=== 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: bf5cdd7210383fb7939885873c6db45d8a334c42 ################# 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 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # 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: 21.06.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/01 job id: 47900705 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/02 job id: 47900706 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/03 job id: 47900707 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/04 job id: 47900708 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/05 job id: 47900709 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/06 job id: 47900710 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/07 job id: 47900711 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/08 job id: 47900712 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/09 job id: 47900713 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/10 job id: 47900714 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/11 job id: 47900715 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/12 job id: 47900717 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/13 job id: 47900719 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/14 job id: 47900721 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/15 job id: 47900723 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/16 job id: 47900727 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/17 job id: 47900728 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/18 job id: 47900729 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/19 job id: 47900730 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/20 job id: 47900732 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/21 job id: 47900733 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/22 job id: 47900734 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/23 job id: 47900735 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/24 job id: 47900736 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/25 job id: 47900737 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/26 job id: 47900738 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/27 job id: 47900740 --- 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/bf5cdd7210383fb7939885873c6db45d8a334c42_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 0.021 0.033 133.992 133.993 farming_run 1 2.0 133.386 133.396 133.965 133.968 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.489125E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.025 115.243 115.244 qs_energies 1 2.0 0.000 0.000 115.057 115.058 mp2_main 1 3.0 0.000 0.000 113.003 113.004 mp2_gpw_main 1 4.0 0.027 0.033 112.002 112.004 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 92.930 93.378 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.001 55.450 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.164 41.400 46.532 get_2c_integrals 1 6.0 0.008 0.009 37.058 37.758 integrate_v_rspace 273 8.0 0.441 0.455 24.793 29.702 pw_transfer 6555 10.6 0.373 0.387 27.430 27.977 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.071 26.674 grid_integrate_task_list 273 9.0 20.590 25.995 20.590 25.995 fft_wrap_pw1pw2_100 2178 12.4 1.180 1.429 23.601 24.207 compute_2c_integrals 1 7.0 0.002 0.002 19.788 19.790 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.877 19.526 mp2_eri_2c_integrate_gpw 1 9.0 2.375 2.429 18.874 19.523 rpa_ri_compute_en 1 5.0 0.019 0.020 18.955 19.288 cp_fm_cholesky_decompose 12 8.2 17.200 17.921 17.200 17.921 cholesky_decomp 1 7.0 0.000 0.000 16.119 16.844 fft3d_s 5443 13.4 16.193 16.697 16.214 16.719 ao_to_mo_and_store_B_mult_1 272 7.0 10.764 15.305 10.764 15.305 calculate_wavefunction 272 8.0 5.396 5.563 12.520 13.128 rpa_num_int 1 6.0 0.000 0.000 10.604 10.613 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.588 10.602 calc_mat_Q 8 8.0 0.000 0.000 9.462 9.550 contract_S_to_Q 8 9.0 0.000 0.000 8.883 8.969 calc_potential_gpw 544 9.5 0.005 0.005 8.250 8.727 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.213 8.609 parallel_gemm_fm 14 9.1 0.000 0.000 8.463 8.561 parallel_gemm_fm_cosma 14 10.1 8.463 8.561 8.463 8.561 potential_pw2rs 545 10.0 0.107 0.109 7.711 8.345 collocate_single_gaussian 272 10.0 0.039 0.041 7.478 7.828 create_integ_mat 1 6.0 0.014 0.027 7.811 7.820 array2fm 1 7.0 0.000 0.000 6.708 7.203 pw_scatter_s 2720 13.7 4.429 4.639 4.429 4.639 pw_gather_s 2722 13.2 3.867 4.072 3.867 4.072 array2fm_buffer_send 1 8.0 2.970 3.164 2.970 3.164 pw_poisson_solve 545 10.5 1.094 1.140 2.162 2.410 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.003704, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2799.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_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 0.029 0.039 396.488 396.489 farming_run 1 2.0 395.653 395.659 396.444 396.447 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.234588E+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 2369023. 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 0.009 0.034 209.154 209.155 qs_energies 1 2.0 0.000 0.000 208.928 208.941 scf_env_do_scf 1 3.0 0.000 0.000 106.671 106.671 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.764 105.772 rebuild_ks_matrix 4 6.0 0.000 0.000 105.763 105.771 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.063 105.763 105.771 hfx_ks_matrix 4 8.0 0.001 0.001 105.387 105.391 integrate_four_center 4 9.0 0.143 0.453 105.386 105.390 mp2_main 1 3.0 0.000 0.000 101.965 101.977 mp2_gpw_main 1 4.0 0.049 0.073 101.126 101.139 integrate_four_center_main 4 10.0 0.096 0.528 96.621 100.000 integrate_four_center_bin 266 11.0 96.526 99.995 96.526 99.995 init_scf_loop 1 4.0 0.000 0.000 92.368 92.369 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.496 75.580 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.052 55.136 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.160 41.793 46.810 integrate_v_rspace 95 8.0 0.397 0.564 28.163 32.966 pw_transfer 2240 10.6 0.143 0.177 29.923 30.413 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.923 29.432 ao_to_mo_and_store_B_mult_1 91 7.0 10.573 29.179 10.573 29.179 mp2_ri_gpw_compute_en 1 5.0 0.056 0.069 26.467 28.598 grid_integrate_task_list 95 9.0 23.464 28.484 23.464 28.484 fft_wrap_pw1pw2_100 730 12.4 1.284 1.475 26.631 27.087 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.839 1.896 24.735 24.745 get_2c_integrals 1 6.0 0.000 0.000 20.360 20.380 compute_2c_integrals 1 7.0 0.002 0.003 19.344 19.347 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.885 19.224 mp2_eri_2c_integrate_gpw 1 9.0 1.734 1.908 18.884 19.222 fft3d_s 1823 13.4 18.442 18.893 18.455 18.906 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.301 14.301 calculate_wavefunction 91 8.0 2.015 2.051 9.741 9.983 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.588 8.698 9.346 potential_pw2rs 186 10.0 0.033 0.035 8.619 9.165 local_gemm 172 8.0 8.143 8.787 8.143 8.787 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.259 8.532 mp2_ri_gpw_compute_en_comm 22 7.0 0.502 0.524 7.793 8.342 calc_potential_gpw 182 9.5 0.002 0.002 7.938 8.200 collocate_single_gaussian 91 10.0 0.017 0.023 7.898 8.137 mp2_ri_gpw_compute_en_ener 172 7.0 6.360 6.468 6.360 6.468 mp_sendrecv_dm3 2068 8.0 5.822 6.377 5.822 6.377 mp_sync 37 10.5 3.420 5.856 3.420 5.856 pw_gather_s 912 13.2 4.887 5.341 4.887 5.341 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.128710, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1494.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.038656E+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 588782. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.026 53.723 53.723 qs_mol_dyn_low 1 2.0 0.003 0.005 53.522 53.531 qs_forces 11 3.9 0.002 0.002 53.459 53.460 qs_energies 11 4.9 0.002 0.002 51.966 51.982 scf_env_do_scf 11 5.9 0.000 0.001 46.037 46.037 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 44.002 44.002 qs_scf_new_mos 108 7.5 0.000 0.001 33.930 34.232 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.929 34.232 dbcsr_multiply_generic 2286 12.5 0.094 0.099 33.669 34.090 ot_scf_mini 108 9.5 0.002 0.002 32.228 32.408 multiply_cannon 2286 13.5 0.191 0.197 26.294 27.816 multiply_cannon_loop 2286 14.5 1.831 1.949 25.655 27.211 velocity_verlet 10 3.0 0.002 0.005 26.397 26.398 ot_mini 108 10.5 0.001 0.001 19.527 19.771 qs_ot_get_derivative 108 11.5 0.001 0.001 16.533 16.709 mp_waitall_1 245248 16.5 8.522 14.576 8.522 14.576 multiply_cannon_metrocomm3 54864 15.5 0.072 0.078 6.126 12.877 multiply_cannon_multrec 54864 15.5 3.643 5.766 7.568 10.922 qs_ot_get_p 119 10.4 0.001 0.001 8.090 8.381 rebuild_ks_matrix 119 8.3 0.000 0.000 7.819 7.971 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.819 7.971 mp_sum_l 7287 12.8 5.503 7.119 5.503 7.119 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.875 7.012 multiply_cannon_sync_h2d 54864 15.5 5.174 6.814 5.174 6.814 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.839 6.282 dbcsr_mm_accdrv_process 76910 16.1 1.828 2.885 3.838 5.516 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.312 5.436 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.273 5.310 init_scf_run 11 5.9 0.000 0.001 4.663 4.663 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.662 4.663 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.469 4.469 sum_up_and_integrate 119 10.3 0.012 0.014 4.418 4.424 integrate_v_rspace 119 11.3 0.002 0.003 4.406 4.413 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.177 4.321 calculate_rho_elec 119 8.7 0.012 0.017 4.176 4.320 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.281 4.281 cp_fm_redistribute_end 50 14.0 2.184 4.258 2.190 4.261 cp_fm_diag_elpa_base 50 14.0 2.065 4.158 2.069 4.164 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.915 3.098 apply_single 119 13.6 0.000 0.000 2.915 3.097 calculate_dm_sparse 119 9.5 0.000 0.001 2.914 3.078 multiply_cannon_metrocomm1 54864 15.5 0.055 0.060 1.803 3.004 acc_transpose_blocks 54864 15.5 0.229 0.252 2.229 2.765 rs_pw_transfer 974 11.9 0.012 0.013 2.632 2.741 ot_diis_step 108 11.5 0.006 0.006 2.684 2.685 jit_kernel_multiply 13 15.8 1.946 2.549 1.946 2.549 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.431 2.501 calculate_first_density_matrix 1 7.0 0.000 0.000 2.427 2.431 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.295 2.297 density_rs2pw 119 9.7 0.004 0.005 2.131 2.272 wfi_extrapolate 11 7.9 0.001 0.001 2.166 2.166 grid_integrate_task_list 119 12.3 2.016 2.109 2.016 2.109 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.978 2.047 init_scf_loop 11 6.9 0.000 0.001 2.019 2.020 mp_sum_d 4135 12.0 1.370 1.942 1.370 1.942 potential_pw2rs 119 12.3 0.004 0.004 1.777 1.783 pw_transfer 1439 11.6 0.053 0.057 1.655 1.721 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.579 1.645 make_m2s 4572 13.5 0.053 0.056 1.554 1.595 make_images 4572 14.5 0.133 0.140 1.472 1.513 acc_transpose_blocks_sync 164592 16.5 1.209 1.456 1.209 1.456 mp_alltoall_d11v 2130 13.8 1.228 1.428 1.228 1.428 grid_collocate_task_list 119 9.7 1.356 1.425 1.356 1.425 mp_waitany 12084 13.8 1.224 1.411 1.224 1.411 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.379 1.407 fft_wrap_pw1pw2_140 487 13.2 0.189 0.206 1.221 1.289 fft3d_ps 1201 14.6 0.367 0.473 1.216 1.281 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.746 1.142 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.723000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.153664E+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 596307. 