 7.5 0.001 0.001 33.114 33.234 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.113 33.233 multiply_cannon 2055 13.4 0.224 0.242 31.304 32.816 ot_scf_mini 99 9.5 0.003 0.003 31.447 31.565 multiply_cannon_loop 2055 14.4 1.153 1.176 30.039 31.213 ot_mini 99 10.5 0.001 0.001 18.361 18.505 multiply_cannon_multrec 24660 15.4 6.965 8.207 14.081 15.528 rebuild_ks_matrix 110 8.3 0.000 0.001 13.519 13.652 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.035 13.518 13.652 qs_ot_get_derivative 99 11.5 0.001 0.001 12.582 12.706 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.858 11.972 mp_waitall_1 176588 16.5 7.450 10.146 7.450 10.146 multiply_cannon_metrocomm3 24660 15.4 0.073 0.076 5.103 7.925 multiply_cannon_sync_h2d 24660 15.4 6.378 7.451 6.378 7.451 init_scf_loop 11 6.9 0.000 0.000 7.445 7.446 dbcsr_mm_accdrv_process 52282 16.1 5.537 6.158 6.946 7.258 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.443 7.168 apply_single 110 13.6 0.000 0.001 6.442 7.168 qs_ot_get_p 110 10.4 0.001 0.001 6.444 6.624 sum_up_and_integrate 110 10.3 0.001 0.002 6.348 6.362 integrate_v_rspace 110 11.3 0.002 0.003 6.322 6.336 init_scf_run 11 5.9 0.000 0.001 6.242 6.243 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.242 6.242 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.859 5.868 calculate_rho_elec 110 8.6 0.039 0.047 5.858 5.867 ot_diis_step 99 11.5 0.010 0.011 5.732 5.732 prepare_preconditioner 11 7.9 0.000 0.000 5.441 5.463 make_preconditioner 11 8.9 0.000 0.000 5.440 5.463 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.707 5.426 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.004 5.163 make_m2s 4110 13.4 0.056 0.060 4.192 4.658 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.552 4.573 make_images 4110 14.4 0.405 0.464 4.080 4.541 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.080 4.080 pw_transfer 1331 11.6 0.065 0.071 3.570 3.714 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.464 3.607 wfi_extrapolate 11 7.9 0.001 0.001 3.501 3.501 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.479 3.494 cp_fm_diag_elpa_base 48 14.0 3.433 3.449 3.476 3.492 density_rs2pw 110 9.6 0.004 0.004 3.115 3.408 grid_integrate_task_list 110 12.3 3.154 3.376 3.154 3.376 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.281 3.348 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.252 3.254 fft_wrap_pw1pw2_140 451 13.1 0.357 0.376 2.960 3.105 calculate_dm_sparse 110 9.5 0.001 0.001 3.007 3.033 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.899 2.954 hybrid_alltoall_any 4261 16.3 0.105 0.454 1.973 2.748 make_images_data 4110 15.4 0.048 0.052 2.271 2.741 fft3d_ps 1111 14.6 1.115 1.346 2.556 2.713 calculate_first_density_matrix 1 7.0 0.000 0.000 2.642 2.645 cp_fm_cholesky_invert 11 10.9 2.520 2.527 2.520 2.527 mp_sum_l 6594 12.7 1.774 2.447 1.774 2.447 grid_collocate_task_list 110 9.6 2.169 2.292 2.169 2.292 potential_pw2rs 110 12.3 0.008 0.009 2.219 2.225 mp_alltoall_d11v 2046 13.8 1.762 2.123 1.762 2.123 acc_transpose_blocks 24660 15.4 0.115 0.119 1.959 1.993 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.960 1.975 jit_kernel_multiply 10 16.2 1.040 1.945 1.040 1.945 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.846 1.847 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.817 1.832 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.779 1.654 transfer_rs2pw 451 10.6 0.006 0.008 1.395 1.607 mp_allgather_i34 2055 14.4 0.571 1.589 0.571 1.589 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.573 1.584 mp_irecv_dv 57340 16.2 0.654 1.534 0.654 1.534 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.388 1.492 dbcsr_complete_redistribute 325 12.2 0.233 0.291 1.160 1.422 mp_waitany 10164 13.8 1.177 1.385 1.177 1.385 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.718000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=561.090909, yerr=7.656899 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 667.770880E+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 0.008 0.025 59.332 59.332 qs_mol_dyn_low 1 2.0 0.003 0.003 59.088 59.098 qs_forces 11 3.9 0.015 0.021 59.019 59.020 qs_energies 11 4.9 0.001 0.002 55.813 55.828 scf_env_do_scf 11 5.9 0.000 0.001 47.796 47.796 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.438 39.439 velocity_verlet 10 3.0 0.003 0.006 32.185 32.187 dbcsr_multiply_generic 2055 12.4 0.109 0.112 28.566 28.828 qs_scf_new_mos 99 7.5 0.001 0.001 24.889 24.982 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.888 24.981 multiply_cannon 2055 13.4 0.213 0.221 22.468 23.770 ot_scf_mini 99 9.5 0.002 0.003 23.642 23.754 multiply_cannon_loop 2055 14.4 0.810 0.854 21.311 22.322 ot_mini 99 10.5 0.001 0.001 13.592 13.703 rebuild_ks_matrix 110 8.3 0.000 0.000 12.070 12.221 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.070 12.221 multiply_cannon_multrec 16440 15.4 3.688 5.351 9.899 11.490 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.618 10.750 mp_waitall_1 139946 16.5 6.765 9.765 6.765 9.765 qs_ot_get_derivative 99 11.5 0.001 0.001 9.212 9.321 init_scf_loop 11 6.9 0.000 0.000 8.324 8.325 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.257 7.180 prepare_preconditioner 11 7.9 0.000 0.000 6.595 6.610 make_preconditioner 11 8.9 0.000 0.000 6.595 6.610 sum_up_and_integrate 110 10.3 0.001 0.002 6.239 6.269 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.906 6.246 integrate_v_rspace 110 11.3 0.003 0.003 6.213 6.242 dbcsr_mm_accdrv_process 34862 16.1 5.395 5.771 6.056 6.192 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.709 5.722 calculate_rho_elec 110 8.6 0.059 0.059 5.709 5.722 init_scf_run 11 5.9 0.000 0.001 5.630 5.630 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.630 5.630 qs_ot_get_p 110 10.4 0.001 0.001 5.237 5.397 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.846 5.363 apply_single 110 13.6 0.000 0.000 4.846 5.363 make_m2s 4110 13.4 0.049 0.052 4.070 4.454 ot_diis_step 99 11.5 0.011 0.011 4.352 4.352 make_images 4110 14.4 0.398 0.523 3.953 4.333 multiply_cannon_sync_h2d 16440 15.4 3.299 3.977 3.299 3.977 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.203 3.797 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.632 3.640 pw_transfer 1331 11.6 0.065 0.071 3.576 3.586 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.470 3.483 grid_integrate_task_list 110 12.3 3.216 3.414 3.216 3.414 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.267 3.267 density_rs2pw 110 9.6 0.004 0.004 2.926 3.135 fft_wrap_pw1pw2_140 451 13.1 0.460 0.471 3.012 3.027 wfi_extrapolate 11 7.9 0.001 0.001 2.931 2.931 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.777 2.778 make_images_data 4110 15.4 0.046 0.050 2.287 2.766 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.746 2.756 cp_fm_diag_elpa_base 48 14.0 2.681 2.709 2.744 2.754 hybrid_alltoall_any 4261 16.3 0.108 0.380 1.991 2.732 calculate_first_density_matrix 1 7.0 0.000 0.000 2.607 2.609 cp_fm_cholesky_invert 11 10.9 2.564 2.572 2.564 2.572 calculate_dm_sparse 110 9.5 0.001 0.001 2.513 2.540 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.465 2.511 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.423 2.486 fft3d_ps 1111 14.6 1.119 1.126 2.387 2.398 grid_collocate_task_list 110 9.6 2.224 2.396 2.224 2.396 multiply_cannon_metrocomm4 14385 15.4 0.047 0.051 0.864 2.364 mp_irecv_dv 48980 15.7 0.790 2.227 0.790 2.227 mp_sum_l 6594 12.7 1.401 2.115 1.401 2.115 potential_pw2rs 110 12.3 0.011 0.011 2.089 2.095 mp_alltoall_d11v 2046 13.8 1.754 2.059 1.754 2.059 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.953 1.966 dbcsr_complete_redistribute 325 12.2 0.322 0.348 1.401 1.836 cp_fm_upper_to_full 70 14.2 1.397 1.752 1.397 1.752 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.643 1.662 acc_transpose_blocks 16440 15.4 0.077 0.080 1.574 1.634 cp_fm_cholesky_decompose 22 10.9 1.541 1.572 1.541 1.572 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.363 1.501 mp_allgather_i34 2055 14.4 0.468 1.478 0.468 1.478 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.450 1.462 transfer_rs2pw 451 10.6 0.005 0.006 1.224 1.414 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.970 1.409 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.272 1.285 mp_waitany 17072 13.8 1.100 1.277 1.100 1.277 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.332000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.909091, yerr=8.229156 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 733.179904E+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 0.015 0.029 66.187 66.188 qs_mol_dyn_low 1 2.0 0.004 0.010 65.887 