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 0.052 0.108 39.528 39.529 qs_mol_dyn_low 1 2.0 0.003 0.005 39.080 39.088 qs_forces 11 3.9 0.003 0.008 39.016 39.017 qs_energies 11 4.9 0.001 0.005 37.297 37.301 scf_env_do_scf 11 5.9 0.001 0.002 32.163 32.163 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.727 29.728 dbcsr_multiply_generic 2286 12.5 0.102 0.107 21.997 22.346 qs_scf_new_mos 108 7.5 0.001 0.001 20.605 20.831 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.604 20.831 ot_scf_mini 108 9.5 0.003 0.003 19.686 19.850 velocity_verlet 10 3.0 0.001 0.002 18.553 18.554 multiply_cannon 2286 13.5 0.210 0.219 16.945 18.457 multiply_cannon_loop 2286 14.5 1.193 1.250 15.792 17.227 ot_mini 108 10.5 0.001 0.001 12.145 12.375 mp_waitall_1 200699 16.5 5.543 10.596 5.543 10.596 qs_ot_get_derivative 108 11.5 0.001 0.002 9.771 9.938 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.086 9.405 multiply_cannon_multrec 27432 15.5 1.834 4.064 6.150 8.951 rebuild_ks_matrix 119 8.3 0.000 0.000 7.231 7.372 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.231 7.372 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.369 6.497 dbcsr_mm_accdrv_process 47894 16.0 3.454 5.636 4.235 6.298 qs_ot_get_p 119 10.4 0.001 0.001 4.648 4.873 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.754 4.569 sum_up_and_integrate 119 10.3 0.025 0.029 4.112 4.121 integrate_v_rspace 119 11.3 0.002 0.003 4.087 4.097 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.977 4.071 apply_single 119 13.6 0.000 0.000 2.976 4.070 init_scf_run 11 5.9 0.000 0.001 3.890 3.890 scf_env_initial_rho_setup 11 6.9 0.001 0.003 3.890 3.890 mp_sum_l 7287 12.8 2.023 3.885 2.023 3.885 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.771 3.796 calculate_rho_elec 119 8.7 0.022 0.029 3.770 3.795 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.011 3.030 rs_pw_transfer 974 11.9 0.010 0.012 2.458 2.974 make_m2s 4572 13.5 0.053 0.054 2.610 2.951 make_images 4572 14.5 0.205 0.246 2.521 2.865 cp_dbcsr_syevd 50 12.0 0.003 0.005 2.615 2.616 density_rs2pw 119 9.7 0.004 0.005 2.019 2.548 init_scf_loop 11 6.9 0.001 0.006 2.415 2.417 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.306 2.390 calculate_first_density_matrix 1 7.0 0.000 0.003 2.363 2.365 ot_diis_step 108 11.5 0.011 0.014 2.325 2.325 cp_fm_diag_elpa 50 13.0 0.000 0.001 2.298 2.298 cp_fm_redistribute_end 50 14.0 1.163 2.271 1.166 2.274 cp_fm_diag_elpa_base 50 14.0 1.075 2.183 1.104 2.221 multiply_cannon_sync_h2d 27432 15.5 1.701 2.192 1.701 2.192 calculate_dm_sparse 119 9.5 0.000 0.001 2.066 2.149 pw_transfer 1439 11.6 0.066 0.070 1.919 1.951 grid_integrate_task_list 119 12.3 1.835 1.949 1.835 1.949 acc_transpose_blocks 27432 15.5 0.114 0.120 1.619 1.946 jit_kernel_multiply 10 16.3 0.722 1.907 0.722 1.907 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.879 1.881 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.827 1.860 potential_pw2rs 119 12.3 0.006 0.006 1.830 1.839 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.680 1.721 make_images_data 4572 15.5 0.048 0.053 1.232 1.595 prepare_preconditioner 11 7.9 0.000 0.000 1.474 1.501 make_preconditioner 11 8.9 0.000 0.001 1.474 1.501 fft_wrap_pw1pw2_140 487 13.2 0.206 0.218 1.460 1.492 wfi_extrapolate 11 7.9 0.001 0.001 1.452 1.453 hybrid_alltoall_any 4725 16.4 0.054 0.115 1.090 1.438 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.380 1.438 grid_collocate_task_list 119 9.7 1.286 1.435 1.286 1.435 fft3d_ps 1201 14.6 0.519 0.577 1.372 1.398 mp_alltoall_d11v 2130 13.8 1.176 1.338 1.176 1.338 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.256 1.265 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.192 1.239 rs_pw_transfer_RS2PW_140 130 11.5 0.141 0.149 0.567 1.079 mp_waitany 5720 13.7 0.541 1.053 0.541 1.053 mp_sum_d 4135 12.0 0.556 1.017 0.556 1.017 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.976 0.978 mp_allgather_i34 2286 14.5 0.571 0.978 0.571 0.978 acc_transpose_blocks_sync 82296 16.5 0.819 0.938 0.819 0.938 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.923 0.936 acc_transpose_blocks_kernels 27432 16.5 0.187 0.274 0.661 0.869 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.529000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.090909, yerr=1.729663 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 520.552448E+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 1901786. 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 0.039 0.090 33.363 33.364 qs_mol_dyn_low 1 2.0 0.003 0.004 33.008 33.016 qs_forces 11 3.9 0.006 0.021 32.563 32.564 qs_energies 11 4.9 0.001 0.005 30.973 30.978 scf_env_do_scf 11 5.9 0.000 0.001 25.801 25.801 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.381 23.382 dbcsr_multiply_generic 2286 12.5 0.096 0.099 17.119 17.191 velocity_verlet 10 3.0 0.001 0.001 15.639 15.652 qs_scf_new_mos 108 7.5 0.001 0.001 15.282 15.296 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.282 15.296 multiply_cannon 2286 13.5 0.194 0.201 13.952 14.709 ot_scf_mini 108 9.5 0.002 0.003 14.533 14.542 multiply_cannon_loop 2286 14.5 0.854 0.894 13.191 13.967 ot_mini 108 10.5 0.001 0.001 9.008 9.021 qs_ot_get_derivative 108 11.5 0.001 0.001 7.530 7.540 multiply_cannon_multrec 18288 15.5 1.875 2.886 7.173 7.522 rebuild_ks_matrix 119 8.3 0.000 0.000 6.330 6.344 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.329 6.344 dbcsr_mm_accdrv_process 38222 16.0 5.114 5.927 5.204 5.992 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.581 5.594 sum_up_and_integrate 119 10.3 0.031 0.032 3.892 3.896 integrate_v_rspace 119 11.3 0.003 0.003 3.861 3.869 init_scf_run 11 5.9 0.000 0.001 3.754 3.754 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.753 3.754 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.890 3.470 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.414 3.419 calculate_rho_elec 119 8.7 0.031 0.032 3.414 3.418 mp_waitall_1 158411 16.6 2.585 3.403 2.585 3.403 qs_ot_get_p 119 10.4 0.001 0.002 3.295 3.320 rs_pw_transfer 974 11.9 0.009 0.010 2.393 2.648 calculate_first_density_matrix 1 7.0 0.000 0.002 2.532 2.533 init_scf_loop 11 6.9 0.001 0.004 2.404 2.405 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.982 2.338 apply_single 119 13.6 0.000 0.000 1.982 2.338 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.482 2.292 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.215 2.221 density_rs2pw 119 9.7 0.004 0.004 1.910 2.207 calculate_dm_sparse 119 9.5 0.000 0.001 1.952 1.963 make_m2s 4572 13.5 0.045 0.049 1.783 1.929 grid_integrate_task_list 119 12.3 1.807 1.927 1.807 1.927 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.922 1.922 pw_transfer 1439 11.6 0.065 0.069 1.860 1.869 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.857 1.858 make_images 4572 14.5 0.194 0.207 1.697 1.843 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.767 1.779 acc_transpose_blocks 18288 15.5 0.077 0.080 1.691 1.745 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.712 1.722 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 1.687 1.695 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.684 1.694 cp_fm_diag_elpa_base 50 14.0 1.661 1.674 1.683 1.692 potential_pw2rs 119 12.3 0.007 0.008 1.660 1.664 prepare_preconditioner 11 7.9 0.000 0.000 1.609 1.612 make_preconditioner 11 8.9 0.000 0.001 1.609 1.612 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.470 1.552 ot_diis_step 108 11.5 0.011 0.011 1.462 1.462 fft_wrap_pw1pw2_140 487 13.2 0.256 0.262 1.436 1.446 mp_sum_l 7287 12.8 1.015 1.427 1.015 1.427 grid_collocate_task_list 119 9.7 1.242 1.422 1.242 1.422 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.298 1.301 multiply_cannon_sync_h2d 18288 15.5 1.071 1.267 1.071 1.267 fft3d_ps 1201 14.6 0.527 0.544 1.251 1.259 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 1.196 1.201 wfi_extrapolate 11 7.9 0.001 0.001 1.168 1.168 mp_waitany 9880 13.7 0.726 1.000 0.726 1.000 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.122 0.720 0.977 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.930 0.947 make_images_data 4572 15.5 0.047 0.051 0.765 0.929 acc_transpose_blocks_kernels 18288 16.5 0.216 0.223 0.859 0.892 hybrid_alltoall_any 4725 16.4 0.058 0.115 0.666 0.863 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.819 0.822 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.728 0.798 acc_transpose_blocks_sync 54864 16.5 0.735 0.791 0.735 0.791 mp_alltoall_d11v 2130 13.8 0.613 0.758 0.613 0.758 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.674 0.746 cp_fm_cholesky_invert 11 10.9 0.674 0.678 0.674 0.678 jit_kernel_transpose 5 15.6 0.643 0.669 0.643 0.669 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.364000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.272727, yerr=1.542778 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 556.281856E+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 0.018 0.031 36.413 36.414 qs_mol_dyn_low 1 2.0 0.003 0.004 36.194 36.205 qs_forces 11 3.9 0.003 0.012 36.134 36.135 qs_energies 11 4.9 0.005 0.010 34.382 34.394 scf_env_do_scf 11 5.9 0.001 0.002 29.322 29.326 scf_env_do_scf_inner_loop 108 6.5 0.003 0.009 25.955 25.958 dbcsr_multiply_generic 2286 12.5 0.100 0.104 19.606 19.760 velocity_verlet 10 3.0 0.002 0.002 18.438 18.440 qs_scf_new_mos 108 7.5 0.001 0.001 17.548 17.600 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.547 17.599 multiply_cannon 2286 13.5 0.218 0.225 16.081 16.934 ot_scf_mini 108 9.5 0.003 0.003 16.529 16.592 multiply_cannon_loop 2286 14.5 1.512 1.608 15.155 15.799 ot_mini 108 10.5 0.001 0.001 10.144 10.216 multiply_cannon_multrec 27432 15.5 2.454 3.132 8.986 9.261 qs_ot_get_derivative 108 11.5 0.001 0.001 8.291 8.357 dbcsr_mm_accdrv_process 47916 15.9 5.901 7.699 6.429 7.760 rebuild_ks_matrix 119 8.3 0.000 0.000 6.572 6.621 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.572 6.620 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.823 5.865 sum_up_and_integrate 119 10.3 0.035 0.037 3.749 3.757 integrate_v_rspace 119 11.3 0.003 0.004 3.713 3.722 init_scf_run 11 5.9 0.000 0.001 3.668 3.668 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.667 3.668 qs_ot_get_p 119 10.4 0.001 0.001 3.576 3.652 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.515 3.546 calculate_rho_elec 119 8.7 0.041 0.048 3.514 3.546 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.149 3.480 init_scf_loop 11 6.9 0.009 0.020 3.348 3.351 acc_transpose_blocks 27432 15.5 0.117 0.124 2.380 2.531 prepare_preconditioner 11 7.9 0.000 0.000 2.463 2.472 make_preconditioner 11 8.9 0.001 0.002 2.463 2.472 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.072 2.402 make_m2s 4572 13.5 0.054 0.056 2.219 2.334 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.282 2.313 calculate_first_density_matrix 1 7.0 0.001 0.003 2.278 2.280 make_images 4572 14.5 0.274 0.332 2.112 2.224 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.029 2.219 apply_single 119 13.6 0.000 0.000 2.028 2.219 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.187 2.199 mp_waitall_1 137007 16.6 1.741 2.185 1.741 2.185 calculate_dm_sparse 119 9.5 0.000 0.000 2.125 2.181 rs_pw_transfer 974 11.9 0.009 0.010 1.910 2.080 density_rs2pw 119 9.7 0.004 0.004 1.882 2.071 pw_transfer 1439 11.6 0.066 0.071 1.983 2.018 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.890 1.928 grid_integrate_task_list 119 12.3 1.832 1.927 1.832 1.927 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.824 1.825 ot_diis_step 108 11.5 0.012 0.012 1.811 1.812 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.787 1.788 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.660 1.674 fft_wrap_pw1pw2_140 487 13.2 0.292 0.312 1.572 1.610 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.555 1.563 cp_fm_diag_elpa_base 50 14.0 1.521 1.537 1.553 1.561 acc_transpose_blocks_sync 82296 16.5 1.407 1.550 1.407 1.550 potential_pw2rs 119 12.3 0.009 0.010 1.497 1.500 grid_collocate_task_list 119 9.7 1.252 1.390 1.252 1.390 fft3d_ps 1201 14.6 0.555 0.606 1.320 1.345 wfi_extrapolate 11 7.9 0.001 0.001 1.327 1.327 multiply_cannon_metrocomm3 27432 15.5 0.041 0.044 0.839 1.307 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.269 1.286 mp_sum_l 7287 12.8 0.881 1.272 0.881 1.272 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.237 1.254 cp_fm_upper_to_full 72 14.2 0.811 1.157 0.811 1.157 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.115 1.117 jit_kernel_multiply 6 16.2 0.462 1.087 0.462 1.087 dbcsr_complete_redistribute 329 12.2 0.124 0.149 0.744 1.017 make_images_data 4572 15.5 0.048 0.052 0.819 0.938 hybrid_alltoall_any 4725 16.4 0.066 0.155 0.693 0.881 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.800 0.877 mp_alltoall_d11v 2130 13.8 0.704 0.867 0.704 0.867 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.844 0.848 acc_transpose_blocks_kernels 27432 16.5 0.269 0.277 0.827 0.839 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.560 0.828 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.414000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.818182, yerr=2.690663 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_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 624.181248E+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 0.012 0.027 29.340 29.341 qs_mol_dyn_low 1 2.0 0.003 0.004 29.175 29.183 qs_forces 11 3.9 0.002 0.003 28.637 28.644 qs_energies 11 4.9 0.043 0.045 26.919 26.927 scf_env_do_scf 11 5.9 0.000 0.001 22.064 22.064 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.627 19.627 velocity_verlet 10 3.0 0.002 0.002 15.153 15.186 dbcsr_multiply_generic 2286 12.5 0.093 0.098 13.057 13.182 qs_scf_new_mos 108 7.5 0.001 0.001 11.719 11.745 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.719 11.744 ot_scf_mini 108 9.5 0.002 0.002 11.020 11.044 multiply_cannon 2286 13.5 0.222 0.230 10.482 10.791 multiply_cannon_loop 2286 14.5 0.645 0.672 9.580 9.807 ot_mini 108 10.5 0.001 0.001 6.431 6.459 multiply_cannon_multrec 9144 15.5 1.731 1.914 5.970 6.156 rebuild_ks_matrix 119 8.3 0.000 0.000 5.953 5.975 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.952 5.975 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.294 5.314 qs_ot_get_derivative 108 11.5 0.001 0.001 5.156 5.180 dbcsr_mm_accdrv_process 12550 15.8 3.380 4.141 4.128 4.195 sum_up_and_integrate 119 10.3 0.037 0.041 3.628 3.635 integrate_v_rspace 119 11.3 0.003 0.003 3.590 3.598 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.477 3.485 calculate_rho_elec 119 8.7 0.060 0.061 3.477 3.484 init_scf_run 11 5.9 0.000 0.003 3.350 3.350 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.350 3.350 qs_ot_get_p 119 10.4 0.001 0.001 2.737 2.771 init_scf_loop 11 6.9 0.000 0.000 2.418 2.419 calculate_first_density_matrix 1 7.0 0.000 0.000 2.211 2.212 pw_transfer 1439 11.6 0.066 0.069 2.011 2.021 grid_integrate_task_list 119 12.3 1.859 1.962 1.859 1.962 make_m2s 4572 13.5 0.034 0.037 1.790 1.944 density_rs2pw 119 9.7 0.004 0.004 1.794 1.938 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.917 1.929 make_images 4572 14.5 0.268 0.299 1.701 1.852 mp_waitall_1 115863 16.7 1.362 1.813 1.362 1.813 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.786 1.789 rs_pw_transfer 974 11.9 0.008 0.009 1.621 1.749 calculate_dm_sparse 119 9.5 0.000 0.000 1.725 1.743 jit_kernel_multiply 10 15.5 0.709 1.737 0.709 1.737 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.691 1.709 prepare_preconditioner 11 7.9 0.000 0.000 1.679 1.683 make_preconditioner 11 8.9 0.000 0.000 1.679 1.682 fft_wrap_pw1pw2_140 487 13.2 0.368 0.377 1.587 1.601 