65.896 qs_forces 11 3.9 0.007 0.023 65.784 65.790 qs_energies 11 4.9 0.003 0.012 62.350 62.363 scf_env_do_scf 11 5.9 0.001 0.001 53.777 53.779 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.140 42.140 velocity_verlet 10 3.0 0.002 0.006 37.229 37.233 dbcsr_multiply_generic 2055 12.4 0.117 0.122 30.978 31.164 qs_scf_new_mos 99 7.5 0.001 0.001 27.489 27.591 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.489 27.590 ot_scf_mini 99 9.5 0.003 0.003 25.806 25.907 multiply_cannon 2055 13.4 0.239 0.255 23.993 25.218 multiply_cannon_loop 2055 14.4 1.437 1.476 22.437 22.966 ot_mini 99 10.5 0.001 0.002 14.841 14.953 multiply_cannon_multrec 24660 15.4 4.076 6.761 13.269 14.542 rebuild_ks_matrix 110 8.3 0.000 0.000 12.070 12.190 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.027 12.070 12.189 init_scf_loop 11 6.9 0.001 0.003 11.594 11.595 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.659 10.763 qs_ot_get_derivative 99 11.5 0.001 0.001 10.652 10.750 dbcsr_mm_accdrv_process 52304 16.0 8.007 9.646 9.033 9.950 prepare_preconditioner 11 7.9 0.000 0.000 9.794 9.808 make_preconditioner 11 8.9 0.003 0.025 9.794 9.808 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.018 9.497 sum_up_and_integrate 110 10.3 0.002 0.006 6.253 6.265 integrate_v_rspace 110 11.3 0.003 0.003 6.226 6.238 mp_waitall_1 121746 16.5 4.086 6.136 4.086 6.136 qs_ot_get_p 110 10.4 0.001 0.001 5.802 5.945 make_m2s 4110 13.4 0.059 0.061 5.363 5.859 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.848 5.856 calculate_rho_elec 110 8.6 0.078 0.081 5.847 5.856 init_scf_run 11 5.9 0.000 0.001 5.748 5.748 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.748 5.748 make_images 4110 14.4 0.572 0.697 5.222 5.715 cp_fm_upper_to_full 70 14.2 3.365 4.883 3.365 4.883 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.060 4.162 apply_single 110 13.6 0.000 0.000 4.060 4.161 ot_diis_step 99 11.5 0.012 0.017 4.152 4.152 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.930 3.945 dbcsr_complete_redistribute 325 12.2 0.405 0.451 2.666 3.816 pw_transfer 1331 11.6 0.065 0.073 3.771 3.802 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.665 3.699 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.558 3.610 grid_integrate_task_list 110 12.3 3.283 3.455 3.283 3.455 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.365 3.417 cp_dbcsr_syevd 48 12.0 0.003 0.005 3.371 3.371 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.140 3.281 fft_wrap_pw1pw2_140 451 13.1 0.529 0.544 3.157 3.194 hybrid_alltoall_any 4261 16.3 0.122 0.456 2.278 3.140 density_rs2pw 110 9.6 0.004 0.004 2.965 3.130 make_images_data 4110 15.4 0.048 0.052 2.638 3.109 calculate_dm_sparse 110 9.5 0.001 0.001 3.056 3.088 wfi_extrapolate 11 7.9 0.001 0.001 3.010 3.010 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.345 2.997 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.974 2.978 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.765 2.887 mp_alltoall_i22 605 13.7 1.687 2.859 1.687 2.859 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.839 2.849 cp_fm_diag_elpa_base 48 14.0 2.691 2.748 2.835 2.846 calculate_first_density_matrix 1 7.0 0.001 0.004 2.643 2.645 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.593 2.632 multiply_cannon_sync_h2d 24660 15.4 2.408 2.625 2.408 2.625 acc_transpose_blocks 24660 15.4 0.111 0.114 2.476 2.595 cp_fm_cholesky_invert 11 10.9 2.566 2.574 2.566 2.574 fft3d_ps 1111 14.6 1.152 1.192 2.492 2.512 grid_collocate_task_list 110 9.6 2.260 2.462 2.260 2.462 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.296 2.300 potential_pw2rs 110 12.3 0.013 0.013 2.042 2.045 mp_alltoall_d11v 2046 13.8 1.797 1.998 1.797 1.998 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.805 1.835 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.005 1.614 1.716 cp_fm_cholesky_decompose 22 10.9 1.645 1.694 1.645 1.694 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.651 1.665 jit_kernel_multiply 8 15.8 0.678 1.652 0.678 1.652 mp_allgather_i34 2055 14.4 0.640 1.641 0.640 1.641 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.615 1.629 multiply_cannon_metrocomm4 20550 15.4 0.059 0.061 0.828 1.562 acc_transpose_blocks_sync 73980 16.4 1.450 1.562 1.450 1.562 mp_irecv_dv 62702 16.1 0.726 1.479 0.726 1.479 mp_sum_l 6594 12.7 0.939 1.424 0.939 1.424 transfer_rs2pw 451 10.6 0.005 0.006 1.207 1.413 mp_waitany 13376 13.8 1.077 1.341 1.077 1.341 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.236 1.335 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.188000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.818182, yerr=8.809274 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 850.239488E+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 0.009 0.024 55.729 55.729 qs_mol_dyn_low 1 2.0 0.003 0.003 55.451 55.461 qs_forces 11 3.9 0.003 0.003 55.386 55.386 qs_energies 11 4.9 0.001 0.002 51.740 51.745 scf_env_do_scf 11 5.9 0.000 0.001 43.083 43.083 scf_env_do_scf_inner_loop 99 6.5 0.019 0.025 35.417 35.418 velocity_verlet 10 3.0 0.002 0.003 31.025 31.028 dbcsr_multiply_generic 2055 12.4 0.107 0.117 23.689 23.849 qs_scf_new_mos 99 7.5 0.001 0.001 20.763 20.810 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.762 20.809 ot_scf_mini 99 9.5 0.002 0.002 19.503 19.532 multiply_cannon 2055 13.4 0.242 0.253 18.015 19.276 multiply_cannon_loop 2055 14.4 0.604 0.626 16.742 17.086 rebuild_ks_matrix 110 8.3 0.000 0.000 11.674 11.695 qs_ks_build_kohn_sham_matrix 110 9.3 0.033 0.037 11.674 11.694 ot_mini 99 10.5 0.001 0.001 10.809 10.830 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.389 10.404 multiply_cannon_multrec 8220 15.4 3.217 4.438 7.978 8.914 init_scf_loop 11 6.9 0.000 0.000 7.620 7.621 mp_waitall_1 103326 16.6 5.756 7.265 5.756 7.265 qs_ot_get_derivative 99 11.5 0.001 0.001 7.137 7.167 sum_up_and_integrate 110 10.3 0.001 0.002 6.215 6.227 integrate_v_rspace 110 11.3 0.003 0.003 6.189 6.201 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.047 6.059 calculate_rho_elec 110 8.6 0.112 0.113 6.046 6.059 prepare_preconditioner 11 7.9 0.000 0.000 5.958 5.980 make_preconditioner 11 8.9 0.000 0.000 5.958 5.980 dbcsr_mm_accdrv_process 17442 15.9 3.191 4.378 4.622 5.650 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.541 5.614 init_scf_run 11 5.9 0.000 0.001 5.540 5.540 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.540 5.540 qs_ot_get_p 110 10.4 0.001 0.001 4.760 4.772 make_m2s 4110 13.4 0.038 0.039 4.231 4.526 make_images 4110 14.4 0.647 0.707 4.100 4.395 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.968 4.379 pw_transfer 1331 11.6 0.065 0.070 4.081 4.089 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.973 3.985 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.727 3.799 apply_single 110 13.6 0.000 0.000 3.727 3.799 ot_diis_step 99 11.5 0.012 0.012 3.649 3.649 grid_integrate_task_list 110 12.3 3.366 3.514 3.366 3.514 fft_wrap_pw1pw2_140 451 13.1 0.719 0.732 3.453 3.466 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.403 3.407 density_rs2pw 110 9.6 0.004 0.004 3.022 3.173 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.062 3.063 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.869 2.871 cp_fm_cholesky_invert 11 10.9 2.840 2.843 2.840 2.843 hybrid_alltoall_any 4261 16.3 0.199 0.849 2.302 2.830 calculate_first_density_matrix 1 7.0 0.000 0.000 2.735 2.736 wfi_extrapolate 11 7.9 0.001 0.001 2.709 2.709 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.663 2.663 make_images_data 4110 15.4 0.040 0.047 2.266 2.662 calculate_dm_sparse 110 9.5 0.001 0.001 2.585 2.619 fft3d_ps 1111 14.6 1.291 1.298 2.517 2.532 grid_collocate_task_list 110 9.6 2.368 2.518 2.368 2.518 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.483 2.516 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.489 2.497 cp_fm_diag_elpa_base 48 14.0 2.434 2.458 2.488 2.495 multiply_cannon_sync_h2d 8220 15.4 2.378 2.441 2.378 2.441 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.051 2.081 potential_pw2rs 110 12.3 0.015 0.016 2.017 2.019 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.787 2.001 mp_alltoall_d11v 2046 13.8 1.747 1.977 1.747 1.977 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.918 1.934 jit_kernel_multiply 8 15.6 1.115 1.744 1.115 1.744 cp_fm_cholesky_decompose 22 10.9 1.667 1.696 1.667 1.696 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.614 1.619 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.494 1.606 dbcsr_complete_redistribute 325 12.2 0.551 0.587 1.446 1.601 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.528 1.538 mp_allgather_i34 2055 14.4 