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.571 1.597 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.582 1.583 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.547 1.548 acc_transpose_blocks 9144 15.5 0.041 0.044 1.450 1.489 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.425 1.433 grid_collocate_task_list 119 9.7 1.296 1.411 1.296 1.411 potential_pw2rs 119 12.3 0.010 0.011 1.348 1.351 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.318 1.332 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.294 1.302 cp_fm_diag_elpa_base 50 14.0 1.267 1.284 1.292 1.301 ot_diis_step 108 11.5 0.012 0.013 1.263 1.263 fft3d_ps 1201 14.6 0.558 0.570 1.245 1.255 qs_energies_init_hamiltonians 11 5.9 0.003 0.005 1.247 1.254 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.211 1.243 apply_single 119 13.6 0.000 0.000 1.211 1.243 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.122 1.126 wfi_extrapolate 11 7.9 0.001 0.001 1.086 1.086 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.733 0.961 make_images_data 4572 15.5 0.042 0.046 0.760 0.935 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.871 0.921 cp_fm_cholesky_invert 11 10.9 0.833 0.835 0.833 0.835 mp_alltoall_d11v 2130 13.8 0.715 0.832 0.715 0.832 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.807 0.813 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.416 0.790 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.781 0.783 acc_transpose_blocks_sync 27432 16.5 0.722 0.757 0.722 0.757 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.690 0.747 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 0.671 0.675 mp_allgather_i34 2286 14.5 0.221 0.631 0.221 0.631 mp_alltoall_z22v 1201 16.6 0.562 0.588 0.562 0.588 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.341000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=591.727273, yerr=5.327304 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 775.512064E+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 0.014 0.033 42.388 42.389 qs_mol_dyn_low 1 2.0 0.003 0.004 42.191 42.199 qs_forces 11 3.9 0.002 0.002 42.124 42.130 qs_energies 11 4.9 0.034 0.035 40.108 40.118 scf_env_do_scf 11 5.9 0.001 0.001 34.326 34.326 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.665 26.667 velocity_verlet 10 3.0 0.002 0.002 23.829 23.834 dbcsr_multiply_generic 2286 12.5 0.100 0.103 18.671 18.911 qs_scf_new_mos 108 7.5 0.001 0.001 16.945 17.041 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.944 17.040 ot_scf_mini 108 9.5 0.002 0.002 15.795 15.895 multiply_cannon 2286 13.5 0.297 0.302 14.884 15.812 multiply_cannon_loop 2286 14.5 0.853 0.861 13.630 14.593 ot_mini 108 10.5 0.001 0.001 9.618 9.732 multiply_cannon_multrec 9144 15.5 3.393 4.615 8.743 8.824 qs_ot_get_derivative 108 11.5 0.001 0.001 7.587 7.688 init_scf_loop 11 6.9 0.000 0.000 7.635 7.639 rebuild_ks_matrix 119 8.3 0.000 0.000 7.295 7.434 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.294 7.433 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.563 6.690 prepare_preconditioner 11 7.9 0.000 0.000 6.657 6.670 make_preconditioner 11 8.9 0.000 0.000 6.657 6.670 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.267 6.550 dbcsr_mm_accdrv_process 12550 15.8 4.575 6.469 5.212 6.501 cp_fm_upper_to_full 72 14.2 3.127 4.460 3.127 4.460 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.240 4.245 calculate_rho_elec 119 8.7 0.118 0.121 4.239 4.245 sum_up_and_integrate 119 10.3 0.064 0.066 4.001 4.007 integrate_v_rspace 119 11.3 0.003 0.004 3.937 3.943 init_scf_run 11 5.9 0.001 0.003 3.658 3.658 scf_env_initial_rho_setup 11 6.9 0.003 0.003 3.658 3.658 qs_ot_get_p 119 10.4 0.001 0.001 3.291 3.426 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.770 3.168 mp_waitall_1 94719 16.7 2.180 3.148 2.180 3.148 dbcsr_complete_redistribute 329 12.2 0.295 0.303 1.917 2.713 pw_transfer 1439 11.6 0.069 0.069 2.618 2.624 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.519 2.525 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.172 2.478 apply_single 119 13.6 0.000 0.000 2.172 2.478 make_m2s 4572 13.5 0.038 0.038 2.286 2.429 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.616 2.394 make_images 4572 14.5 0.355 0.388 2.165 2.308 calculate_dm_sparse 119 9.5 0.000 0.000 2.225 2.243 calculate_first_density_matrix 1 7.0 0.000 0.000 2.204 2.206 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.281 2.173 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.384 2.156 density_rs2pw 119 9.7 0.004 0.004 2.132 2.153 fft_wrap_pw1pw2_140 487 13.2 0.620 0.626 2.137 2.142 mp_alltoall_i22 627 13.8 1.359 2.129 1.359 2.129 grid_integrate_task_list 119 12.3 2.067 2.095 2.067 2.095 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.028 2.078 ot_diis_step 108 11.5 0.014 0.014 2.005 2.005 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.953 1.955 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.824 1.824 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.801 1.808 acc_transpose_blocks 9144 15.5 0.043 0.044 1.782 1.802 mp_sum_l 7287 12.8 0.953 1.686 0.953 1.686 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.575 1.613 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.599 1.600 fft3d_ps 1201 14.6 0.593 0.602 1.532 1.539 grid_collocate_task_list 119 9.7 1.513 1.524 1.513 1.524 rs_pw_transfer 974 11.9 0.009 0.009 1.467 1.483 wfi_extrapolate 11 7.9 0.001 0.001 1.388 1.388 cp_fm_cholesky_invert 11 10.9 1.382 1.385 1.382 1.385 potential_pw2rs 119 12.3 0.014 0.014 1.374 1.376 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.316 1.317 cp_fm_diag_elpa_base 50 14.0 1.173 1.224 1.314 1.315 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.290 1.309 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.235 1.243 hybrid_alltoall_any 4725 16.4 0.090 0.149 1.024 1.222 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.126 1.144 make_images_data 4572 15.5 0.046 0.049 0.937 1.107 mp_alltoall_d11v 2130 13.8 1.079 1.099 1.079 1.099 acc_transpose_blocks_sync 27432 16.5 1.081 1.096 1.081 1.096 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.991 1.033 jit_kernel_multiply 6 15.6 0.608 1.021 0.608 1.021 qs_create_task_list 11 7.9 0.016 0.024 0.952 0.965 generate_qs_task_list 11 8.9 0.368 0.387 0.935 0.955 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.941 0.954 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.389000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=730.545455, yerr=11.515710 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.604352E+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 596172. 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 0.030 80.668 80.669 qs_mol_dyn_low 1 2.0 0.003 0.003 80.417 80.428 qs_forces 11 3.9 0.003 0.003 79.854 79.864 qs_energies 11 4.9 0.001 0.002 76.969 76.994 scf_env_do_scf 11 5.9 0.002 0.014 68.154 68.157 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 62.743 62.743 dbcsr_multiply_generic 2055 12.4 0.104 0.107 49.537 49.865 qs_scf_new_mos 99 7.5 0.000 0.001 45.350 45.446 qs_scf_loop_do_ot 99 8.5 0.001 0.001 45.350 45.446 ot_scf_mini 99 9.5 0.002 0.002 43.084 43.186 multiply_cannon 2055 13.4 0.186 0.191 41.951 42.782 velocity_verlet 10 3.0 0.001 0.002 42.298 42.304 multiply_cannon_loop 2055 14.4 1.779 1.847 40.979 41.784 ot_mini 99 10.5 0.001 0.001 25.399 25.491 qs_ot_get_derivative 99 11.5 0.001 0.001 18.657 18.747 multiply_cannon_multrec 49320 15.4 11.702 12.657 17.558 18.618 rebuild_ks_matrix 110 8.3 0.000 0.000 14.674 14.761 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.674 14.761 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.848 12.925 mp_waitall_1 220248 16.4 9.730 10.763 9.730 10.763 multiply_cannon_sync_h2d 49320 15.4 9.844 10.350 9.844 10.350 qs_ot_get_p 110 10.4 0.001 0.003 9.408 9.529 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.330 7.819 apply_single 110 13.6 0.000 0.001 7.330 7.819 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.972 7.567 sum_up_and_integrate 110 10.3 0.036 0.043 7.146 7.163 integrate_v_rspace 110 11.3 0.003 0.004 7.109 7.134 multiply_cannon_metrocomm3 49320 15.4 0.086 0.090 5.966 6.970 init_scf_run 11 5.9 0.000 0.001 6.762 6.763 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.762 6.763 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.579 6.695 calculate_rho_elec 110 8.6 0.021 0.026 6.579 6.694 ot_diis_step 99 11.5 0.005 0.006 6.604 6.604 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.388 6.411 dbcsr_mm_accdrv_process 87628 16.1 2.914 3.100 5.724 6.036 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.608 5.608 init_scf_loop 11 6.9 0.000 0.003 5.384 5.386 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.195 5.247 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.059 5.082 cp_fm_diag_elpa_base 48 14.0 5.047 5.071 5.057 5.081 mp_sum_l 6594 12.7 3.320 4.246 3.320 4.246 rs_pw_transfer 902 11.9 0.011 0.013 3.644 4.199 wfi_extrapolate 11 7.9 0.001 0.001 4.065 4.066 make_m2s 4110 13.4 0.060 0.065 3.807 3.942 density_rs2pw 110 9.6 0.004 0.005 3.348 3.909 make_images 4110 14.4 0.176 0.190 3.712 3.847 calculate_dm_sparse 110 9.5 0.001 0.001 3.639 3.736 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.556 3.561 grid_integrate_task_list 110 12.3 3.271 3.535 3.271 3.535 multiply_cannon_metrocomm1 49320 15.4 0.064 0.068 1.994 3.379 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.227 3.274 prepare_preconditioner 11 7.9 0.000 0.000 3.208 3.228 make_preconditioner 11 8.9 0.000 0.000 3.208 3.228 pw_transfer 1331 11.6 0.056 0.067 3.031 3.105 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.996 3.040 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 2.942 3.015 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.950 3.005 acc_transpose_blocks 49320 15.4 0.205 0.216 2.512 2.624 fft_wrap_pw1pw2_140 451 13.1 0.457 0.501 2.522 2.614 calculate_first_density_matrix 1 7.0 0.000 0.000 2.591 2.599 jit_kernel_multiply 13 15.9 2.534 2.582 2.534 2.582 potential_pw2rs 110 12.3 0.006 0.007 2.550 2.570 mp_waitany 14300 13.8 1.851 2.550 1.851 2.550 mp_alltoall_d11v 2046 13.8 2.077 2.532 2.077 2.532 grid_collocate_task_list 110 9.6 2.168 2.444 2.168 2.444 fft3d_ps 1111 14.6 0.782 0.866 2.128 2.183 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.955 1.969 make_images_data 4110 15.4 0.045 0.048 1.754 1.888 hybrid_alltoall_any 4261 16.3 0.086 0.490 1.521 1.819 cp_fm_cholesky_invert 11 10.9 1.765 1.768 1.765 1.768 mp_sum_d 3889 11.9 1.226 1.718 1.226 1.718 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.654 1.688 rs_gather_matrices 110 12.3 0.159 0.174 1.190 1.618 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=80.669000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.000000, yerr=3.133398 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_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 591.552512E+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 2581006. 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 0.012 0.031 68.835 68.835 qs_mol_dyn_low 1 2.0 0.003 0.004 68.614 68.624 qs_forces 11 3.9 0.003 0.003 68.470 68.470 qs_energies 11 4.9 0.001 0.001 65.081 65.084 scf_env_do_scf 11 5.9 0.001 0.001 56.621 56.624 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 49.170 49.171 dbcsr_multiply_generic 2055 12.4 0.115 0.120 37.729 37.927 velocity_verlet 10 3.0 0.017 0.041 36.130 36.139 qs_scf_new_mos 99 7.5 0.001 0.001 32.863 33.020 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.863 33.020 multiply_cannon 2055 13.4 0.224 0.247 31.262 32.467 ot_scf_mini 99 9.5 0.003 0.003 31.205 31.366 multiply_cannon_loop 2055 14.4 1.161 1.186 30.016 30.771 ot_mini 99 10.5 0.001 0.001 18.245 18.414 multiply_cannon_multrec 24660 15.4 7.015 8.838 14.005 15.756 rebuild_ks_matrix 110 8.3 0.000 0.000 13.788 13.928 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.788 13.928 qs_ot_get_derivative 99 11.5 0.001 0.001 12.472 12.636 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.114 12.234 mp_waitall_1 176588 16.5 7.414 10.176 7.414 10.176 multiply_cannon_metrocomm3 24660 15.4 0.072 0.076 5.119 7.877 multiply_cannon_sync_h2d 24660 15.4 6.396 7.692 6.396 7.692 init_scf_loop 11 6.9 0.000 0.001 7.416 7.417 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.456 7.157 apply_single 110 13.6 0.000 0.001 6.455 7.157 dbcsr_mm_accdrv_process 52282 16.1 5.434 6.414 6.820 7.143 sum_up_and_integrate 110 10.3 0.053 0.060 6.616 6.629 integrate_v_rspace 110 11.3 0.003 0.003 6.563 6.578 qs_ot_get_p 110 10.4 0.001 0.001 6.321 6.513 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.167 6.176 calculate_rho_elec 110 8.6 0.040 0.048 6.167 6.175 init_scf_run 11 5.9 0.000 0.001 6.088 6.089 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.088 6.088 ot_diis_step 99 11.5 0.010 0.010 5.726 5.726 prepare_preconditioner 11 7.9 0.000 0.000 5.389 5.415 make_preconditioner 11 8.9 0.000 0.000 5.389 5.415 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.681 5.385 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.970 5.133 make_m2s 4110 13.4 0.057 0.060 4.125 4.583 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.466 4.484 make_images 4110 14.4 0.407 0.464 4.018 4.473 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.006 4.006 pw_transfer 1331 11.6 0.067 0.072 3.609 3.741 density_rs2pw 110 9.6 0.004 0.005 3.242 3.704 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.502 3.637 wfi_extrapolate 11 7.9 0.001 0.001 3.517 3.517 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.452 3.463 cp_fm_diag_elpa_base 48 14.0 3.405 3.425 3.449 3.460 rs_pw_transfer 902 11.9 0.012 0.014 2.902 3.412 grid_integrate_task_list 110 12.3 3.166 3.378 3.166 3.378 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.256 3.345 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.245 3.247 fft_wrap_pw1pw2_140 451 13.1 0.528 0.543 3.011 3.151 calculate_dm_sparse 110 9.5 0.001 0.001 2.961 2.998 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.904 2.961 hybrid_alltoall_any 4261 16.3 0.107 0.462 1.970 2.695 make_images_data 4110 15.4 0.049 0.054 2.235 2.688 fft3d_ps 1111 14.6 1.105 1.325 2.463 2.607 grid_collocate_task_list 110 9.6 2.101 2.537 2.101 2.537 cp_fm_cholesky_invert 11 10.9 2.491 2.497 2.491 2.497 calculate_first_density_matrix 1 7.0 0.000 0.000 2.472 2.474 potential_pw2rs 110 12.3 0.008 0.009 2.381 2.388 mp_sum_l 6594 12.7 1.710 2.354 1.710 2.354 jit_kernel_multiply 11 16.2 1.028 2.109 1.028 2.109 mp_alltoall_d11v 2046 13.8 1.742 2.087 1.742 2.087 acc_transpose_blocks 24660 15.4 0.115 0.119 1.981 2.010 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.949 1.970 qs_energies_init_hamiltonians 11 5.9 0.001 0.006 1.866 1.867 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.830 1.845 mp_waitany 10164 13.8 1.224 1.748 1.224 1.748 multiply_cannon_metrocomm4 22605 15.4 0.080 0.084 0.764 1.673 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.538 1.548 mp_irecv_dv 57340 16.2 0.632 1.548 0.632 1.548 mp_allgather_i34 2055 14.4 0.554 1.529 0.554 1.529 build_core_hamiltonian_matrix_ 11 4.9 0.004 0.026 1.390 1.495 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.217 0.974 1.477 dbcsr_complete_redistribute 325 12.2 0.244 0.305 1.121 1.383 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.835000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=558.818182, yerr=6.846607 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_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 665.133056E+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 0.021 0.034 59.939 59.940 qs_mol_dyn_low 1 2.0 0.003 0.004 59.573 59.583 qs_forces 11 3.9 0.004 0.011 59.507 59.508 qs_energies 11 4.9 0.003 0.011 56.259 56.263 scf_env_do_scf 11 5.9 0.001 0.001 48.344 48.344 scf_env_do_scf_inner_loop 99 6.5 0.003 0.011 39.871 39.872 velocity_verlet 10 3.0 0.001 0.002 32.503 32.505 dbcsr_multiply_generic 2055 12.4 0.106 0.111 28.426 28.715 qs_scf_new_mos 99 7.5 0.001 0.001 24.826 24.927 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.826 24.926 ot_scf_mini 99 9.5 0.002 0.003 23.588 23.711 multiply_cannon 2055 13.4 0.211 0.225 22.273 23.248 multiply_cannon_loop 2055 14.4 0.810 0.841 21.250 22.262 ot_mini 99 10.5 0.001 0.001 13.561 13.688 rebuild_ks_matrix 110 8.3 0.000 0.000 12.375 12.551 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.375 12.551 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.867 11.024 multiply_cannon_multrec 16440 15.4 3.694 4.818 9.846 10.911 mp_waitall_1 139946 16.5 6.749 9.836 6.749 9.836 qs_ot_get_derivative 99 11.5 0.001 0.001 9.153 9.277 init_scf_loop 11 6.9 0.001 0.005 8.440 8.443 multiply_cannon_metrocomm3 16440 15.4 0.045 0.049 4.289 7.290 prepare_preconditioner 11 7.9 0.000 0.001 6.629 6.649 make_preconditioner 11 8.9 0.000 0.002 6.629 6.649 sum_up_and_integrate 110 10.3 0.061 0.062 6.547 6.562 integrate_v_rspace 110 11.3 0.003 0.004 6.486 6.502 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.957 6.295 dbcsr_mm_accdrv_process 34862 16.1 5.315 5.671 5.997 6.146 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.021 6.034 calculate_rho_elec 110 8.6 0.059 0.059 6.021 6.034 init_scf_run 11 5.9 0.000 0.001 5.467 5.467 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.467 5.467 qs_ot_get_p 110 10.4 0.005 0.018 5.280 5.432 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.921 5.378 apply_single 110 13.6 0.000 0.000 4.921 5.378 ot_diis_step 99 11.5 0.010 0.011 4.379 4.379 density_rs2pw 110 9.6 0.005 0.006 3.073 4.342 make_m2s 4110 13.4 0.050 0.051 4.086 4.331 make_images 4110 14.4 0.398 0.518 3.972 4.217 rs_pw_transfer 902 11.9 0.010 0.011 2.664 3.928 multiply_cannon_sync_h2d 16440 15.4 3.234 3.815 3.234 3.815 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.121 3.759 qs_ot_p2m_diag 48 11.0 0.042 0.045 3.663 3.668 pw_transfer 1331 11.6 0.066 0.073 3.646 3.653 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.538 3.548 grid_integrate_task_list 110 12.3 3.196 3.427 3.196 3.427 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.302 3.303 fft_wrap_pw1pw2_140 451 13.1 0.646 0.660 3.075 3.085 wfi_extrapolate 11 7.9 0.001 0.001 2.939 2.939 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.780 2.790 cp_fm_diag_elpa_base 48 14.0 2.713 2.746 2.778 2.788 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.769 2.771 make_images_data 4110 15.4 0.045 0.051 2.300 2.623 hybrid_alltoall_any 4261 16.3 0.110 0.383 2.079 2.579 cp_fm_cholesky_invert 11 10.9 2.553 2.558 2.553 2.558 calculate_dm_sparse 110 9.5 0.001 0.001 2.505 2.535 grid_collocate_task_list 110 9.6 2.124 2.526 2.124 2.526 mp_waitany 17072 13.8 1.226 2.523 1.226 2.523 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.450 2.500 multiply_cannon_metrocomm4 14385 15.4 0.048 0.051 0.890 2.470 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.394 2.466 calculate_first_density_matrix 1 7.0 0.000 0.002 2.425 2.427 fft3d_ps 1111 14.6 1.092 1.102 2.349 2.362 mp_irecv_dv 48980 15.7 0.815 2.338 0.815 2.338 potential_pw2rs 110 12.3 0.011 0.011 2.262 2.270 rs_pw_transfer_RS2PW_140 121 11.5 0.176 0.182 0.935 2.193 mp_alltoall_d11v 2046 13.8 1.791 2.043 1.791 2.043 mp_sum_l 6594 12.7 1.450 2.024 1.450 2.024 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.007 2.008 dbcsr_complete_redistribute 325 12.2 0.331 0.374 1.440 1.885 cp_fm_upper_to_full 70 14.2 1.370 1.734 1.370 1.734 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.703 1.723 acc_transpose_blocks 16440 15.4 0.079 0.082 1.596 1.647 cp_fm_cholesky_decompose 22 10.9 1.524 1.546 1.524 1.546 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.378 1.485 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.451 1.465 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.968 1.411 mp_allgather_i34 2055 14.4 0.338 1.359 0.338 1.359 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.265 1.272 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.940000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.000000, yerr=8.644494 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_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 739.500032E+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 0.098 0.113 65.989 65.990 qs_mol_dyn_low 1 2.0 0.004 0.022 65.623 65.636 qs_forces 11 3.9 0.014 0.015 65.554 65.556 qs_energies 11 4.9 0.001 0.001 62.097 62.100 scf_env_do_scf 11 5.9 0.000 0.001 53.848 53.851 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.365 42.365 velocity_verlet 10 3.0 0.001 0.002 37.212 37.214 dbcsr_multiply_generic 2055 12.4 0.113 0.119 30.643 30.859 qs_scf_new_mos 99 7.5 0.001 0.001 27.343 27.442 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.342 27.441 ot_scf_mini 99 9.5 0.002 0.003 25.663 25.752 multiply_cannon 2055 13.4 0.238 0.255 23.637 24.952 multiply_cannon_loop 2055 14.4 1.401 1.491 22.236 22.860 ot_mini 99 10.5 0.001 0.001 14.879 15.002 multiply_cannon_multrec 24660 15.4 4.158 6.803 13.040 14.254 rebuild_ks_matrix 110 8.3 0.000 0.000 12.208 12.320 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.208 12.320 init_scf_loop 11 6.9 0.000 0.000 11.444 11.444 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.760 10.859 qs_ot_get_derivative 99 11.5 0.001 0.001 10.706 10.802 prepare_preconditioner 11 7.9 0.000 0.000 9.704 9.719 make_preconditioner 11 8.9 0.000 0.000 9.704 9.719 dbcsr_mm_accdrv_process 52304 16.0 7.704 9.185 8.725 9.673 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.921 9.399 sum_up_and_integrate 110 10.3 0.067 0.071 6.483 6.498 integrate_v_rspace 110 11.3 0.003 0.003 6.416 6.431 mp_waitall_1 121746 16.5 4.224 6.125 4.224 6.125 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.094 6.103 calculate_rho_elec 110 8.6 0.078 0.082 6.094 6.103 qs_ot_get_p 110 10.4 0.001 0.001 5.680 5.831 make_m2s 4110 13.4 0.059 0.063 5.383 5.743 make_images 4110 14.4 0.580 0.696 5.243 5.600 init_scf_run 11 5.9 0.000 0.002 5.521 5.521 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.521 5.521 cp_fm_upper_to_full 70 14.2 3.334 4.856 3.334 4.856 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.066 4.175 apply_single 110 13.6 0.000 0.000 4.065 4.175 ot_diis_step 99 11.5 0.011 0.011 4.133 4.133 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.868 3.888 dbcsr_complete_redistribute 325 12.2 0.443 0.487 2.647 3.788 pw_transfer 1331 11.6 0.066 0.075 3.706 3.734 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.599 3.632 density_rs2pw 110 9.6 0.004 0.005 2.987 3.505 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.455 3.502 grid_integrate_task_list 110 12.3 3.282 3.450 3.282 3.450 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.317 3.366 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.308 3.309 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.135 3.266 fft_wrap_pw1pw2_140 451 13.1 0.674 0.697 3.105 3.141 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.428 3.085 calculate_dm_sparse 110 9.5 0.001 0.001 2.973 3.008 hybrid_alltoall_any 4261 16.3 0.124 0.464 2.278 3.007 wfi_extrapolate 11 7.9 0.001 0.001 2.975 2.975 make_images_data 4110 15.4 0.048 0.052 2.635 2.965 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.774 2.889 mp_alltoall_i22 605 13.7 1.661 2.855 1.661 2.855 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.848 2.850 rs_pw_transfer 902 11.9 0.010 0.011 2.323 2.848 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.788 2.796 cp_fm_diag_elpa_base 48 14.0 2.640 2.696 2.785 2.793 acc_transpose_blocks 24660 15.4 0.109 0.112 2.405 2.580 cp_fm_cholesky_invert 11 10.9 2.557 2.564 2.557 2.564 multiply_cannon_sync_h2d 24660 15.4 2.381 2.556 2.381 2.556 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.488 2.521 grid_collocate_task_list 110 9.6 2.222 2.491 2.222 2.491 calculate_first_density_matrix 1 7.0 0.000 0.000 2.448 2.450 fft3d_ps 1111 14.6 1.083 1.112 2.352 2.369 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.240 2.242 potential_pw2rs 110 12.3 0.012 0.013 2.123 2.131 mp_alltoall_d11v 2046 13.8 1.706 1.888 1.706 1.888 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.787 1.823 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.613 1.717 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.690 1.701 cp_fm_cholesky_decompose 22 10.9 1.605 1.646 1.605 1.646 mp_sum_l 6594 12.7 0.937 1.617 0.937 1.617 mp_allgather_i34 2055 14.4 0.489 1.589 0.489 1.589 multiply_cannon_metrocomm4 20550 15.4 0.062 0.066 0.834 1.582 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.536 1.552 mp_waitany 13376 13.8 1.048 1.551 1.048 1.551 acc_transpose_blocks_sync 73980 16.4 1.388 1.549 1.388 1.549 mp_irecv_dv 62702 16.1 0.729 1.498 0.729 1.498 rs_pw_transfer_RS2PW_140 121 11.5 0.169 0.177 0.807 1.339 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.990000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=698.727273, yerr=11.458152 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/13/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 807.299199E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1438408 0.0% 0.0% 100.0% average stack size 0.0 0.0 684.2 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 875.466752E+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 0.012 0.028 55.204 55.205 qs_mol_dyn_low 1 2.0 0.003 0.004 54.945 54.956 qs_forces 11 3.9 0.003 0.003 54.881 54.882 qs_energies 11 4.9 0.002 0.002 51.177 51.181 scf_env_do_scf 11 5.9 0.000 0.001 42.844 42.845 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.256 35.257 velocity_verlet 10 3.0 0.001 0.002 31.113 31.116 dbcsr_multiply_generic 2055 12.4 0.105 0.110 23.101 23.308 qs_scf_new_mos 99 7.5 0.001 0.001 20.298 20.370 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.297 20.369 ot_scf_mini 99 9.5 0.002 0.002 19.063 19.113 multiply_cannon 2055 13.4 0.235 0.249 17.594 18.543 multiply_cannon_loop 2055 14.4 0.602 0.623 16.351 16.645 rebuild_ks_matrix 110 8.3 0.000 0.000 11.872 11.922 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.872 11.922 ot_mini 99 10.5 0.001 0.001 10.643 10.688 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.526 10.572 multiply_cannon_multrec 8220 15.4 3.164 4.741 7.557 8.757 init_scf_loop 11 6.9 0.000 0.000 7.543 7.544 mp_waitall_1 103326 16.6 5.775 7.468 5.775 7.468 qs_ot_get_derivative 99 11.5 0.001 0.001 6.873 6.922 sum_up_and_integrate 110 10.3 0.080 0.081 6.495 6.506 integrate_v_rspace 110 11.3 0.003 0.004 6.415 6.426 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.277 6.291 calculate_rho_elec 110 8.6 0.115 0.116 6.276 6.291 prepare_preconditioner 11 7.9 0.000 0.000 5.869 5.876 make_preconditioner 11 8.9 0.000 0.000 5.869 5.876 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.448 5.523 init_scf_run 11 5.9 0.000 0.001 5.213 5.214 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.213 5.213 dbcsr_mm_accdrv_process 17442 15.9 3.054 4.134 4.253 5.190 qs_ot_get_p 110 10.4 0.001 0.001 4.523 4.589 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 3.085 4.516 make_m2s 4110 13.4 0.039 0.040 4.202 4.393 make_images 4110 14.4 0.650 0.711 4.073 4.260 pw_transfer 1331 11.6 0.067 0.074 3.910 3.931 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.793 3.865 apply_single 110 13.6 0.000 0.000 3.793 3.865 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.802 3.827 ot_diis_step 99 11.5 0.012 0.012 3.750 3.750 grid_integrate_task_list 110 12.3 3.368 3.592 3.368 3.592 density_rs2pw 