0.440 1.535 0.440 1.535 acc_transpose_blocks 8220 15.4 0.040 0.041 1.439 1.505 qs_create_task_list 11 7.9 0.000 0.001 1.204 1.298 generate_qs_task_list 11 8.9 0.373 0.440 1.203 1.297 transfer_rs2pw 451 10.6 0.005 0.005 1.113 1.293 mp_waitany 9240 13.8 1.057 1.240 1.057 1.240 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.203 1.229 multiply_cannon_metrocomm1 8220 15.4 0.022 0.022 0.700 1.119 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.729000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=803.272727, yerr=9.430038 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 1.379410E+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 0.013 0.033 87.883 87.884 qs_mol_dyn_low 1 2.0 0.003 0.003 87.508 87.520 qs_forces 11 3.9 0.010 0.010 87.438 87.438 qs_energies 11 4.9 0.008 0.009 83.247 83.249 scf_env_do_scf 11 5.9 0.000 0.001 73.055 73.056 velocity_verlet 10 3.0 0.002 0.002 55.850 55.857 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.939 44.941 dbcsr_multiply_generic 2055 12.4 0.124 0.128 29.487 29.580 init_scf_loop 11 6.9 0.000 0.000 28.045 28.050 qs_scf_new_mos 99 7.5 0.001 0.001 26.892 26.917 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.891 26.917 prepare_preconditioner 11 7.9 0.000 0.000 26.019 26.025 make_preconditioner 11 8.9 0.000 0.000 26.019 26.025 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.334 25.484 ot_scf_mini 99 9.5 0.002 0.002 25.088 25.105 multiply_cannon 2055 13.4 0.345 0.372 22.379 23.196 multiply_cannon_loop 2055 14.4 0.824 0.843 20.588 20.856 cp_fm_upper_to_full 70 14.2 12.575 17.969 12.575 17.969 ot_mini 99 10.5 0.001 0.001 14.107 14.122 rebuild_ks_matrix 110 8.3 0.000 0.001 14.041 14.048 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.040 14.048 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.741 12.752 dbcsr_complete_redistribute 325 12.2 1.014 1.038 7.265 10.364 multiply_cannon_multrec 8220 15.4 4.055 4.219 9.737 9.851 qs_ot_get_derivative 99 11.5 0.001 0.001 9.539 9.556 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.236 9.337 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.669 8.748 mp_alltoall_i22 605 13.7 5.292 8.384 5.292 8.384 mp_waitall_1 84994 16.7 7.450 8.235 7.450 8.235 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.528 7.559 calculate_rho_elec 110 8.6 0.222 0.223 7.527 7.558 sum_up_and_integrate 110 10.3 0.002 0.002 6.957 6.973 integrate_v_rspace 110 11.3 0.004 0.004 6.929 6.945 init_scf_run 11 5.9 0.000 0.001 5.920 5.920 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.920 5.920 pw_transfer 1331 11.6 0.074 0.075 5.851 5.857 make_m2s 4110 13.4 0.043 0.043 5.339 5.826 dbcsr_mm_accdrv_process 11614 15.7 3.811 4.450 5.534 5.755 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.735 5.741 qs_ot_get_p 110 10.4 0.001 0.001 5.687 5.724 make_images 4110 14.4 0.880 0.936 5.149 5.635 cp_fm_cholesky_invert 11 10.9 5.241 5.245 5.241 5.245 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.815 5.201 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.698 5.161 apply_single 110 13.6 0.000 0.000 4.698 5.161 fft_wrap_pw1pw2_140 451 13.1 1.366 1.371 5.031 5.037 ot_diis_step 99 11.5 0.015 0.016 4.550 4.550 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.112 4.119 density_rs2pw 110 9.6 0.004 0.004 3.981 4.007 grid_integrate_task_list 110 12.3 3.679 3.735 3.679 3.735 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.649 3.651 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.592 3.593 hybrid_alltoall_any 4261 16.3 0.264 0.565 2.802 3.538 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.037 3.481 fft3d_ps 1111 14.6 1.881 1.888 3.442 3.450 make_images_data 4110 15.4 0.045 0.048 2.789 3.417 wfi_extrapolate 11 7.9 0.001 0.001 3.323 3.323 calculate_dm_sparse 110 9.5 0.001 0.001 3.179 3.207 multiply_cannon_sync_h2d 8220 15.4 3.129 3.147 3.129 3.147 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.001 3.001 cp_fm_diag_elpa_base 48 14.0 2.465 2.661 2.999 2.999 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.889 2.890 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.685 2.708 grid_collocate_task_list 110 9.6 2.669 2.705 2.669 2.705 calculate_first_density_matrix 1 7.0 0.000 0.001 2.474 2.477 potential_pw2rs 110 12.3 0.021 0.022 2.466 2.469 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.392 2.397 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.180 2.240 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.119 2.232 mp_alltoall_d11v 2046 13.8 2.076 2.121 2.076 2.121 cp_fm_cholesky_decompose 22 10.9 1.987 2.007 1.987 2.007 qs_create_task_list 11 7.9 0.000 0.001 1.871 1.916 generate_qs_task_list 11 8.9 0.727 0.780 1.871 1.915 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.901 1.905 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.756 1.778 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.884000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1250.727273, yerr=52.152035 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 631.492608E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57505. MP_Allreduce 11129 795. MP_Sync 87 MP_Alltoall 2226 3004121. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.083 207.705 207.706 qs_mol_dyn_low 1 2.0 0.003 0.004 207.102 207.115 qs_forces 11 3.9 0.010 0.015 207.017 207.018 qs_energies 11 4.9 0.006 0.009 201.450 201.475 scf_env_do_scf 11 5.9 0.001 0.001 185.002 185.006 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 164.674 164.676 dbcsr_multiply_generic 2507 12.6 0.178 0.183 126.302 127.020 qs_scf_new_mos 117 7.6 0.001 0.001 125.359 125.669 qs_scf_loop_do_ot 117 8.6 0.001 0.001 125.358 125.669 velocity_verlet 10 3.0 0.002 0.005 123.406 123.407 ot_scf_mini 117 9.6 0.003 0.003 118.666 118.923 multiply_cannon 2507 13.6 0.237 0.248 102.044 103.738 multiply_cannon_loop 2507 14.6 2.391 2.438 99.879 101.403 ot_mini 117 10.6 0.001 0.001 67.005 67.274 multiply_cannon_multrec 60168 15.6 31.722 33.646 41.635 43.824 qs_ot_get_derivative 117 11.6 0.001 0.001 42.151 42.388 rebuild_ks_matrix 128 8.3 0.001 0.001 33.434 33.821 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.433 33.820 mp_waitall_1 267128 16.5 29.450 32.503 29.450 32.503 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.025 30.373 qs_ot_get_p 128 10.4 0.001 0.001 29.522 29.879 multiply_cannon_sync_h2d 60168 15.6 26.182 27.829 26.182 27.829 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.352 25.381 apply_single 128 13.6 0.001 0.001 24.352 25.380 ot_diis_step 117 11.6 0.008 0.008 24.614 24.615 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.790 22.867 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.471 20.661 init_scf_loop 11 6.9 0.001 0.001 20.254 20.255 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.981 19.981 multiply_cannon_metrocomm3 60168 15.6 0.116 0.123 16.509 18.834 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.009 17.038 cp_fm_diag_elpa_base 83 14.4 16.936 16.968 17.004 17.035 prepare_preconditioner 11 7.9 0.000 0.000 15.684 15.718 make_preconditioner 11 8.9 0.000 0.000 15.684 15.718 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.913 15.069 make_m2s 5014 13.6 0.105 0.114 13.860 14.323 make_images 5014 14.6 0.397 0.415 13.674 14.148 sum_up_and_integrate 128 10.3 0.002 0.004 13.934 13.950 integrate_v_rspace 128 11.3 0.003 0.004 13.876 13.893 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.101 13.214 calculate_rho_elec 128 8.7 0.045 0.064 13.100 13.213 init_scf_run 11 5.9 0.000 0.001 12.289 12.290 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.289 12.290 mp_sum_l 7950 12.9 9.137 10.303 9.137 10.303 dbcsr_mm_accdrv_process 124484 16.2 4.794 4.936 9.487 10.003 wfi_extrapolate 11 7.9 0.001 0.001 9.005 9.005 cp_fm_cholesky_invert 11 10.9 8.902 8.911 8.902 8.911 calculate_dm_sparse 128 9.5 0.001 0.001 8.540 8.641 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.189 8.302 multiply_cannon_metrocomm1 60168 15.6 0.093 0.097 6.481 8.279 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.053 8.125 pw_transfer 1547 11.6 0.073 0.094 7.594 7.768 make_images_data 5014 15.6 0.065 0.073 6.782 7.750 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.393 7.571 grid_integrate_task_list 128 12.3 7.072 7.492 7.072 7.492 density_rs2pw 128 9.7 0.006 0.006 6.743 7.454 hybrid_alltoall_any 5200 16.5 0.294 2.267 5.954 7.156 fft_wrap_pw1pw2_140 523 13.2 0.857 0.896 6.487 6.670 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.641 6.652 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.790 5.868 fft3d_ps 1291 14.7 2.194 2.843 5.402 5.722 mp_alltoall_d11v 2415 14.1 4.321 5.547 4.321 5.547 grid_collocate_task_list 128 9.7 4.844 5.277 4.844 5.277 cp_fm_cholesky_decompose 22 10.9 4.591 4.604 4.591 4.604 potential_pw2rs 128 12.3 0.009 0.010 4.421 4.440 mp_sum_d 4481 12.1 3.447 4.215 3.447 4.215 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.706000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.181818, yerr=5.669988 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 844.001280E+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 4020 57958. MP_Allreduce 11128 958. MP_Sync 87 MP_Alltoall 1969 6378638. 