110 9.6 0.004 0.004 2.991 3.364 fft_wrap_pw1pw2_140 451 13.1 0.839 0.853 3.306 3.343 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.204 3.208 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.871 2.871 cp_fm_cholesky_invert 11 10.9 2.812 2.816 2.812 2.816 wfi_extrapolate 11 7.9 0.001 0.001 2.688 2.688 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.669 2.670 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.638 2.640 hybrid_alltoall_any 4261 16.3 0.202 0.856 2.151 2.572 grid_collocate_task_list 110 9.6 2.315 2.571 2.315 2.571 rs_pw_transfer 902 11.9 0.010 0.010 2.133 2.538 calculate_dm_sparse 110 9.5 0.001 0.001 2.471 2.512 make_images_data 4110 15.4 0.041 0.047 2.185 2.477 multiply_cannon_sync_h2d 8220 15.4 2.349 2.472 2.349 2.472 calculate_first_density_matrix 1 7.0 0.000 0.000 2.424 2.425 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.376 2.383 cp_fm_diag_elpa_base 48 14.0 2.319 2.341 2.374 2.381 fft3d_ps 1111 14.6 1.135 1.176 2.339 2.351 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.291 2.309 potential_pw2rs 110 12.3 0.015 0.015 2.028 2.032 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.975 2.013 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.786 2.000 mp_alltoall_d11v 2046 13.8 1.606 1.915 1.606 1.915 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.867 1.894 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.674 1.679 cp_fm_cholesky_decompose 22 10.9 1.611 1.624 1.611 1.624 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.492 1.613 dbcsr_complete_redistribute 325 12.2 0.565 0.583 1.406 1.500 mp_waitany 9240 13.8 1.056 1.479 1.056 1.479 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.439 1.451 acc_transpose_blocks 8220 15.4 0.039 0.040 1.363 1.396 mp_allgather_i34 2055 14.4 0.374 1.376 0.374 1.376 multiply_cannon_metrocomm4 6165 15.4 0.018 0.020 0.479 1.323 qs_create_task_list 11 7.9 0.001 0.001 1.210 1.316 generate_qs_task_list 11 8.9 0.375 0.443 1.210 1.315 mp_irecv_dv 24056 15.7 0.453 1.283 0.453 1.283 rs_gather_matrices 110 12.3 0.328 0.369 0.930 1.218 rs_pw_transfer_RS2PW_140 121 11.5 0.167 0.175 0.804 1.209 jit_kernel_multiply 7 15.6 0.891 1.207 0.891 1.207 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.129 1.167 multiply_cannon_metrocomm1 8220 15.4 0.022 0.022 0.702 1.131 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.205000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=829.454545, yerr=9.413372 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_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 1.446887E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.036 88.428 88.428 qs_mol_dyn_low 1 2.0 0.003 0.004 88.089 88.099 qs_forces 11 3.9 0.003 0.003 88.022 88.023 qs_energies 11 4.9 0.001 0.001 83.809 83.811 scf_env_do_scf 11 5.9 0.001 0.001 73.720 73.721 velocity_verlet 10 3.0 0.002 0.002 56.296 56.302 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 45.337 45.339 dbcsr_multiply_generic 2055 12.4 0.131 0.137 29.718 29.766 init_scf_loop 11 6.9 0.000 0.000 28.312 28.314 qs_scf_new_mos 99 7.5 0.001 0.001 26.872 26.905 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.871 26.905 prepare_preconditioner 11 7.9 0.000 0.000 26.257 26.264 make_preconditioner 11 8.9 0.000 0.000 26.257 26.264 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.512 25.740 ot_scf_mini 99 9.5 0.002 0.002 25.064 25.097 multiply_cannon 2055 13.4 0.331 0.349 22.403 23.044 multiply_cannon_loop 2055 14.4 0.840 0.863 20.612 21.068 cp_fm_upper_to_full 70 14.2 12.846 18.317 12.846 18.317 rebuild_ks_matrix 110 8.3 0.000 0.000 14.315 14.356 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.315 14.356 ot_mini 99 10.5 0.001 0.001 14.173 14.196 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.955 12.992 dbcsr_complete_redistribute 325 12.2 1.019 1.050 7.288 10.481 multiply_cannon_multrec 8220 15.4 4.061 4.219 9.733 9.837 qs_ot_get_derivative 99 11.5 0.001 0.001 9.487 9.522 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.281 9.469 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.729 8.891 mp_alltoall_i22 605 13.7 5.355 8.525 5.355 8.525 mp_waitall_1 84994 16.7 7.675 8.393 7.675 8.393 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.776 7.809 calculate_rho_elec 110 8.6 0.226 0.227 7.776 7.809 sum_up_and_integrate 110 10.3 0.151 0.152 7.430 7.445 integrate_v_rspace 110 11.3 0.004 0.004 7.279 7.295 make_m2s 4110 13.4 0.044 0.044 5.582 6.078 make_images 4110 14.4 0.882 0.925 5.393 5.889 init_scf_run 11 5.9 0.000 0.001 5.818 5.818 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.817 5.817 dbcsr_mm_accdrv_process 11614 15.7 3.806 4.118 5.525 5.725 qs_ot_get_p 110 10.4 0.001 0.001 5.549 5.575 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.825 5.270 apply_single 110 13.6 0.000 0.000 4.825 5.270 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.815 5.266 cp_fm_cholesky_invert 11 10.9 5.205 5.209 5.205 5.209 pw_transfer 1331 11.6 0.075 0.076 4.989 4.996 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.871 4.878 ot_diis_step 99 11.5 0.015 0.015 4.659 4.659 fft_wrap_pw1pw2_140 451 13.1 1.340 1.345 4.252 4.260 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.008 4.015 grid_integrate_task_list 110 12.3 3.753 3.866 3.753 3.866 density_rs2pw 110 9.6 0.004 0.004 3.715 3.791 hybrid_alltoall_any 4261 16.3 0.264 0.565 2.987 3.739 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.668 3.669 make_images_data 4110 15.4 0.044 0.047 2.987 3.627 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.489 3.489 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.016 3.479 wfi_extrapolate 11 7.9 0.001 0.001 3.347 3.347 calculate_dm_sparse 110 9.5 0.001 0.001 3.169 3.198 multiply_cannon_sync_h2d 8220 15.4 3.094 3.126 3.094 3.126 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.978 2.983 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.903 2.903 cp_fm_diag_elpa_base 48 14.0 2.363 2.555 2.901 2.901 fft3d_ps 1111 14.6 1.299 1.306 2.795 2.802 grid_collocate_task_list 110 9.6 2.731 2.782 2.731 2.782 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.647 2.665 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.487 2.504 potential_pw2rs 110 12.3 0.021 0.021 2.406 2.411 rs_pw_transfer 902 11.9 0.010 0.011 2.294 2.353 calculate_first_density_matrix 1 7.0 0.000 0.000 2.349 2.349 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.169 2.222 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.103 2.188 mp_alltoall_d11v 2046 13.8 1.965 2.128 1.965 2.128 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.960 1.964 cp_fm_cholesky_decompose 22 10.9 1.944 1.964 1.944 1.964 qs_create_task_list 11 7.9 0.001 0.001 1.880 1.931 generate_qs_task_list 11 8.9 0.732 0.784 1.880 1.931 acc_transpose_blocks 8220 15.4 0.041 0.041 1.768 1.813 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.747 1.789 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.428000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1282.272727, yerr=77.211741 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_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 631.353344E+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 4010 57647. MP_Allreduce 11104 795. MP_Sync 87 MP_Alltoall 2226 2805752. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 207.514 207.515 qs_mol_dyn_low 1 2.0 0.003 0.004 207.073 207.087 qs_forces 11 3.9 0.005 0.005 206.989 206.989 qs_energies 11 4.9 0.007 0.014 201.359 201.373 scf_env_do_scf 11 5.9 0.001 0.001 184.737 184.742 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 164.452 164.454 dbcsr_multiply_generic 2507 12.6 0.178 0.181 125.275 125.898 qs_scf_new_mos 117 7.6 0.001 0.001 124.262 124.539 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.261 124.538 velocity_verlet 10 3.0 0.001 0.002 124.062 124.062 ot_scf_mini 117 9.6 0.003 0.003 117.557 117.812 multiply_cannon 2507 13.6 0.239 0.251 101.177 102.581 multiply_cannon_loop 2507 14.6 2.423 2.492 99.025 100.288 ot_mini 117 10.6 0.001 0.001 65.934 66.196 multiply_cannon_multrec 60168 15.6 31.860 33.530 41.564 43.537 qs_ot_get_derivative 117 11.6 0.001 0.001 41.177 41.425 rebuild_ks_matrix 128 8.3 0.001 0.001 33.674 33.880 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.673 33.879 mp_waitall_1 267128 16.5 28.414 31.713 28.414 31.713 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.198 30.372 qs_ot_get_p 128 10.4 0.001 0.001 29.398 29.682 multiply_cannon_sync_h2d 60168 15.6 26.342 27.690 26.342 27.690 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.275 25.061 apply_single 128 13.6 0.001 0.001 24.275 25.060 ot_diis_step 117 11.6 0.008 0.008 24.522 24.524 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.870 22.939 init_scf_loop 11 6.9 0.000 0.001 20.212 20.213 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.971 19.972 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.772 19.949 multiply_cannon_metrocomm3 60168 15.6 0.119 0.123 15.992 17.509 cp_fm_diag_elpa 83 13.4 0.000 0.001 17.025 17.065 cp_fm_diag_elpa_base 83 14.4 16.958 17.004 17.021 17.063 prepare_preconditioner 11 7.9 0.000 0.000 15.645 15.696 make_preconditioner 11 8.9 0.000 0.000 15.645 15.696 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.886 15.050 sum_up_and_integrate 128 10.3 0.089 0.107 14.225 14.243 integrate_v_rspace 128 11.3 0.004 0.005 14.135 14.157 make_m2s 5014 13.6 0.104 0.113 13.824 14.149 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.878 14.062 calculate_rho_elec 128 8.7 0.046 0.065 13.877 14.061 make_images 5014 14.6 0.401 0.422 13.647 13.980 init_scf_run 11 5.9 0.000 0.001 12.447 12.448 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.447 12.447 density_rs2pw 128 9.7 0.006 0.007 7.109 10.327 mp_sum_l 7950 12.9 9.010 9.994 9.010 9.994 dbcsr_mm_accdrv_process 124484 16.2 4.683 4.849 9.276 9.742 wfi_extrapolate 11 7.9 0.001 0.001 9.133 9.133 rs_pw_transfer 1046 11.9 0.016 0.018 5.756 8.982 cp_fm_cholesky_invert 11 10.9 8.899 8.907 8.899 8.907 calculate_dm_sparse 128 9.5 0.001 0.001 8.517 8.589 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.149 8.250 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.039 8.211 pw_transfer 1547 11.6 0.075 0.093 7.766 7.954 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 7.563 7.749 make_images_data 5014 15.6 0.070 0.075 6.773 7.702 grid_integrate_task_list 128 12.3 7.133 7.571 7.133 7.571 multiply_cannon_metrocomm1 60168 15.6 0.092 0.095 5.952 7.395 hybrid_alltoall_any 5200 16.5 0.297 2.280 5.930 7.261 fft_wrap_pw1pw2_140 523 13.2 1.281 1.328 6.628 6.803 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.684 6.694 grid_collocate_task_list 128 9.7 4.670 5.975 4.670 5.975 mp_waitany 16020 13.9 2.732 5.962 2.732 5.962 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.807 5.914 fft3d_ps 1291 14.7 2.134 2.794 5.316 5.660 mp_alltoall_d11v 2415 14.1 4.189 5.467 4.189 5.467 rs_pw_transfer_RS2PW_140 139 11.5 0.281 0.297 2.163 5.391 potential_pw2rs 128 12.3 0.009 0.011 4.584 4.602 cp_fm_cholesky_decompose 22 10.9 4.583 4.597 4.583 4.597 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.515000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.909091, yerr=6.960336 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_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 846.442496E+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 4910117. 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 0.058 0.140 190.654 190.656 qs_mol_dyn_low 1 2.0 0.003 0.005 190.091 190.105 qs_forces 11 3.9 0.014 0.047 190.007 190.009 qs_energies 11 4.9 0.003 0.007 183.130 183.153 scf_env_do_scf 11 5.9 0.001 0.004 166.691 166.701 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 133.848 133.851 velocity_verlet 10 3.0 0.002 0.005 119.719 119.720 dbcsr_multiply_generic 2507 12.6 0.191 0.198 96.933 98.192 qs_scf_new_mos 117 7.6 0.001 0.001 94.185 94.829 qs_scf_loop_do_ot 117 8.6 0.001 0.002 94.184 94.828 ot_scf_mini 117 9.6 0.004 0.006 89.390 90.099 multiply_cannon 2507 13.6 0.503 0.557 76.749 81.311 multiply_cannon_loop 2507 14.6 1.565 1.626 73.881 76.681 ot_mini 117 10.6 0.001 0.001 50.198 50.854 mp_waitall_1 214728 16.6 23.966 37.911 23.966 37.911 multiply_cannon_multrec 30084 15.6 21.019 26.061 31.535 37.054 rebuild_ks_matrix 128 8.3 0.001 0.001 32.851 33.597 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.021 32.850 33.596 init_scf_loop 11 6.9 0.001 0.004 32.752 32.755 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.498 30.176 multiply_cannon_metrocomm3 30084 15.6 0.096 0.101 15.702 29.450 qs_ot_get_derivative 117 11.6 0.001 0.002 28.157 28.867 prepare_preconditioner 11 7.9 0.000 0.000 28.392 28.461 make_preconditioner 11 8.9 0.001 0.002 28.392 28.461 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.078 27.636 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.114 23.237 apply_single 128 13.6 0.001 0.001 22.114 23.237 qs_ot_get_p 128 10.4 0.001 0.002 21.354 22.157 ot_diis_step 117 11.6 0.014 0.015 21.860 21.863 multiply_cannon_sync_h2d 30084 15.6 18.006 20.153 18.006 20.153 cp_fm_cholesky_invert 11 10.9 16.663 16.675 16.663 16.675 qs_ot_p2m_diag 83 11.4 0.188 0.217 16.473 16.508 cp_dbcsr_syevd 83 12.4 0.006 0.006 15.269 15.270 make_m2s 5014 13.6 0.093 0.099 14.140 15.247 make_images 5014 14.6 1.171 1.376 13.928 15.033 sum_up_and_integrate 128 10.3 0.117 0.134 14.663 14.694 integrate_v_rspace 128 11.3 0.004 0.005 14.547 14.580 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.138 14.165 calculate_rho_elec 128 8.7 0.089 0.107 14.137 14.164 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.100 12.132 cp_fm_diag_elpa_base 83 14.4 11.845 11.939 12.094 12.121 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.184 11.702 init_scf_run 11 5.9 0.000 0.001 11.533 11.534 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.533 11.534 multiply_cannon_metrocomm4 27577 15.6 0.109 0.122 3.833 11.091 mp_irecv_dv 69486 16.3 3.624 10.684 3.624 10.684 dbcsr_mm_accdrv_process 62242 16.2 5.406 6.201 9.969 10.518 density_rs2pw 128 9.7 0.007 0.007 7.433 9.885 make_images_data 5014 15.6 0.068 0.079 8.299 9.679 hybrid_alltoall_any 5200 16.5 0.352 1.507 7.042 9.251 pw_transfer 1547 11.6 0.086 0.097 8.772 8.828 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 8.546 8.605 wfi_extrapolate 11 7.9 0.001 0.001 8.358 8.358 rs_pw_transfer 1046 11.9 0.015 0.017 5.717 8.207 fft_wrap_pw1pw2_140 523 13.2 1.346 1.367 7.563 7.634 grid_integrate_task_list 128 12.3 7.173 7.556 7.173 7.556 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.281 7.069 cp_fm_cholesky_decompose 22 10.9 6.860 6.925 6.860 6.925 calculate_dm_sparse 128 9.5 0.001 0.001 6.450 6.586 mp_sum_l 7950 12.9 4.305 6.534 4.305 6.534 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.091 6.102 grid_collocate_task_list 128 9.7 4.788 5.960 4.788 5.960 fft3d_ps 1291 14.7 2.811 2.983 5.837 5.888 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.369 5.453 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.184 5.370 mp_waitany 11748 13.9 2.541 5.085 2.541 5.085 potential_pw2rs 128 12.3 0.015 0.017 4.892 4.909 mp_alltoall_d11v 2415 14.1 4.119 4.853 4.119 4.853 rs_pw_transfer_RS2PW_140 139 11.5 0.351 0.378 2.167 4.629 mp_allgather_i34 2507 14.6 1.390 4.363 1.390 4.363 mp_sum_d 4479 12.1 2.687 4.121 2.687 4.121 dbcsr_complete_redistribute 395 12.7 0.776 0.863 3.140 3.973 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.743 3.815 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.656000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=804.545455, yerr=4.008256 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 965.959680E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+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 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58192. MP_Allreduce 11085 1000. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.071 175.261 175.263 qs_mol_dyn_low 1 2.0 0.003 0.004 174.687 174.702 qs_forces 11 3.9 0.012 0.023 174.594 174.600 qs_energies 11 4.9 0.002 0.008 167.902 167.923 scf_env_do_scf 11 5.9 0.001 0.004 152.231 152.232 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 117.886 117.887 velocity_verlet 10 3.0 0.001 0.002 111.370 111.373 dbcsr_multiply_generic 2485 12.5 0.179 0.183 81.946 83.212 qs_scf_new_mos 116 7.6 0.001 0.001 79.705 80.003 qs_scf_loop_do_ot 116 8.6 0.001 0.002 79.705 80.002 ot_scf_mini 116 9.6 0.003 0.004 75.540 75.904 multiply_cannon 2485 13.5 0.503 0.522 62.460 67.835 multiply_cannon_loop 2485 14.5 1.115 1.177 59.548 62.566 ot_mini 116 10.6 0.001 0.001 42.823 43.188 mp_waitall_1 169034 16.6 24.828 34.552 24.828 34.552 init_scf_loop 11 6.9 0.001 0.004 34.247 34.248 rebuild_ks_matrix 127 8.3 0.001 0.001 30.718 31.182 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 30.717 31.181 prepare_preconditioner 11 7.9 0.000 0.001 30.203 30.243 make_preconditioner 11 8.9 0.000 0.002 30.203 30.242 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.893 29.256 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.594 28.018 multiply_cannon_multrec 19880 15.5 13.092 15.944 22.499 25.609 multiply_cannon_metrocomm3 19880 15.5 0.061 0.063 15.619 25.067 qs_ot_get_derivative 116 11.6 0.001 0.002 23.436 23.794 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.420 20.551 apply_single 127 13.6 0.001 0.001 19.420 20.550 ot_diis_step 116 11.6 0.018 0.024 19.278 19.278 qs_ot_get_p 127 10.4 0.001 0.001 18.177 18.646 make_m2s 4970 13.5 0.082 0.088 14.443 15.629 make_images 4970 14.5 1.135 1.236 14.210 15.391 multiply_cannon_sync_h2d 19880 15.5 13.587 15.242 13.587 15.242 sum_up_and_integrate 127 10.3 0.130 0.145 14.501 14.529 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.401 14.436 calculate_rho_elec 127 8.7 0.131 0.147 14.400 14.435 integrate_v_rspace 127 11.3 0.004 0.004 14.371 14.400 cp_fm_cholesky_invert 11 10.9 14.022 14.030 14.022 14.030 qs_ot_p2m_diag 82 11.4 0.262 0.269 13.919 13.926 cp_dbcsr_syevd 82 12.4 0.005 0.006 12.888 12.890 init_scf_run 11 5.9 0.000 0.002 10.564 10.565 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.564 10.564 make_images_data 4970 15.5 0.064 0.075 8.623 10.213 cp_fm_diag_elpa 82 13.4 0.000 0.001 9.793 9.811 cp_fm_diag_elpa_base 82 14.4 9.390 9.540 9.790 9.808 hybrid_alltoall_any 5155 16.4 0.446 2.023 7.486 9.787 dbcsr_mm_accdrv_process 41158 16.2 5.755 6.385 8.875 9.379 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.970 9.231 multiply_cannon_metrocomm4 17395 15.5 0.065 0.074 3.426 9.116 density_rs2pw 127 9.7 0.006 0.007 7.089 9.066 pw_transfer 1535 11.6 0.085 0.100 8.804 8.910 mp_irecv_dv 49801 16.2 3.300 8.862 3.300 8.862 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.581 8.691 grid_integrate_task_list 127 12.3 7.240 7.780 7.240 7.780 fft_wrap_pw1pw2_140 519 13.2 1.415 1.450 7.635 7.759 wfi_extrapolate 11 7.9 0.001 0.001 7.484 7.484 cp_fm_cholesky_decompose 22 10.9 7.175 7.207 7.175 7.207 cp_fm_upper_to_full 104 14.8 5.647 7.131 5.647 7.131 rs_pw_transfer 1038 11.9 0.014 0.014 5.081 7.071 dbcsr_complete_redistribute 393 12.7 1.176 1.209 4.505 6.244 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.223 5.901 calculate_dm_sparse 127 9.5 0.001 0.001 5.762 5.863 fft3d_ps 1281 14.7 2.720 2.959 5.706 5.788 grid_collocate_task_list 127 9.7 4.951 5.773 4.951 5.773 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.513 5.519 mp_alltoall_d11v 2401 14.1 4.415 5.333 4.415 5.333 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.365 5.105 mp_sum_l 7884 12.9 3.237 4.965 3.237 4.965 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.629 4.777 potential_pw2rs 127 12.3 0.020 0.022 4.613 4.627 mp_allgather_i34 2485 14.5 1.419 4.517 1.419 4.517 mp_waitany 11660 13.9 2.369 4.387 2.369 4.387 transfer_fm_to_dbcsr 11 9.9 0.017 0.022 2.291 4.011 rs_pw_transfer_RS2PW_140 138 11.5 0.328 0.348 1.958 3.933 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.871 3.904 mp_alltoall_i22 712 14.1 1.924 3.826 1.924 3.826 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.815 3.826 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.617 3.657 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.263000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=903.000000, yerr=13.087121 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.164947E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58132. MP_Allreduce 11095 1083. MP_Sync 86 MP_Alltoall 1700 12496371. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.033 187.587 187.587 qs_mol_dyn_low 1 2.0 0.003 0.003 187.209 187.222 qs_forces 11 3.9 0.004 0.004 186.865 186.870 qs_energies 11 4.9 0.001 0.002 179.624 179.633 scf_env_do_scf 11 5.9 0.001 0.001 162.921 162.927 velocity_verlet 10 3.0 0.002 0.002 124.188 124.201 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 116.772 116.773 dbcsr_multiply_generic 2485 12.5 0.185 0.192 79.847 80.445 qs_scf_new_mos 116 7.6 0.001 0.001 79.589 79.902 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.588 79.902 ot_scf_mini 116 9.6 0.003 0.003 75.039 75.353 multiply_cannon 2485 13.5 0.546 0.576 55.240 58.370 multiply_cannon_loop 2485 14.5 1.810 1.916 51.567 53.117 init_scf_loop 11 6.9 0.000 0.000 46.029 46.030 ot_mini 116 10.6 0.001 0.001 42.541 42.837 prepare_preconditioner 11 7.9 0.000 0.000 41.929 41.953 make_preconditioner 11 8.9 0.000 0.000 41.929 41.953 make_full_inverse_cholesky 11 9.9 0.011 0.023 35.635 40.611 multiply_cannon_multrec 29820 15.5 13.729 18.908 26.478 31.239 rebuild_ks_matrix 127 8.3 0.001 0.001 29.733 30.020 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 29.732 30.020 mp_waitall_1 146592 16.7 16.893 27.177 16.893 27.177 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.721 26.991 qs_ot_get_derivative 116 11.6 0.002 0.002 23.011 23.317 make_m2s 4970 13.5 0.096 0.098 20.278 21.461 make_images 4970 14.5 1.932 2.260 19.973 21.155 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.934 19.508 apply_single 127 13.6 0.001 0.001 18.934 19.508 ot_diis_step 116 11.6 0.017 0.018 19.391 19.393 qs_ot_get_p 127 10.4 0.001 0.001 18.748 19.038 cp_fm_upper_to_full 104 14.8 11.379 16.739 11.379 16.739 cp_fm_cholesky_invert 11 10.9 16.027 16.036 16.027 16.036 multiply_cannon_metrocomm3 29820 15.5 0.050 0.053 6.071 15.063 qs_ot_p2m_diag 82 11.4 0.338 0.385 14.663 14.718 sum_up_and_integrate 127 10.3 0.138 0.150 14.520 14.545 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.475 14.530 calculate_rho_elec 127 8.7 0.175 0.191 14.474 14.529 integrate_v_rspace 127 11.3 0.004 0.004 14.382 14.411 cp_dbcsr_syevd 82 12.4 0.005 0.006 13.283 13.287 dbcsr_mm_accdrv_process 61748 16.2 8.394 9.132 12.326 12.849 dbcsr_complete_redistribute 393 12.7 1.511 1.622 8.971 12.733 make_images_data 4970 15.5 0.065 0.071 10.946 12.666 hybrid_alltoall_any 5155 16.4 0.528 2.208 9.593 11.635 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.614 11.371 multiply_cannon_sync_h2d 29820 15.5 10.387 11.101 10.387 11.101 init_scf_run 11 5.9 0.000 0.001 10.816 10.817 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.816 10.817 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.129 10.147 cp_fm_diag_elpa_base 82 14.4 9.180 9.497 10.122 10.139 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.272 9.963 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.586 9.814 mp_alltoall_i22 712 14.1 5.536 9.345 5.536 9.345 pw_transfer 1535 11.6 0.086 0.100 9.037 9.126 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.812 8.908 density_rs2pw 127 9.7 0.006 0.007 6.896 8.155 fft_wrap_pw1pw2_140 519 13.2 1.557 1.587 7.861 7.963 grid_integrate_task_list 127 12.3 7.412 7.897 7.412 7.897 wfi_extrapolate 11 7.9 0.001 0.001 7.720 7.720 cp_fm_cholesky_decompose 22 10.9 7.508 7.589 7.508 7.589 multiply_cannon_metrocomm4 24850 15.5 0.083 0.094 2.727 6.664 mp_irecv_dv 75445 16.2 2.575 6.395 2.575 6.395 calculate_dm_sparse 127 9.5 0.001 0.001 6.236 6.336 rs_pw_transfer 1038 11.9 0.013 0.014 4.556 5.886 grid_collocate_task_list 127 9.7 5.120 5.869 5.120 5.869 fft3d_ps 1281 14.7 2.822 2.895 5.745 5.811 mp_alltoall_d11v 2401 14.1 4.875 5.550 4.875 5.550 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.465 5.523 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.453 4.529 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.384 4.471 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.433 4.434 potential_pw2rs 127 12.3 0.023 0.024 4.415 4.426 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.186 4.262 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.587000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1092.000000, yerr=15.713341 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/19/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 5.909858E+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 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.507889E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+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 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591885696 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57807. MP_Allreduce 11152 1164. MP_Sync 87 MP_Alltoall 1724 18848449. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.034 170.580 170.580 qs_mol_dyn_low 1 2.0 0.003 0.004 170.169 170.191 qs_forces 11 3.9 0.004 0.005 170.064 170.067 qs_energies 11 4.9 0.002 0.002 162.469 162.478 scf_env_do_scf 11 5.9 0.001 0.001 145.238 145.242 velocity_verlet 10 3.0 0.002 0.002 112.583 112.587 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 109.939 109.941 dbcsr_multiply_generic 2527 12.6 0.181 0.194 72.005 72.355 qs_scf_new_mos 118 7.6 0.001 0.001 71.086 71.201 qs_scf_loop_do_ot 118 8.6 0.001 0.001 71.086 71.200 ot_scf_mini 118 9.6 0.003 0.004 66.638 66.741 multiply_cannon 2527 13.6 0.563 0.602 53.406 56.888 multiply_cannon_loop 2527 14.6 0.820 0.859 50.486 51.140 ot_mini 118 10.6 0.001 0.001 37.438 37.527 init_scf_loop 11 6.9 0.000 0.000 35.153 35.154 prepare_preconditioner 11 7.9 0.000 0.000 31.293 31.317 make_preconditioner 11 8.9 0.000 0.000 31.293 31.317 mp_waitall_1 126780 16.7 24.528 30.735 24.528 30.735 rebuild_ks_matrix 129 8.3 0.001 0.001 30.026 30.081 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 30.026 30.081 make_full_inverse_cholesky 11 9.9 0.016 0.026 29.259 29.538 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.267 27.316 multiply_cannon_multrec 10108 15.6 10.538 14.489 18.103 21.071 qs_ot_get_derivative 118 11.6 0.002 0.002 20.367 20.464 multiply_cannon_metrocomm3 10108 15.6 0.025 0.027 12.463 19.413 cp_fm_cholesky_invert 11 10.9 17.844 17.849 17.844 17.849 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.060 17.194 apply_single 129 13.6 0.001 0.001 17.059 17.194 ot_diis_step 118 11.6 0.020 0.020 17.000 17.001 qs_ot_get_p 129 10.4 0.001 0.001 16.430 16.538 make_m2s 5054 13.6 0.067 0.071 14.796 15.609 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.490 15.520 calculate_rho_elec 129 8.7 0.261 0.271 15.489 15.520 make_images 5054 14.6 2.194 2.638 14.488 15.296 sum_up_and_integrate 129 10.3 0.183 0.193 15.038 15.081 integrate_v_rspace 129 11.3 0.004 0.004 14.855 14.907 qs_ot_p2m_diag 83 11.4 0.495 0.501 12.704 12.720 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.547 11.548 multiply_cannon_sync_h2d 10108 15.6 10.950 11.318 10.950 11.318 init_scf_run 11 5.9 0.000 0.002 10.460 10.460 