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 0.018 0.061 189.335 189.337 qs_mol_dyn_low 1 2.0 0.003 0.004 188.841 188.909 qs_forces 11 3.9 0.005 0.006 188.741 188.742 qs_energies 11 4.9 0.005 0.026 181.982 181.998 scf_env_do_scf 11 5.9 0.001 0.002 165.618 165.628 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 133.037 133.040 velocity_verlet 10 3.0 0.003 0.008 118.806 118.808 dbcsr_multiply_generic 2507 12.6 0.188 0.194 97.276 98.453 qs_scf_new_mos 117 7.6 0.001 0.001 94.411 94.867 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.411 94.866 ot_scf_mini 117 9.6 0.004 0.005 89.612 90.167 multiply_cannon 2507 13.6 0.505 0.570 77.237 81.412 multiply_cannon_loop 2507 14.6 1.557 1.628 73.834 76.540 ot_mini 117 10.6 0.001 0.001 50.180 50.671 mp_waitall_1 214728 16.6 23.680 39.493 23.680 39.493 multiply_cannon_multrec 30084 15.6 21.127 25.664 31.735 36.602 rebuild_ks_matrix 128 8.3 0.001 0.001 32.295 32.973 qs_ks_build_kohn_sham_matrix 128 9.3 0.022 0.031 32.294 32.973 init_scf_loop 11 6.9 0.001 0.003 32.493 32.494 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.033 29.670 multiply_cannon_metrocomm3 30084 15.6 0.098 0.106 15.374 29.466 qs_ot_get_derivative 117 11.6 0.001 0.002 28.312 28.854 prepare_preconditioner 11 7.9 0.000 0.000 28.169 28.244 make_preconditioner 11 8.9 0.000 0.004 28.169 28.244 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.862 27.391 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.987 23.043 apply_single 128 13.6 0.001 0.001 21.986 23.043 qs_ot_get_p 128 10.4 0.001 0.002 21.437 22.019 ot_diis_step 117 11.6 0.015 0.015 21.688 21.690 multiply_cannon_sync_h2d 30084 15.6 18.162 20.758 18.162 20.758 cp_fm_cholesky_invert 11 10.9 16.520 16.532 16.520 16.532 qs_ot_p2m_diag 83 11.4 0.188 0.216 16.463 16.496 make_m2s 5014 13.6 0.086 0.091 14.082 15.781 make_images 5014 14.6 1.161 1.361 13.865 15.565 cp_dbcsr_syevd 83 12.4 0.006 0.007 15.259 15.260 sum_up_and_integrate 128 10.3 0.002 0.004 13.966 13.993 integrate_v_rspace 128 11.3 0.003 0.004 13.906 13.935 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.447 13.476 calculate_rho_elec 128 8.7 0.087 0.103 13.446 13.476 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.967 11.996 cp_fm_diag_elpa_base 83 14.4 11.712 11.799 11.960 11.984 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.262 11.656 init_scf_run 11 5.9 0.000 0.001 11.474 11.475 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.474 11.475 multiply_cannon_metrocomm4 27577 15.6 0.099 0.113 3.809 10.796 dbcsr_mm_accdrv_process 62242 16.2 5.541 6.101 10.052 10.611 mp_irecv_dv 69486 16.3 3.609 10.398 3.609 10.398 make_images_data 5014 15.6 0.065 0.076 8.301 10.302 hybrid_alltoall_any 5200 16.5 0.348 1.565 7.308 10.157 pw_transfer 1547 11.6 0.084 0.098 8.550 8.609 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.327 8.388 wfi_extrapolate 11 7.9 0.001 0.001 8.382 8.382 density_rs2pw 128 9.7 0.006 0.006 7.020 7.516 grid_integrate_task_list 128 12.3 7.161 7.494 7.161 7.494 fft_wrap_pw1pw2_140 523 13.2 0.934 0.955 7.371 7.444 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.298 7.111 cp_fm_cholesky_decompose 22 10.9 6.813 6.893 6.813 6.893 calculate_dm_sparse 128 9.5 0.001 0.001 6.467 6.602 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.095 6.104 mp_sum_l 7950 12.9 4.176 6.053 4.176 6.053 fft3d_ps 1291 14.7 2.826 3.004 5.882 5.941 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.361 5.434 grid_collocate_task_list 128 9.7 5.007 5.364 5.007 5.364 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.186 5.358 mp_alltoall_d11v 2415 14.1 4.598 5.314 4.598 5.314 mp_allgather_i34 2507 14.6 1.920 5.008 1.920 5.008 potential_pw2rs 128 12.3 0.016 0.018 4.451 4.467 dbcsr_complete_redistribute 395 12.7 0.757 0.829 3.076 3.932 mp_sum_d 4481 12.1 2.619 3.919 2.619 3.919 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.337000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=804.000000, yerr=2.132007 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 988.618752E+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 4002 58208. MP_Allreduce 11082 999. 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 0.011 0.027 173.931 173.932 qs_mol_dyn_low 1 2.0 0.003 0.003 173.556 173.569 qs_forces 11 3.9 0.004 0.004 173.460 173.462 qs_energies 11 4.9 0.001 0.002 166.919 166.930 scf_env_do_scf 11 5.9 0.001 0.001 151.306 151.308 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.628 116.629 velocity_verlet 10 3.0 0.001 0.002 111.441 111.443 dbcsr_multiply_generic 2507 12.6 0.186 0.197 81.606 82.864 qs_scf_new_mos 117 7.6 0.001 0.001 79.742 80.124 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.741 80.123 ot_scf_mini 117 9.6 0.003 0.004 75.498 75.922 multiply_cannon 2507 13.6 0.520 0.548 62.111 66.661 multiply_cannon_loop 2507 14.6 1.123 1.193 59.033 62.313 ot_mini 117 10.6 0.001 0.001 42.186 42.623 init_scf_loop 11 6.9 0.000 0.002 34.581 34.582 mp_waitall_1 170520 16.6 24.156 33.187 24.156 33.187 prepare_preconditioner 11 7.9 0.000 0.000 30.604 30.651 make_preconditioner 11 8.9 0.000 0.000 30.604 30.651 rebuild_ks_matrix 128 8.3 0.001 0.001 30.044 30.617 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 30.044 30.616 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.239 29.660 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.084 27.608 multiply_cannon_multrec 20056 15.6 13.255 15.926 22.571 25.373 multiply_cannon_metrocomm3 20056 15.6 0.061 0.065 15.068 24.659 qs_ot_get_derivative 117 11.6 0.001 0.002 22.764 23.187 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.445 20.393 apply_single 128 13.6 0.001 0.001 19.445 20.392 ot_diis_step 117 11.6 0.018 0.019 19.309 19.310 qs_ot_get_p 128 10.4 0.001 0.001 18.668 19.197 make_m2s 5014 13.6 0.078 0.083 14.469 15.776 make_images 5014 14.6 1.145 1.235 14.231 15.529 multiply_cannon_sync_h2d 20056 15.6 13.700 15.211 13.700 15.211 qs_ot_p2m_diag 83 11.4 0.264 0.271 14.395 14.403 cp_fm_cholesky_invert 11 10.9 14.168 14.176 14.168 14.176 sum_up_and_integrate 128 10.3 0.002 0.003 13.812 13.836 integrate_v_rspace 128 11.3 0.003 0.003 13.753 13.773 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.499 13.538 calculate_rho_elec 128 8.7 0.129 0.143 13.498 13.537 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.346 13.346 init_scf_run 11 5.9 0.000 0.001 10.626 10.626 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.626 10.626 make_images_data 5014 15.6 0.063 0.072 8.632 10.395 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.186 10.204 cp_fm_diag_elpa_base 83 14.4 9.778 9.944 10.183 10.201 hybrid_alltoall_any 5200 16.5 0.446 2.021 7.502 9.809 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.998 9.299 multiply_cannon_metrocomm4 17549 15.6 0.065 0.074 3.401 9.040 dbcsr_mm_accdrv_process 41502 16.2 5.632 5.941 8.771 8.891 pw_transfer 1547 11.6 0.084 0.100 8.697 8.796 mp_irecv_dv 50230 16.2 3.273 8.792 3.273 8.792 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 8.474 8.579 grid_integrate_task_list 128 12.3 7.311 7.669 7.311 7.669 fft_wrap_pw1pw2_140 523 13.2 1.054 1.086 7.545 7.654 wfi_extrapolate 11 7.9 0.001 0.001 7.405 7.405 cp_fm_upper_to_full 105 14.8 5.803 7.336 5.803 7.336 cp_fm_cholesky_decompose 22 10.9 7.208 7.250 7.208 7.250 density_rs2pw 128 9.7 0.005 0.006 6.876 7.192 dbcsr_complete_redistribute 395 12.7 1.161 1.196 4.644 6.368 calculate_dm_sparse 128 9.5 0.001 0.001 5.838 5.928 fft3d_ps 1291 14.7 2.818 3.061 5.797 5.893 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.609 5.616 grid_collocate_task_list 128 9.7 5.151 5.518 5.151 5.518 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.498 5.219 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.588 5.187 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.621 4.807 mp_sum_l 7950 12.9 3.185 4.766 3.185 4.766 mp_allgather_i34 2507 14.6 1.563 4.749 1.563 4.749 mp_alltoall_d11v 2415 14.1 4.228 4.697 4.228 4.697 potential_pw2rs 128 12.3 0.021 0.023 4.275 4.283 transfer_fm_to_dbcsr 11 9.9 0.020 0.024 2.345 4.061 mp_alltoall_i22 716 14.1 2.046 3.985 2.046 3.985 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.897 3.925 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.705 3.706 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.457 3.505 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=173.932000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=925.727273, yerr=20.578413 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 4.387043E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6024576 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.8 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.168290E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1152312 MPI messages size (bytes): total size 2.039354E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.769794E+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 321600 36333158400 131072 < size <= 4194304 721560 792041881600 4194304 < size <= 16777216 70800 669921260160 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57894. MP_Allreduce 11137 1080. MP_Sync 87 MP_Alltoall 1724 12509439. 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 0.010 0.029 188.044 188.045 qs_mol_dyn_low 1 2.0 0.003 0.005 187.640 187.654 qs_forces 11 3.9 0.007 0.008 187.508 187.515 qs_energies 11 4.9 0.008 0.010 180.421 180.434 scf_env_do_scf 11 5.9 0.004 0.011 163.834 163.845 velocity_verlet 10 3.0 0.002 0.005 125.130 125.133 scf_env_do_scf_inner_loop 118 6.6 0.003 0.010 117.699 117.700 dbcsr_multiply_generic 2527 12.6 0.193 0.198 81.426 82.293 qs_scf_new_mos 118 7.6 0.001 0.001 81.091 81.391 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.090 81.390 ot_scf_mini 118 9.6 0.003 0.004 76.351 76.675 multiply_cannon 2527 13.6 0.561 0.600 56.334 58.601 multiply_cannon_loop 2527 14.6 1.879 1.931 52.419 54.362 init_scf_loop 11 6.9 0.001 0.003 46.011 46.016 ot_mini 118 10.6 0.001 0.001 43.393 43.709 prepare_preconditioner 11 7.9 0.000 0.000 41.910 41.945 make_preconditioner 11 8.9 0.000 0.000 41.910 41.945 make_full_inverse_cholesky 11 9.9 0.010 0.022 35.562 40.589 multiply_cannon_multrec 30324 15.6 13.825 18.895 26.813 31.538 rebuild_ks_matrix 129 8.3 0.001 0.001 29.519 29.818 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 29.518 29.817 mp_waitall_1 149060 16.7 17.277 28.432 17.277 28.432 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.662 26.934 qs_ot_get_derivative 118 11.6 0.001 0.002 23.574 23.911 make_m2s 5054 13.6 0.092 0.097 20.672 21.858 make_images 5054 14.6 1.950 2.465 20.362 21.544 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.218 19.846 apply_single 129 13.6 0.001 0.001 19.218 19.845 ot_diis_step 118 11.6 0.018 0.019 19.690 19.692 qs_ot_get_p 129 10.4 0.001 0.001 18.931 19.259 cp_fm_upper_to_full 105 14.8 11.488 16.918 11.488 16.918 multiply_cannon_metrocomm3 30324 15.6 0.049 0.051 6.262 16.055 cp_fm_cholesky_invert 11 10.9 15.926 15.934 15.926 15.934 qs_ot_p2m_diag 83 11.4 0.342 0.389 14.726 14.779 sum_up_and_integrate 129 10.3 0.002 0.003 14.019 14.043 integrate_v_rspace 129 11.3 0.003 0.004 13.958 13.985 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.840 13.888 calculate_rho_elec 129 8.7 0.171 0.187 13.839 13.887 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.328 13.330 make_images_data 5054 15.6 0.065 0.070 11.184 13.127 dbcsr_mm_accdrv_process 62756 16.2 8.538 9.507 12.556 13.052 dbcsr_complete_redistribute 395 12.7 1.473 1.598 9.043 12.843 hybrid_alltoall_any 5240 16.5 0.529 2.264 9.872 12.165 multiply_cannon_sync_h2d 30324 15.6 10.566 11.642 10.566 11.642 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.694 11.481 init_scf_run 11 5.9 0.000 0.001 10.690 10.691 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.689 10.691 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.946 10.185 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.140 10.157 cp_fm_diag_elpa_base 83 14.4 9.173 9.476 10.132 10.148 transfer_fm_to_dbcsr 11 9.9 0.001 0.008 6.326 10.048 mp_alltoall_i22 718 14.1 5.602 9.322 5.602 9.322 pw_transfer 1559 11.6 0.086 0.102 9.198 9.249 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.971 9.029 fft_wrap_pw1pw2_140 527 13.2 1.255 1.293 7.999 8.072 grid_integrate_task_list 129 12.3 7.589 8.007 7.589 8.007 wfi_extrapolate 11 7.9 0.001 0.001 7.633 7.633 cp_fm_cholesky_decompose 22 10.9 7.428 7.524 7.428 7.524 density_rs2pw 129 9.7 0.005 0.006 6.826 7.278 multiply_cannon_metrocomm4 25270 15.6 0.080 0.089 2.749 7.112 mp_irecv_dv 76685 16.2 2.598 6.844 2.598 6.844 calculate_dm_sparse 129 9.5 0.001 0.001 6.430 6.509 fft3d_ps 1301 14.7 3.004 3.046 5.998 6.046 grid_collocate_task_list 129 9.7 5.330 5.799 5.330 5.799 mp_alltoall_d11v 2423 14.1 4.978 5.693 4.978 5.693 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.454 5.509 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.601 4.689 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.380 4.461 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.401 4.405 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.314 4.374 potential_pw2rs 129 12.3 0.023 0.023 4.267 4.276 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.045000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1096.454545, yerr=15.423498 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.535484E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108638608 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58273. MP_Allreduce 11070 1167. MP_Sync 86 MP_Alltoall 1700 18828160. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.029 167.099 167.100 qs_mol_dyn_low 1 2.0 0.003 0.003 166.644 166.684 qs_forces 11 3.9 0.004 0.004 166.526 166.528 qs_energies 11 4.9 0.002 0.002 159.091 159.097 scf_env_do_scf 11 5.9 0.001 0.001 141.951 141.959 velocity_verlet 10 3.0 0.001 0.002 109.296 109.302 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 106.720 106.722 dbcsr_multiply_generic 2485 12.5 0.183 0.190 71.030 71.379 qs_scf_new_mos 116 7.6 0.001 0.001 70.099 70.185 qs_scf_loop_do_ot 116 8.6 0.001 0.001 70.098 70.184 ot_scf_mini 116 9.6 0.003 0.004 65.732 65.782 multiply_cannon 2485 13.5 0.559 0.578 52.622 56.032 multiply_cannon_loop 2485 14.5 0.798 0.839 49.653 50.308 ot_mini 116 10.6 0.001 0.001 36.812 36.860 init_scf_loop 11 6.9 0.000 0.001 35.084 35.086 prepare_preconditioner 11 7.9 0.000 0.000 31.152 31.187 make_preconditioner 11 8.9 0.000 0.000 31.152 31.187 mp_waitall_1 124680 16.7 24.201 30.389 24.201 30.389 make_full_inverse_cholesky 11 9.9 0.014 0.029 29.098 29.344 rebuild_ks_matrix 127 8.3 0.001 0.001 28.932 29.027 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 28.932 29.027 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.350 26.439 multiply_cannon_multrec 9940 15.5 10.300 14.509 17.803 21.111 qs_ot_get_derivative 116 11.6 0.001 0.002 20.066 20.127 multiply_cannon_metrocomm3 9940 15.5 0.023 0.025 12.427 19.350 cp_fm_cholesky_invert 11 10.9 17.721 17.727 17.721 17.727 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 16.813 16.987 apply_single 127 13.6 0.001 0.001 16.812 16.986 ot_diis_step 116 11.6 0.020 0.020 16.677 16.677 qs_ot_get_p 127 10.4 0.001 0.001 16.337 16.382 make_m2s 4970 13.5 0.063 0.068 14.599 15.501 make_images 4970 14.5 2.136 2.578 14.292 15.188 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.058 14.092 calculate_rho_elec 127 8.7 0.251 0.262 14.058 14.091 sum_up_and_integrate 127 10.3 0.002 0.002 14.005 14.056 integrate_v_rspace 127 11.3 0.004 0.004 13.945 13.998 qs_ot_p2m_diag 82 11.4 0.489 0.494 12.677 12.692 cp_dbcsr_syevd 82 12.4 0.005 0.005 11.535 11.535 multiply_cannon_sync_h2d 9940 15.5 10.840 11.316 10.840 11.316 init_scf_run 11 5.9 0.000 0.001 10.225 10.226 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.225 10.226 pw_transfer 1535 11.6 0.084 0.093 9.784 9.829 hybrid_alltoall_any 5155 16.4 0.834 3.740 8.251 9.752 make_images_data 4970 15.5 0.054 0.063 8.395 9.685 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 9.563 9.615 cp_fm_diag_elpa 82 13.4 0.000 0.000 8.727 8.738 cp_fm_diag_elpa_base 82 14.4 8.490 8.565 8.721 8.733 fft_wrap_pw1pw2_140 519 13.2 1.663 1.708 8.494 8.546 grid_integrate_task_list 127 12.3 7.702 8.069 7.702 8.069 cp_fm_cholesky_decompose 22 10.9 7.819 7.943 7.819 7.943 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.825 7.873 dbcsr_mm_accdrv_process 20590 16.1 3.324 4.418 7.156 7.849 wfi_extrapolate 11 7.9 0.001 0.001 7.259 7.259 density_rs2pw 127 9.7 0.005 0.005 6.859 7.129 multiply_cannon_metrocomm1 9940 15.5 0.029 0.030 4.154 6.927 calculate_dm_sparse 127 9.5 0.001 0.001 6.009 6.077 fft3d_ps 1281 14.7 3.119 3.218 6.040 6.070 grid_collocate_task_list 127 9.7 5.528 5.784 5.528 5.784 dbcsr_complete_redistribute 393 12.7 