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.460 10.460 hybrid_alltoall_any 5240 16.5 0.852 3.815 8.356 9.803 make_images_data 5054 15.6 0.056 0.066 8.509 9.758 pw_transfer 1559 11.6 0.087 0.097 9.477 9.496 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.252 9.276 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.724 8.734 cp_fm_diag_elpa_base 83 14.4 8.486 8.551 8.720 8.730 grid_integrate_task_list 129 12.3 7.819 8.236 7.819 8.236 fft_wrap_pw1pw2_140 527 13.2 1.938 1.975 8.204 8.230 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 7.912 7.982 cp_fm_cholesky_decompose 22 10.9 7.810 7.919 7.810 7.919 dbcsr_mm_accdrv_process 20926 16.1 3.359 4.539 7.211 7.904 density_rs2pw 129 9.7 0.006 0.006 6.890 7.799 wfi_extrapolate 11 7.9 0.001 0.001 7.444 7.444 multiply_cannon_metrocomm1 10108 15.6 0.030 0.031 4.376 7.160 grid_collocate_task_list 129 9.7 5.514 6.221 5.514 6.221 calculate_dm_sparse 129 9.5 0.001 0.001 6.098 6.169 fft3d_ps 1301 14.7 2.780 2.851 5.688 5.723 dbcsr_complete_redistribute 395 12.7 2.119 2.176 5.043 5.479 mp_alltoall_d11v 2423 14.1 4.758 5.460 4.758 5.460 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.267 5.268 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.135 5.160 rs_pw_transfer 1054 12.0 0.012 0.013 4.118 5.031 mp_allgather_i34 2527 14.6 1.097 4.570 1.097 4.570 potential_pw2rs 129 12.3 0.026 0.026 4.404 4.415 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.069 4.127 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.571 3.884 multiply_cannon_metrocomm4 7581 15.6 0.026 0.029 1.777 3.848 mp_irecv_dv 29086 15.9 1.738 3.779 1.738 3.779 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.660 3.713 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.667 3.680 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.588 3.622 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.257 3.592 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.382 3.483 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.580000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1419.545455, yerr=24.246325 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456868864 0.0% 0.0% 100.0% flops 32 x 32 x 32 1967095021568 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992004743168 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753954734080 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 11613074706432 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239170400256 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239170400256 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911117062144 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.237285E+12 0.0% 0.0% 100.0% flops max/rank 11.786603E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806448352 0.0% 0.0% 100.0% number of processed stacks 1981024 0.0% 0.0% 100.0% average stack size 0.0 0.0 3435.8 marketing flops 145.655226E+12 ------------------------------------------------------------------------------- # multiplications 2531 max memory usage/rank 3.230982E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101240 MPI messages size (bytes): total size 1.145036E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.310119E+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 45728 35500589056 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592178544 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4080 58046. MP_Allreduce 11277 1498. MP_Sync 87 MP_Alltoall 1724 36993608. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.038 297.267 297.268 qs_mol_dyn_low 1 2.0 0.003 0.004 296.715 296.734 qs_forces 11 3.9 0.004 0.005 296.622 296.625 qs_energies 11 4.9 0.011 0.012 287.190 287.201 scf_env_do_scf 11 5.9 0.001 0.001 264.307 264.311 velocity_verlet 10 3.0 0.002 0.002 214.704 214.712 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 139.938 139.941 init_scf_loop 11 6.9 0.000 0.000 124.089 124.093 prepare_preconditioner 11 7.9 0.000 0.000 119.055 119.073 make_preconditioner 11 8.9 0.000 0.000 119.055 119.073 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.611 116.229 qs_scf_new_mos 118 7.6 0.001 0.001 90.048 90.144 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.047 90.143 ot_scf_mini 118 9.6 0.004 0.004 85.101 85.139 dbcsr_multiply_generic 2531 12.6 0.229 0.241 82.547 82.994 cp_fm_upper_to_full 105 14.8 53.426 75.845 53.426 75.845 multiply_cannon 2531 13.6 0.689 0.721 59.518 60.428 multiply_cannon_loop 2531 14.6 1.076 1.099 55.545 56.880 ot_mini 118 10.6 0.001 0.001 44.499 44.538 dbcsr_complete_redistribute 395 12.7 4.039 4.082 29.424 41.721 copy_fm_to_dbcsr 209 11.7 0.001 0.002 25.893 38.163 rebuild_ks_matrix 129 8.3 0.001 0.001 37.525 37.542 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 37.524 37.541 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.401 35.457 qs_ks_update_qs_env 129 7.6 0.001 0.001 34.477 34.498 mp_alltoall_i22 718 14.1 21.166 33.568 21.166 33.568 cp_fm_cholesky_invert 11 10.9 33.167 33.173 33.167 33.173 mp_waitall_1 104660 16.8 28.117 32.268 28.117 32.268 qs_ot_get_p 129 10.4 0.001 0.001 25.455 25.519 qs_ot_get_derivative 118 11.6 0.002 0.002 24.713 24.751 qs_ot_p2m_diag 83 11.4 0.879 0.886 21.439 21.471 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.522 20.542 calculate_rho_elec 129 8.7 0.486 0.487 20.522 20.541 multiply_cannon_metrocomm3 10124 15.6 0.025 0.028 19.174 20.529 make_m2s 5062 13.6 0.077 0.081 18.458 19.869 ot_diis_step 118 11.6 0.022 0.023 19.756 19.757 cp_dbcsr_syevd 83 12.4 0.006 0.006 19.645 19.646 sum_up_and_integrate 129 10.3 0.322 0.323 19.439 19.523 make_images 5062 14.6 3.087 3.292 17.974 19.387 integrate_v_rspace 129 11.3 0.004 0.004 19.116 19.200 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.042 19.151 apply_single 129 13.6 0.001 0.001 19.042 19.150 multiply_cannon_multrec 10124 15.6 10.440 12.269 18.293 18.373 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.468 16.468 cp_fm_diag_elpa_base 83 14.4 12.026 13.617 16.464 16.465 multiply_cannon_sync_h2d 10124 15.6 14.385 14.417 14.385 14.417 init_scf_run 11 5.9 0.000 0.001 12.515 12.516 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.515 12.516 hybrid_alltoall_any 5248 16.5 1.325 3.091 10.418 12.510 make_images_data 5062 15.6 0.064 0.069 10.233 12.349 pw_transfer 1559 11.6 0.093 0.094 11.518 11.525 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 11.282 11.288 fft_wrap_pw1pw2_140 527 13.2 3.130 3.197 10.049 10.054 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.702 9.737 dbcsr_mm_accdrv_process 20942 16.1 4.299 6.188 7.603 9.477 mp_alltoall_d11v 2423 14.1 8.032 9.324 8.032 9.324 wfi_extrapolate 11 7.9 0.001 0.001 9.237 9.237 cp_fm_cholesky_decompose 22 10.9 8.880 8.897 8.880 8.897 grid_integrate_task_list 129 12.3 8.636 8.832 8.636 8.832 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.183 8.183 density_rs2pw 129 9.7 0.005 0.006 7.970 8.124 calculate_dm_sparse 129 9.5 0.001 0.001 6.842 6.943 grid_collocate_task_list 129 9.7 6.480 6.554 6.480 6.554 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.490 6.524 fft3d_ps 1301 14.7 2.872 2.880 6.282 6.336 rs_scatter_matrices 140 9.7 3.847 4.857 6.051 6.275 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.198 6.245 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=297.268000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2928.090909, yerr=162.697407 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_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 1.259676E+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 6799628. 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 0.018 0.047 85.072 85.073 qs_energies 1 2.0 0.000 0.000 84.574 84.588 ls_scf 1 3.0 0.000 0.000 83.595 83.608 dbcsr_multiply_generic 111 6.7 0.016 0.016 72.585 72.739 multiply_cannon 111 7.7 0.017 0.020 56.025 57.006 multiply_cannon_loop 111 8.7 0.228 0.242 52.602 53.701 ls_scf_main 1 4.0 0.000 0.000 52.338 52.338 density_matrix_trs4 2 5.0 0.002 0.003 46.780 46.864 ls_scf_init_scf 1 4.0 0.000 0.000 28.205 28.206 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.070 27.122 mp_waitall_1 11031 10.9 22.656 26.012 22.656 26.012 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 24.982 25.000 multiply_cannon_multrec 2664 9.7 8.152 8.937 15.578 17.387 multiply_cannon_sync_h2d 2664 9.7 13.408 15.445 13.408 15.445 make_m2s 222 7.7 0.010 0.012 13.001 13.593 make_images 222 8.7 0.101 0.110 12.979 13.574 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.908 12.251 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.463 8.571 make_images_data 222 9.7 0.005 0.006 7.622 8.133 dbcsr_mm_accdrv_process 4760 10.4 0.592 0.705 7.041 8.022 hybrid_alltoall_any 227 10.6 0.219 1.860 6.536 7.756 dbcsr_mm_accdrv_process_sort 4760 11.4 6.250 7.165 6.250 7.165 calculate_norms 4752 9.8 5.508 6.045 5.508 6.045 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.030 5.134 mp_sum_l 887 5.1 3.021 4.519 3.021 4.519 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.055 3.447 mp_irecv_dv 6231 10.9 2.039 3.417 2.039 3.417 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.249 3.382 make_images_sizes 222 9.7 0.000 0.000 0.651 3.342 mp_alltoall_i44 222 10.7 0.650 3.342 0.650 3.342 arnoldi_extremal 4 6.8 0.000 0.000 3.209 3.236 arnoldi_normal_ev 4 7.8 0.001 0.004 3.209 3.236 build_subspace 16 8.4 0.009 0.012 3.119 3.121 ls_scf_post 1 4.0 0.000 0.000 3.052 3.065 ls_scf_store_result 1 5.0 0.000 0.000 2.868 2.908 dbcsr_special_finalize 555 9.7 0.006 0.007 2.364 2.770 dbcsr_merge_single_wm 555 10.7 0.456 0.578 2.354 2.760 make_images_pack 222 9.7 2.214 2.637 2.216 2.639 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.308 2.560 dbcsr_sort_data 658 11.4 2.156 2.492 2.156 2.492 dbcsr_matrix_vector_mult_local 304 10.0 2.068 2.471 2.070 2.473 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.315 2.400 buffer_matrices_ensure_size 222 8.7 1.744 2.072 1.744 2.072 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.779 1.780 rebuild_ks_matrix 3 7.3 0.000 0.000 1.769 1.771 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.769 1.771 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.073000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_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 2.163184E+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 0.066 0.182 90.548 90.550 qs_energies 1 2.0 0.000 0.001 89.887 89.902 ls_scf 1 3.0 0.001 0.006 88.549 88.553 dbcsr_multiply_generic 111 6.7 0.016 0.020 74.633 74.941 multiply_cannon 111 7.7 0.027 0.040 52.817 57.662 ls_scf_main 1 4.0 0.000 0.005 55.062 55.067 multiply_cannon_loop 111 8.7 0.136 0.147 49.954 53.474 density_matrix_trs4 2 5.0 0.002 0.004 49.371 49.613 mp_waitall_1 9105 10.9 20.709 29.798 20.709 29.798 ls_scf_init_scf 1 4.0 0.000 0.002 29.764 29.765 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.559 28.640 multiply_cannon_multrec 1332 9.7 13.305 17.422 22.658 27.844 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.202 26.212 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.526 20.424 make_m2s 222 7.7 0.008 0.009 15.069 15.694 make_images 222 8.7 1.377 1.711 15.039 15.665 dbcsr_mm_accdrv_process 4041 10.4 0.334 0.543 8.946 10.546 dbcsr_mm_accdrv_process_sort 4041 11.4 8.464 10.003 8.464 10.003 make_images_data 222 9.7 0.005 0.006 8.751 9.727 hybrid_alltoall_any 227 10.6 0.545 2.583 8.082 9.537 mp_sum_l 887 5.1 5.322 8.911 5.322 8.911 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.221 7.974 mp_irecv_dv 3311 11.0 3.201 7.917 3.201 7.917 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.990 7.010 calculate_norms 2376 9.8 6.076 6.774 6.076 6.774 multiply_cannon_sync_h2d 1332 9.7 4.876 6.161 4.876 6.161 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.897 5.137 arnoldi_extremal 4 6.8 0.000 0.000 4.649 4.662 arnoldi_normal_ev 4 7.8 0.001 0.004 4.649 4.662 build_subspace 16 8.4 0.014 0.021 4.392 4.394 ls_scf_post 1 4.0 0.000 0.002 3.723 3.727 ls_scf_store_result 1 5.0 0.000 0.000 3.420 3.548 dbcsr_matrix_vector_mult 304 9.0 0.010 0.023 3.166 3.405 dbcsr_matrix_vector_mult_local 304 10.0 2.785 3.275 2.787 3.277 mp_allgather_i34 111 8.7 0.847 3.097 0.847 3.097 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.085 3.028 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.643 2.746 dbcsr_data_new 4174 10.1 2.113 2.394 2.113 2.394 make_images_pack 222 9.7 1.823 2.126 1.826 2.128 dbcsr_sort_data 436 11.2 1.829 2.044 1.829 2.044 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.885 1.887 rebuild_ks_matrix 3 7.3 0.000 0.000 1.872 1.875 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.008 1.872 1.875 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.550000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1805.