2.083 2.122 5.091 5.491 qs_energies_init_hamiltonians 11 5.9 0.061 0.086 5.332 5.333 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.068 5.081 mp_alltoall_d11v 2401 14.1 4.450 5.058 4.450 5.058 mp_allgather_i34 2485 14.5 1.174 4.609 1.174 4.609 potential_pw2rs 127 12.3 0.026 0.027 4.277 4.284 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.020 4.065 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.583 3.916 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.559 3.643 multiply_cannon_metrocomm4 7455 15.5 0.026 0.028 1.689 3.623 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.278 3.604 mp_irecv_dv 28618 15.9 1.651 3.550 1.651 3.550 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.524 3.548 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.390 3.485 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.476 3.484 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=167.100000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1442.909091, yerr=18.971052 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 11.696233E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 1964048 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 3.026043E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 100280 MPI messages size (bytes): total size 1.136195E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.330227E+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 45208 35089547264 4194304 < size <= 16777216 44352 379752284160 16777216 < size 10104 721350092304 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 59279. MP_Allreduce 11057 1504. MP_Sync 87 MP_Alltoall 1712 36974159. MP_SendRecv 1792 218624. MP_ISendRecv 1792 218624. MP_Wait 9802 MP_ISend 6408 1080322. MP_IRecv 6408 1080322. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.054 291.216 291.233 qs_mol_dyn_low 1 2.0 0.003 0.004 290.683 290.702 qs_forces 11 3.9 0.006 0.019 290.585 290.588 qs_energies 11 4.9 0.002 0.003 281.698 281.707 scf_env_do_scf 11 5.9 0.001 0.001 259.176 259.182 velocity_verlet 10 3.0 0.004 0.008 209.179 209.187 scf_env_do_scf_inner_loop 117 6.6 0.004 0.009 136.000 136.003 init_scf_loop 11 6.9 0.000 0.000 122.903 122.908 prepare_preconditioner 11 7.9 0.000 0.000 117.994 118.010 make_preconditioner 11 8.9 0.000 0.000 117.994 118.010 make_full_inverse_cholesky 11 9.9 0.038 0.039 94.868 115.109 qs_scf_new_mos 117 7.6 0.001 0.001 91.206 91.257 qs_scf_loop_do_ot 117 8.6 0.001 0.001 91.205 91.256 ot_scf_mini 117 9.6 0.004 0.004 86.316 86.358 dbcsr_multiply_generic 2507 12.6 0.217 0.227 82.256 82.710 cp_fm_upper_to_full 105 14.8 52.404 74.436 52.404 74.436 multiply_cannon 2507 13.6 0.687 0.743 59.443 60.356 multiply_cannon_loop 2507 14.6 1.055 1.069 55.444 56.681 ot_mini 117 10.6 0.001 0.001 44.442 44.490 dbcsr_complete_redistribute 395 12.7 3.973 4.048 29.037 41.700 copy_fm_to_dbcsr 209 11.7 0.001 0.002 25.556 38.262 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 23.084 35.708 rebuild_ks_matrix 128 8.3 0.001 0.001 34.767 34.802 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.019 34.767 34.801 mp_alltoall_i22 716 14.1 20.866 33.770 20.866 33.770 cp_fm_cholesky_invert 11 10.9 33.365 33.373 33.365 33.373 mp_waitall_1 103674 16.8 28.233 32.707 28.233 32.707 qs_ks_update_qs_env 128 7.6 0.001 0.001 32.280 32.321 qs_ot_get_p 128 10.4 0.001 0.001 26.744 26.786 qs_ot_get_derivative 117 11.6 0.002 0.002 24.759 24.804 qs_ot_p2m_diag 83 11.4 0.880 0.886 22.549 22.581 cp_dbcsr_syevd 83 12.4 0.006 0.006 20.730 20.732 multiply_cannon_metrocomm3 10028 15.6 0.024 0.025 19.406 20.587 ot_diis_step 117 11.6 0.022 0.022 19.651 19.651 make_m2s 5014 13.6 0.074 0.076 18.168 19.622 make_images 5014 14.6 3.055 3.253 17.684 19.140 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 18.969 19.078 apply_single 128 13.6 0.001 0.001 18.969 19.078 multiply_cannon_multrec 10028 15.6 10.258 12.076 18.054 18.133 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.498 17.499 cp_fm_diag_elpa_base 83 14.4 13.108 14.708 17.493 17.493 qs_rho_update_rho_low 128 7.7 0.001 0.001 17.308 17.319 calculate_rho_elec 128 8.7 0.478 0.479 17.307 17.318 sum_up_and_integrate 128 10.3 0.002 0.002 16.426 16.512 integrate_v_rspace 128 11.3 0.004 0.004 16.363 16.450 multiply_cannon_sync_h2d 10028 15.6 14.298 14.315 14.298 14.315 pw_transfer 1547 11.6 0.095 0.096 13.198 13.208 fft_wrap_pw1pw2 1291 12.7 0.012 0.012 12.961 12.972 hybrid_alltoall_any 5200 16.5 1.310 3.038 10.259 12.352 init_scf_run 11 5.9 0.000 0.001 12.252 12.253 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.252 12.253 make_images_data 5014 15.6 0.063 0.068 10.088 12.224 fft_wrap_pw1pw2_140 523 13.2 3.196 3.232 11.510 11.519 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.873 9.911 dbcsr_mm_accdrv_process 20762 16.1 4.270 6.001 7.546 9.429 cp_fm_cholesky_decompose 22 10.9 8.934 8.951 8.934 8.951 wfi_extrapolate 11 7.9 0.001 0.001 8.938 8.938 grid_integrate_task_list 128 12.3 8.553 8.741 8.553 8.741 density_rs2pw 128 9.7 0.005 0.005 8.348 8.397 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.144 8.146 fft3d_ps 1291 14.7 3.969 3.986 7.446 7.478 mp_alltoall_d11v 2415 14.1 6.782 6.972 6.782 6.972 calculate_dm_sparse 128 9.5 0.001 0.001 6.822 6.882 grid_collocate_task_list 128 9.7 6.412 6.489 6.412 6.489 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.363 6.406 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.117 6.181 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=291.233000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2720.181818, yerr=154.703128 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 1.260589E+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 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 0.055 0.116 85.550 85.557 qs_energies 1 2.0 0.000 0.001 84.869 84.880 ls_scf 1 3.0 0.008 0.041 83.908 83.920 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.609 72.783 multiply_cannon 111 7.7 0.022 0.026 55.828 57.186 multiply_cannon_loop 111 8.7 0.227 0.243 52.435 54.105 ls_scf_main 1 4.0 0.003 0.036 52.370 52.377 density_matrix_trs4 2 5.0 0.002 0.006 46.836 46.892 ls_scf_init_scf 1 4.0 0.009 0.045 28.481 28.486 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.137 27.207 mp_waitall_1 11031 10.9 22.538 25.770 22.538 25.770 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.941 24.961 multiply_cannon_multrec 2664 9.7 8.141 8.955 15.539 17.311 multiply_cannon_sync_h2d 2664 9.7 13.486 14.872 13.486 14.872 make_m2s 222 7.7 0.009 0.012 13.139 13.648 make_images 222 8.7 0.098 0.108 13.116 13.628 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.747 12.903 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.437 8.274 make_images_data 222 9.7 0.004 0.005 7.683 8.261 dbcsr_mm_accdrv_process 4760 10.4 0.595 0.695 7.016 7.956 hybrid_alltoall_any 227 10.6 0.215 1.848 6.603 7.831 dbcsr_mm_accdrv_process_sort 4760 11.4 6.200 7.089 6.200 7.089 calculate_norms 4752 9.8 5.498 6.130 5.498 6.130 mp_sum_l 887 5.1 3.141 5.254 3.141 5.254 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.063 5.199 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.323 3.982 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.059 3.669 mp_irecv_dv 6231 10.9 2.042 3.644 2.042 3.644 make_images_sizes 222 9.7 0.000 0.000 0.777 3.625 mp_alltoall_i44 222 10.7 0.777 3.624 0.777 3.624 arnoldi_extremal 4 6.8 0.000 0.000 3.264 3.289 arnoldi_normal_ev 4 7.8 0.001 0.002 3.264 3.289 build_subspace 16 8.4 0.009 0.012 3.129 3.131 ls_scf_post 1 4.0 0.004 0.021 3.049 3.058 ls_scf_store_result 1 5.0 0.000 0.000 2.868 2.914 dbcsr_special_finalize 555 9.7 0.005 0.006 2.326 2.846 dbcsr_merge_single_wm 555 10.7 0.462 0.594 2.318 2.838 make_images_pack 222 9.7 2.206 2.639 2.208 2.640 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.328 2.576 dbcsr_sort_data 658 11.4 2.114 2.551 2.114 2.551 dbcsr_matrix_vector_mult_local 304 10.0 2.066 2.471 2.068 2.473 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.313 2.387 buffer_matrices_ensure_size 222 8.7 1.751 2.129 1.751 2.129 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.945 1.948 rebuild_ks_matrix 3 7.3 0.000 0.000 1.935 1.939 qs_ks_build_kohn_sham_matrix 3 8.3 0.010 0.024 1.935 1.939 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.557000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1132.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 2.139341E+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 0.024 0.040 90.547 90.548 qs_energies 1 2.0 0.000 0.001 90.064 90.070 ls_scf 1 3.0 0.000 0.001 88.680 88.687 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.806 75.111 multiply_cannon 111 7.7 0.028 0.040 52.811 57.385 ls_scf_main 1 4.0 0.000 0.001 55.086 55.089 multiply_cannon_loop 111 8.7 0.135 0.145 49.951 53.631 density_matrix_trs4 2 5.0 0.002 0.004 49.410 49.635 mp_waitall_1 9105 10.9 20.951 31.014 20.951 31.014 ls_scf_init_scf 1 4.0 0.000 0.001 29.889 29.891 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.678 28.795 multiply_cannon_multrec 1332 9.7 13.314 17.021 22.591 27.424 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.327 26.341 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.571 21.348 make_m2s 222 7.7 0.006 0.007 15.271 15.896 make_images 222 8.7 1.368 1.697 15.241 15.867 dbcsr_mm_accdrv_process 4041 10.4 0.333 0.533 8.870 10.419 dbcsr_mm_accdrv_process_sort 4041 11.4 8.390 9.885 8.390 9.885 make_images_data 222 9.7 0.004 0.005 8.869 9.784 hybrid_alltoall_any 227 10.6 0.542 2.572 8.188 9.671 mp_sum_l 887 5.1 5.268 8.753 5.268 8.753 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.198 7.576 mp_irecv_dv 3311 11.0 3.178 7.523 3.178 7.523 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.014 6.943 calculate_norms 2376 9.8 6.063 6.702 6.063 6.702 multiply_cannon_sync_h2d 1332 9.7 4.881 6.146 4.881 6.146 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.907 5.153 arnoldi_extremal 4 6.8 0.000 0.000 4.667 4.686 arnoldi_normal_ev 4 7.8 0.001 0.004 4.667 4.686 build_subspace 16 8.4 0.014 0.021 4.417 4.419 ls_scf_post 1 4.0 0.000 0.001 3.705 3.711 ls_scf_store_result 1 5.0 0.000 0.000 3.402 3.529 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.147 3.380 dbcsr_matrix_vector_mult_local 304 10.0 2.741 3.224 2.743 3.225 mp_allgather_i34 111 8.7 0.845 3.112 0.845 3.112 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.602 2.686 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.154 2.590 dbcsr_data_new 4174 10.1 2.112 2.409 2.112 2.409 make_images_pack 222 9.7 1.816 2.128 1.819 2.130 dbcsr_sort_data 436 11.2 1.840 2.123 1.840 2.123 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.853 1.855 rebuild_ks_matrix 3 7.3 0.000 0.000 1.840 1.842 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.840 1.842 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.548000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1785.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 2.904404E+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 0.028 0.040 92.821 92.822 qs_energies 1 2.0 0.000 0.000 92.276 92.279 ls_scf 1 3.0 0.000 0.000 90.832 90.836 dbcsr_multiply_generic 111 6.7 0.016 0.018 75.560 75.880 ls_scf_main 1 4.0 0.000 0.000 56.814 56.818 multiply_cannon 111 7.7 0.032 0.084 52.114 55.992 multiply_cannon_loop 111 8.7 0.116 0.130 49.348 53.770 density_matrix_trs4 2 5.0 0.002 0.003 50.927 51.165 mp_waitall_1 7281 11.0 23.764 33.023 23.764 33.023 ls_scf_init_scf 1 4.0 0.000 0.000 30.351 30.353 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.139 29.235 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.731 26.747 multiply_cannon_multrec 888 9.7 12.637 15.477 21.190 24.620 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.260 22.507 make_m2s 222 7.7 0.006 0.007 16.498 17.263 make_images 222 8.7 1.581 1.849 16.460 17.223 make_images_data 222 9.7 0.004 0.004 9.626 10.685 hybrid_alltoall_any 227 10.6 0.640 2.939 9.165 10.400 dbcsr_mm_accdrv_process 3754 10.4 0.315 0.485 8.088 9.306 mp_sum_l 887 5.1 5.237 9.146 5.237 9.146 dbcsr_mm_accdrv_process_sort 3754 11.4 7.648 8.821 7.648 8.821 multiply_cannon_sync_h2d 888 9.7 6.032 7.405 6.032 7.405 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.108 7.293 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.507 7.194 mp_irecv_dv 2335 11.1 2.490 7.133 2.490 7.133 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.521 6.631 arnoldi_extremal 4 6.8 0.000 0.000 5.084 5.103 arnoldi_normal_ev 4 7.8 0.001 0.005 5.084 5.102 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.740 5.008 calculate_norms 1584 9.8 4.377 4.827 4.377 4.827 build_subspace 16 8.4 0.014 0.020 4.786 4.792 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.440 3.771 mp_allgather_i34 111 8.7 0.878 3.770 0.878 3.770 ls_scf_post 1 4.0 0.000 0.000 3.668 3.673 dbcsr_matrix_vector_mult_local 304 10.0 3.022 3.593 3.024 3.595 ls_scf_store_result 1 5.0 0.000 0.000 3.418 3.493 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.830 2.930 dbcsr_data_new 4116 9.9 2.106 2.475 2.106 2.475 make_images_sizes 222 9.7 0.000 0.000 1.051 2.188 mp_alltoall_i44 222 10.7 1.050 2.187 1.050 2.187 dbcsr_sort_data 325 11.1 1.873 2.165 1.873 2.165 dbcsr_finalize 304 7.8 0.026 0.031 1.610 1.907 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.881 1.883 make_images_pack 222 9.7 1.622 1.877 1.625 1.880 rebuild_ks_matrix 3 7.3 0.000 0.000 1.863 1.864 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.863 1.864 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.822000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2222.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 3.335582E+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 0.042 97.582 97.583 qs_energies 1 2.0 0.000 0.000 96.986 96.991 ls_scf 1 3.0 0.000 0.000 95.311 95.316 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.959 79.237 ls_scf_main 1 4.0 0.000 0.000 59.270 59.271 multiply_cannon 111 7.7 0.039 0.073 51.777 56.305 density_matrix_trs4 2 5.0 0.002 0.003 53.149 53.270 multiply_cannon_loop 111 8.7 0.154 0.166 46.750 50.136 ls_scf_init_scf 1 4.0 0.000 0.000 32.821 32.823 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.599 31.668 mp_waitall_1 6369 11.0 22.881 29.025 22.881 29.025 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.995 29.007 multiply_cannon_multrec 1332 9.7 14.103 17.749 22.009 24.713 make_m2s 222 7.7 0.007 0.008 21.196 22.606 make_images 222 8.7 3.138 3.596 21.146 22.557 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.166 17.569 make_images_data 222 9.7 0.004 0.004 11.818 13.430 hybrid_alltoall_any 227 10.6 0.797 3.758 11.231 12.842 dbcsr_mm_accdrv_process 3641 10.4 0.294 0.482 7.550 9.042 mp_sum_l 887 5.1 4.387 8.674 4.387 8.674 dbcsr_mm_accdrv_process_sort 3641 11.4 7.095 8.546 7.095 8.546 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.409 7.055 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.113 6.127 mp_irecv_dv 3229 10.9 2.087 6.045 2.087 6.045 multiply_cannon_sync_h2d 1332 9.7 5.448 5.987 5.448 5.987 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.728 5.891 arnoldi_extremal 4 6.8 0.000 0.000 5.237 5.253 arnoldi_normal_ev 4 7.8 0.001 0.005 5.237 5.253 build_subspace 16 8.4 0.014 0.021 4.893 4.903 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.584 4.770 mp_allgather_i34 111 8.7 2.154 4.503 2.154 4.503 calculate_norms 2376 9.8 4.183 4.502 4.183 4.502 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.596 3.903 dbcsr_matrix_vector_mult_local 304 10.0 3.192 3.700 3.194 3.702 dbcsr_sort_data 658 11.4 3.108 3.491 3.108 3.491 dbcsr_special_finalize 555 9.7 0.006 0.007 2.850 3.259 dbcsr_merge_single_wm 555 10.7 0.541 0.673 2.841 3.251 ls_scf_post 1 4.0 0.000 0.000 3.219 3.223 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.022 3.081 ls_scf_store_result 1 5.0 0.000 0.000 2.959 3.023 dbcsr_data_release 10477 10.7 1.581 2.370 1.581 2.370 dbcsr_finalize 304 7.8 0.049 0.061 1.811 1.984 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.583000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2752.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 4.734775E+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 0.035 0.051 92.482 92.482 qs_energies 1 2.0 0.000 0.000 91.759 91.764 ls_scf 1 3.0 0.000 0.000 89.800 89.804 dbcsr_multiply_generic 111 6.7 0.017 0.019 71.345 71.561 ls_scf_main 1 4.0 0.000 0.000 57.130 57.131 multiply_cannon 111 7.7 0.069 0.105 52.977 56.199 multiply_cannon_loop 111 8.7 0.088 0.092 50.382 51.758 density_matrix_trs4 2 5.0 0.002 0.003 50.170 50.234 ls_scf_init_scf 1 4.0 0.000 0.000 29.304 29.305 mp_waitall_1 5436 11.0 25.091 29.101 25.091 29.101 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.062 28.094 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.948 25.959 multiply_cannon_multrec 444 9.7 13.678 16.145 20.769 22.054 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.310 16.409 make_m2s 222 7.7 0.005 0.005 13.708 14.673 make_images 222 8.7 2.037 2.479 13.640 14.602 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.101 13.829 hybrid_alltoall_any 227 10.6 0.803 3.828 8.296 9.964 make_images_data 222 9.7 0.003 0.004 8.501 9.944 multiply_cannon_sync_h2d 444 9.7 6.712 8.390 6.712 8.390 dbcsr_mm_accdrv_process 3003 10.4 0.358 0.409 6.779 7.865 dbcsr_mm_accdrv_process_sort 3003 11.4 6.409 7.470 6.409 7.470 arnoldi_extremal 4 6.8 0.000 0.000 5.875 5.889 arnoldi_normal_ev 4 7.8 0.001 0.004 5.875 5.889 build_subspace 16 8.4 0.015 0.019 5.484 5.497 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.402 4.545 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.183 4.382 mp_sum_l 887 5.1 2.665 4.212 2.665 4.212 dbcsr_matrix_vector_mult_local 304 10.0 3.706 4.173 3.708 4.175 mp_allgather_i34 111 8.7 1.166 3.788 1.166 3.788 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.546 3.750 calculate_norms 792 9.8 3.604 3.725 3.604 3.725 mp_irecv_dv 1241 11.2 1.533 3.717 1.533 3.717 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 3.366 3.371 make_images_sizes 222 9.7 0.000 0.000 0.825 3.332 mp_alltoall_i44 222 10.7 0.825 3.332 0.825 3.332 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.877 3.242 ls_scf_store_result 1 5.0 0.000 0.000 3.156 3.189 dbcsr_finalize 304 7.8 0.062 0.077 2.195 2.290 dbcsr_data_new 4608 9.7 1.808 2.246 1.808 2.246 dbcsr_merge_all 275 8.9 0.479 0.521 2.054 2.129 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.034 2.035 rebuild_ks_matrix 3 7.3 0.000 0.000 2.001 2.003 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.001 2.003 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 1.943 1.943 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.482000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3743.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 8.816509E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.070 0.088 105.510 105.510 qs_energies 1 2.0 0.000 0.000 104.114 104.118 ls_scf 1 3.0 0.000 0.000 101.108 101.112 dbcsr_multiply_generic 111 6.7 0.024 0.027 74.690 74.842 ls_scf_main 1 4.0 0.000 0.000 63.664 63.676 density_matrix_trs4 2 5.0 0.002 0.003 54.619 54.672 multiply_cannon 111 7.7 0.151 0.263 48.595 50.884 multiply_cannon_loop 111 8.7 0.098 0.099 45.468 46.069 ls_scf_init_scf 1 4.0 0.000 0.000 33.667 33.679 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.096 32.122 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.333 29.340 mp_waitall_1 4527 11.1 21.780 25.753 21.780 25.753 make_m2s 222 7.7 0.005 0.005 22.517 23.712 make_images 222 8.7 3.570 3.882 22.409 23.602 multiply_cannon_multrec 444 9.7 17.853 18.516 22.485 23.153 hybrid_alltoall_any 227 10.6 1.653 3.624 12.756 15.536 make_images_data 222 9.7 0.003 0.004 12.935 15.148 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.236 10.645 multiply_cannon_sync_h2d 444 9.7 8.794 8.848 8.794 8.848 arnoldi_extremal 4 6.8 0.000 0.000 7.382 7.387 arnoldi_normal_ev 4 7.8 0.002 0.009 7.382 7.387 build_subspace 16 8.4 0.026 0.036 6.843 6.851 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.467 5.627 dbcsr_matrix_vector_mult_local 304 10.0 5.053 5.379 5.055 5.381 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.170 5.259 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.902 5.152 dbcsr_mm_accdrv_process 1814 10.4 0.243 0.353 4.455 4.573 dbcsr_mm_accdrv_process_sort 1814 11.4 4.116 4.246 4.116 4.246 mp_allgather_i34 111 8.7 1.229 3.907 1.229 3.907 ls_scf_post 1 4.0 0.000 0.000 3.777 3.782 make_images_sizes 222 9.7 0.000 0.000 1.448 3.538 mp_alltoall_i44 222 10.7 1.448 3.538 1.448 3.538 ls_scf_store_result 1 5.0 0.000 0.000 3.496 3.531 calculate_norms 792 9.8 3.233 3.278 3.233 3.278 dbcsr_finalize 304 7.8 0.082 0.089 3.077 3.182 qs_energies_init_hamiltonians 1 3.0 0.018 0.025 2.975 2.975 dbcsr_merge_all 275 8.9 0.887 0.918 2.863 2.961 dbcsr_complete_redistribute 5 7.6 1.439 1.481 2.741 2.861 matrix_ls_to_qs 2 6.0 0.000 0.000 2.399 2.524 dbcsr_sort_data 325 11.1 2.440 2.508 2.440 2.508 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.470 2.483 rebuild_ks_matrix 3 7.3 0.000 0.000 2.402 2.415 qs_ks_build_kohn_sham_matrix 3 8.3 0.013 0.025 2.402 2.415 dbcsr_data_new 6591 9.6 1.868 2.409 1.868 2.409 dbcsr_new_transposed 4 7.5 0.241 0.252 2.287 2.308 dbcsr_frobenius_norm 74 6.6 2.056 2.135 2.204 2.239 dbcsr_add_d 103 6.2 0.000 0.000 2.127 2.207 dbcsr_add_anytype 103 7.2 0.859 0.891 2.126 2.207 dbcsr_data_release 12724 10.6 1.975 2.187 1.975 2.187 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.510000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6998.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e1b64ec72d8c911e7967845d04dd1ccfaacfc398_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 589.484032E+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 1214887567. 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 0.023 0.079 242.170 242.177 qs_mol_dyn_low 1 2.0 0.006 0.072 241.113 241.174 qs_forces 5 3.8 0.005 0.048 240.932 240.935 qs_energies 5 4.8 0.004 0.033 237.653 237.682 scf_env_do_scf 5 5.8 0.001 0.003 223.243 223.246 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 194.466 194.468 qs_scf_new_mos 105 7.6 0.000 0.001 151.325 151.514 qs_scf_loop_do_ot 105 8.6 0.001 0.001 151.325 151.514 ot_scf_mini 105 9.6 0.003 0.004 141.475 141.617 dbcsr_multiply_generic 1445 12.2 0.129 0.137 133.768 134.234 multiply_cannon 1445 13.2 0.276 0.286 113.913 116.013 multiply_cannon_loop 1445 14.2 2.807 2.973 111.955 113.309 velocity_verlet 4 3.0 0.001 0.024 109.576 109.578 ot_mini 105 10.6 0.001 0.001 61.045 61.157 qs_ot_get_p 112 10.4 0.001 0.002 50.243 50.521 multiply_cannon_multrec 69360 15.2 29.626 34.713 39.500 44.235 mp_waitall_1 488190 16.1 35.855 42.695 35.855 42.695 qs_ot_get_derivative 55 11.6 0.001 0.001 39.108 39.241 qs_ot_p2m_diag 40 11.0 0.020 0.030 38.774 38.887 cp_dbcsr_syevd 40 12.0 0.002 0.003 35.387 35.388 multiply_cannon_sync_h2d 69360 15.2 29.170 33.222 29.170 33.222 multiply_cannon_metrocomm3 69360 15.2 0.199 0.211 25.832 33.220 rebuild_ks_matrix 110 8.4 0.000 0.001 30.208 30.368 qs_ks_build_kohn_sham_matrix 110 9.4 0.018 0.074 30.207 30.367 cp_fm_syevd 40 13.0 0.000 0.002 29.864 30.020 init_scf_loop 7 6.6 0.000 0.005 28.735 28.737 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.649 27.802 cp_fm_redistribute_end 40 14.0 12.451 24.830 12.457 24.833 cp_fm_syevd_base 40 14.0 12.365 24.746 12.365 24.746 prepare_preconditioner 7 7.6 0.000 0.000 23.536 23.576 make_preconditioner 7 8.6 0.000 0.003 23.536 23.576 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.130 23.418 apply_single 62 13.6 0.000 0.000 23.130 23.418 ot_new_cg_direction 55 11.6 0.001 0.001 21.206 21.206 qs_rho_update_rho_low 110 7.6 0.001 0.001 19.195 19.488 calculate_rho_elec 110 8.6 0.030 0.049 19.195 19.488 make_full_inverse_cholesky 7 9.6 0.000 0.000 16.007 16.066 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.977 15.176 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.971 14.092 density_rs2pw 110 9.6 0.005 0.006 13.343 13.760 mp_sum_l 4764 12.2 12.363 13.086 12.363 13.086 pw_transfer 1645 12.4 0.080 0.104 12.049 12.381 fft_wrap_pw1pw2 1425 13.5 0.012 0.016 11.910 12.240 calculate_dm_sparse 110 9.5 0.000 0.001 11.419 11.701 init_scf_run 5 5.8 0.000 0.001 11.542 11.543 scf_env_initial_rho_setup 5 6.8 0.003 0.048 11.542 11.543 cp_fm_cholesky_invert 7 10.6 11.238 11.251 11.238 11.251 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.695 10.768 fft_wrap_pw1pw2_240 915 15.0 0.828 0.914 10.399 10.681 qs_vxc_create 110 10.4 0.008 0.082 10.550 10.601 dbcsr_mm_accdrv_process 154766 15.8 6.265 6.470 9.742 10.599 transfer_rs2pw 445 10.6 0.007 0.009 9.902 10.351 check_diag 80 13.5 8.594 8.889 9.989 10.144 sum_up_and_integrate 60 10.3 0.001 0.003 9.536 9.602 integrate_v_rspace 60 11.3 0.002 0.003 9.519 9.586 fft3d_pb 915 16.0 2.399 2.697 8.593 9.010 xc_rho_set_and_dset_create 110 12.4 0.078 0.155 7.374 7.655 make_m2s 2890 13.2 0.078 0.086 7.020 7.648 make_images 2890 14.2 0.237 0.259 6.913 7.540 acc_transpose_blocks 69360 15.2 0.355 0.371 7.010 7.333 make_full_single_inverse 7 9.6 0.001 0.027 7.238 7.278 multiply_cannon_metrocomm1 69360 15.2 0.095 0.100 4.537 7.242 xc_vxc_pw_create 60 11.3 0.039 0.056 7.082 7.137 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 6.938 6.954 calculate_first_density_matrix 1 7.0 0.001 0.017 6.580 6.609 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.450 6.525 xc_pw_derive 510 13.4 0.005 0.006 6.264 6.352 mp_alltoall_z22v 2340 17.7 5.642 6.019 5.642 6.019 potential_pw2rs 60 12.3 0.002 0.003 5.707 5.825 multiply_cannon_metrocomm4 67915 15.2 0.188 0.204 2.021 5.165 mp_waitany 7680 13.5 4.503 5.093 4.503 5.093 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=242.177000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=561.600000, yerr=1.959592 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: e1b64ec72d8c911e7967845d04dd1ccfaacfc398 Summary: empty Status: OK