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_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 2.904117E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.042 92.867 92.868 qs_energies 1 2.0 0.000 0.000 92.287 92.293 ls_scf 1 3.0 0.000 0.000 90.839 90.844 dbcsr_multiply_generic 111 6.7 0.016 0.017 75.485 75.730 ls_scf_main 1 4.0 0.000 0.000 56.786 56.790 multiply_cannon 111 7.7 0.032 0.090 52.245 56.042 multiply_cannon_loop 111 8.7 0.117 0.131 49.469 53.551 density_matrix_trs4 2 5.0 0.002 0.003 50.828 51.050 mp_waitall_1 7281 11.0 23.551 33.403 23.551 33.403 ls_scf_init_scf 1 4.0 0.000 0.000 30.411 30.414 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.218 29.281 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 26.820 26.834 multiply_cannon_multrec 888 9.7 12.755 15.146 21.357 24.381 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.892 23.106 make_m2s 222 7.7 0.007 0.007 16.416 17.181 make_images 222 8.7 1.589 1.850 16.379 17.141 make_images_data 222 9.7 0.004 0.005 9.491 10.650 hybrid_alltoall_any 227 10.6 0.644 2.950 9.069 10.321 dbcsr_mm_accdrv_process 3754 10.4 0.310 0.507 8.139 9.371 mp_sum_l 887 5.1 5.126 8.891 5.126 8.891 dbcsr_mm_accdrv_process_sort 3754 11.4 7.699 8.864 7.699 8.864 multiply_cannon_sync_h2d 888 9.7 6.101 7.309 6.101 7.309 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.821 7.290 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.963 7.156 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.465 7.050 mp_irecv_dv 2335 11.1 2.448 6.983 2.448 6.983 arnoldi_extremal 4 6.8 0.000 0.000 5.107 5.119 arnoldi_normal_ev 4 7.8 0.001 0.005 5.107 5.119 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.766 5.033 build_subspace 16 8.4 0.014 0.020 4.807 4.815 calculate_norms 1584 9.8 4.341 4.741 4.341 4.741 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.468 3.812 mp_allgather_i34 111 8.7 0.886 3.800 0.886 3.800 dbcsr_matrix_vector_mult_local 304 10.0 3.067 3.655 3.069 3.657 ls_scf_post 1 4.0 0.000 0.000 3.642 3.647 ls_scf_store_result 1 5.0 0.000 0.000 3.392 3.465 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.882 2.982 dbcsr_data_new 4116 9.9 2.109 2.439 2.109 2.439 make_images_sizes 222 9.7 0.000 0.000 1.069 2.295 mp_alltoall_i44 222 10.7 1.068 2.294 1.068 2.294 dbcsr_sort_data 325 11.1 1.884 2.135 1.884 2.135 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.909 1.911 rebuild_ks_matrix 3 7.3 0.000 0.000 1.891 1.893 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.891 1.893 make_images_pack 222 9.7 1.631 1.885 1.634 1.889 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.868000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2271.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_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 3.359822E+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 0.032 0.047 97.371 97.372 qs_energies 1 2.0 0.000 0.000 96.738 96.747 ls_scf 1 3.0 0.000 0.001 95.060 95.068 dbcsr_multiply_generic 111 6.7 0.018 0.021 78.757 78.996 ls_scf_main 1 4.0 0.000 0.002 59.053 59.054 multiply_cannon 111 7.7 0.041 0.111 52.057 55.979 density_matrix_trs4 2 5.0 0.002 0.004 52.891 53.000 multiply_cannon_loop 111 8.7 0.152 0.167 47.102 49.745 ls_scf_init_scf 1 4.0 0.000 0.002 32.756 32.758 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.523 31.595 mp_waitall_1 6369 11.0 22.978 30.154 22.978 30.154 matrix_sqrt_Newton_Schulz 2 6.5 0.010 0.017 29.022 29.038 multiply_cannon_multrec 1332 9.7 14.242 17.117 22.221 24.990 make_m2s 222 7.7 0.008 0.010 21.208 22.550 make_images 222 8.7 3.155 3.606 21.156 22.500 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.332 16.915 make_images_data 222 9.7 0.004 0.005 11.833 13.418 hybrid_alltoall_any 227 10.6 0.803 3.783 11.229 12.947 dbcsr_mm_accdrv_process 3641 10.4 0.284 0.509 7.620 9.149 dbcsr_mm_accdrv_process_sort 3641 11.4 7.171 8.653 7.171 8.653 mp_sum_l 887 5.1 3.889 7.005 3.889 7.005 multiply_cannon_sync_h2d 1332 9.7 5.505 6.016 5.505 6.016 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.090 5.953 mp_irecv_dv 3229 10.9 2.064 5.868 2.064 5.868 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.983 5.395 arnoldi_extremal 4 6.8 0.000 0.000 5.222 5.243 arnoldi_normal_ev 4 7.8 0.008 0.017 5.222 5.242 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.655 5.136 build_subspace 16 8.4 0.014 0.021 4.881 4.887 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.546 4.727 calculate_norms 2376 9.8 4.178 4.550 4.178 4.550 mp_allgather_i34 111 8.7 2.058 4.320 2.058 4.320 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.623 3.917 dbcsr_matrix_vector_mult_local 304 10.0 3.245 3.741 3.247 3.743 dbcsr_sort_data 658 11.4 3.059 3.452 3.059 3.452 ls_scf_post 1 4.0 0.001 0.005 3.250 3.259 dbcsr_special_finalize 555 9.7 0.006 0.007 2.802 3.204 dbcsr_merge_single_wm 555 10.7 0.536 0.663 2.793 3.196 ls_scf_dm_to_ks 2 5.0 0.000 0.001 3.057 3.133 ls_scf_store_result 1 5.0 0.000 0.000 2.996 3.053 dbcsr_data_release 10477 10.7 1.600 2.413 1.600 2.413 dbcsr_finalize 304 7.8 0.050 0.062 1.821 1.993 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.372000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2725.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_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 4.762010E+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 0.039 0.056 92.096 92.096 qs_energies 1 2.0 0.000 0.000 91.348 91.352 ls_scf 1 3.0 0.000 0.000 89.411 89.415 dbcsr_multiply_generic 111 6.7 0.018 0.018 70.875 71.062 ls_scf_main 1 4.0 0.000 0.000 56.897 56.898 multiply_cannon 111 7.7 0.084 0.129 52.901 56.266 multiply_cannon_loop 111 8.7 0.089 0.096 50.311 51.460 density_matrix_trs4 2 5.0 0.002 0.003 49.905 49.969 ls_scf_init_scf 1 4.0 0.000 0.000 29.126 29.127 mp_waitall_1 5436 11.0 24.688 28.614 24.688 28.614 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.850 27.878 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.768 25.778 multiply_cannon_multrec 444 9.7 13.567 16.662 20.676 22.223 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.650 15.611 make_m2s 222 7.7 0.005 0.006 13.419 14.364 make_images 222 8.7 2.046 2.489 13.351 14.295 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.691 13.333 make_images_data 222 9.7 0.003 0.004 8.168 9.598 hybrid_alltoall_any 227 10.6 0.804 3.810 8.083 9.585 dbcsr_mm_accdrv_process 3003 10.4 0.340 0.404 6.799 7.952 multiply_cannon_sync_h2d 444 9.7 6.761 7.792 6.761 7.792 dbcsr_mm_accdrv_process_sort 3003 11.4 6.436 7.548 6.436 7.548 arnoldi_extremal 4 6.8 0.000 0.000 5.878 5.889 arnoldi_normal_ev 4 7.8 0.002 0.005 5.877 5.889 build_subspace 16 8.4 0.015 0.019 5.492 5.502 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.367 4.519 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.240 4.455 dbcsr_matrix_vector_mult_local 304 10.0 3.782 4.267 3.784 4.269 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.552 3.911 mp_sum_l 887 5.1 2.532 3.910 2.532 3.910 mp_irecv_dv 1241 11.2 1.539 3.872 1.539 3.872 calculate_norms 792 9.8 3.638 3.747 3.638 3.747 mp_allgather_i34 111 8.7 1.157 3.710 1.157 3.710 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.592 3.650 ls_scf_post 1 4.0 0.000 0.000 3.388 3.392 ls_scf_store_result 1 5.0 0.000 0.000 3.176 3.213 make_images_sizes 222 9.7 0.000 0.000 0.845 3.107 mp_alltoall_i44 222 10.7 0.844 3.106 0.844 3.106 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.782 2.887 dbcsr_finalize 304 7.8 0.062 0.077 2.205 2.276 dbcsr_data_new 4608 9.7 1.798 2.230 1.798 2.230 dbcsr_merge_all 275 8.9 0.481 0.531 2.063 2.115 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.076 2.077 rebuild_ks_matrix 3 7.3 0.000 0.000 2.044 2.045 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.044 2.045 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 1.920 1.920 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.096000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3695.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_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 8.819180E+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 0.075 0.095 110.827 110.828 qs_energies 1 2.0 0.000 0.000 109.443 109.451 ls_scf 1 3.0 0.000 0.000 106.486 106.494 dbcsr_multiply_generic 111 6.7 0.024 0.027 79.775 79.928 ls_scf_main 1 4.0 0.000 0.000 64.170 64.171 density_matrix_trs4 2 5.0 0.002 0.003 55.094 55.165 multiply_cannon 111 7.7 0.162 0.361 53.302 55.130 multiply_cannon_loop 111 8.7 0.099 0.101 50.285 50.999 ls_scf_init_scf 1 4.0 0.000 0.000 38.516 38.517 ls_scf_init_matrix_S 1 5.0 0.000 0.000 36.938 36.966 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 34.115 34.127 mp_waitall_1 4527 11.1 22.462 26.882 22.462 26.882 make_m2s 222 7.7 0.006 0.006 22.658 23.739 make_images 222 8.7 3.578 3.894 22.547 23.628 multiply_cannon_multrec 444 9.7 17.889 18.491 22.553 23.147 hybrid_alltoall_any 227 10.6 1.657 3.617 12.777 15.765 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.866 15.727 make_images_data 222 9.7 0.004 0.004 13.018 15.190 multiply_cannon_sync_h2d 444 9.7 8.789 8.847 8.789 8.847 arnoldi_extremal 4 6.8 0.000 0.000 7.541 7.558 arnoldi_normal_ev 4 7.8 0.003 0.009 7.541 7.557 build_subspace 16 8.4 0.026 0.036 6.980 6.990 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.611 5.773 dbcsr_matrix_vector_mult_local 304 10.0 5.180 5.502 5.183 5.505 compute_matrix_preconditioner 1 6.0 0.002 0.002 5.458 5.466 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.179 5.277 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.929 5.184 acc_transpose_blocks 444 9.7 0.003 0.003 4.361 4.958 acc_transpose_blocks_kernels 444 10.7 0.006 0.006 4.276 4.870 jit_kernel_transpose 1 13.0 4.271 4.865 4.271 4.865 dbcsr_mm_accdrv_process 1814 10.4 0.248 0.357 4.485 4.605 dbcsr_mm_accdrv_process_sort 1814 11.4 4.146 4.274 4.146 4.274 ls_scf_post 1 4.0 0.000 0.000 3.800 3.808 make_images_sizes 222 9.7 0.000 0.000 1.474 3.742 mp_alltoall_i44 222 10.7 1.474 3.742 1.474 3.742 mp_allgather_i34 111 8.7 1.109 3.620 1.109 3.620 ls_scf_store_result 1 5.0 0.000 0.000 3.514 3.551 calculate_norms 792 9.8 3.230 3.276 3.230 3.276 dbcsr_finalize 304 7.8 0.082 0.090 3.079 3.188 dbcsr_merge_all 275 8.9 0.888 0.918 2.862 2.966 qs_energies_init_hamiltonians 1 3.0 0.003 0.005 2.926 2.926 dbcsr_complete_redistribute 5 7.6 1.457 1.518 2.762 2.885 matrix_ls_to_qs 2 6.0 0.000 0.000 2.433 2.556 dbcsr_sort_data 325 11.1 2.438 2.516 2.438 2.516 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.466 2.467 dbcsr_data_new 6591 9.6 1.864 2.432 1.864 2.432 rebuild_ks_matrix 3 7.3 0.000 0.000 2.398 2.400 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.398 2.400 dbcsr_new_transposed 4 7.5 0.255 0.285 2.338 2.358 mp_sum_l 887 5.1 1.396 2.309 1.396 2.309 dbcsr_frobenius_norm 74 6.6 2.057 2.132 2.186 2.223 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=110.828000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6984.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/bf5cdd7210383fb7939885873c6db45d8a334c42_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 588.161024E+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 4037197. 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 0.019 0.050 227.661 227.664 qs_mol_dyn_low 1 2.0 0.003 0.006 226.693 226.720 qs_forces 5 3.8 0.004 0.007 226.611 226.613 qs_energies 5 4.8 0.002 0.003 223.683 223.704 scf_env_do_scf 5 5.8 0.000 0.001 206.764 206.767 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 180.560 180.562 qs_scf_new_mos 105 7.6 0.000 0.001 140.229 140.408 qs_scf_loop_do_ot 105 8.6 0.001 0.001 140.228 140.407 dbcsr_multiply_generic 1445 12.2 0.134 0.143 135.431 135.827 ot_scf_mini 105 9.6 0.003 0.004 130.460 130.599 multiply_cannon 1445 13.2 0.275 0.284 116.707 118.789 multiply_cannon_loop 1445 14.2 2.856 3.024 114.988 116.098 velocity_verlet 4 3.0 0.001 0.002 105.722 105.723 ot_mini 105 10.6 0.001 0.001 59.893 60.008 multiply_cannon_multrec 69360 15.2 29.779 34.729 39.534 44.528 mp_waitall_1 488190 16.1 34.319 41.389 34.319 41.389 qs_ot_get_p 112 10.4 0.001 0.001 40.685 40.970 qs_ot_get_derivative 55 11.6 0.001 0.001 38.095 38.202 multiply_cannon_sync_h2d 69360 15.2 29.265 33.495 29.265 33.495 multiply_cannon_metrocomm3 69360 15.2 0.206 0.218 25.557 32.704 qs_ot_p2m_diag 40 11.0 0.020 0.030 29.688 29.756 rebuild_ks_matrix 110 8.4 0.000 0.000 28.856 29.020 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.013 28.855 29.020 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.512 26.660 cp_dbcsr_syevd 40 12.0 0.002 0.003 26.364 26.365 init_scf_loop 7 6.6 0.000 0.001 26.175 26.176 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.126 23.344 apply_single 62 13.6 0.000 0.000 23.126 23.344 cp_fm_syevd 40 13.0 0.000 0.000 21.279 21.430 prepare_preconditioner 7 7.6 0.000 0.000 21.326 21.359 make_preconditioner 7 8.6 0.000 0.000 21.326 21.359 ot_new_cg_direction 55 11.6 0.001 0.001 21.077 21.078 qs_rho_update_rho_low 110 7.6 0.001 0.001 17.219 17.584 calculate_rho_elec 110 8.6 0.030 0.033 17.218 17.583 cp_fm_redistribute_end 40 14.0 8.322 16.603 8.326 16.605 cp_fm_syevd_base 40 14.0 8.270 16.552 8.270 16.552 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.713 14.872 init_scf_run 5 5.8 0.000 0.001 14.358 14.359 scf_env_initial_rho_setup 5 6.8 0.001 0.002 14.358 14.359 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.241 14.307 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.666 13.767 mp_sum_l 4764 12.2 11.921 12.645 11.921 12.645 density_rs2pw 110 9.6 0.006 0.007 11.257 12.440 rs_pw_transfer 690 11.5 0.011 0.013 11.246 12.410 pw_transfer 1645 12.4 0.083 0.107 11.967 12.225 fft_wrap_pw1pw2 1425 13.5 0.013 0.017 11.824 12.085 calculate_dm_sparse 110 9.5 0.000 0.001 11.320 11.489 acc_transpose_blocks 69360 15.2 0.354 0.371 10.105 10.530 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.423 10.505 dbcsr_mm_accdrv_process 154766 15.8 6.151 6.372 9.623 10.436 qs_vxc_create 110 10.4 0.002 0.004 10.384 10.431 fft_wrap_pw1pw2_240 915 15.0 1.191 1.298 10.031 10.261 cp_fm_cholesky_invert 7 10.6 9.984 9.991 9.984 9.991 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 9.959 9.976 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 9.596 9.649 check_diag 80 13.5 8.595 8.887 9.399 9.550 calculate_first_density_matrix 1 7.0 0.000 0.001 9.497 9.520 sum_up_and_integrate 60 10.3 0.028 0.031 8.495 8.506 integrate_v_rspace 60 11.3 0.002 0.002 8.467 8.479 fft3d_pb 915 16.0 2.377 2.682 8.099 8.401 multiply_cannon_metrocomm1 69360 15.2 0.093 0.099 4.587 7.745 acc_transpose_blocks_kernels 69360 16.2 0.835 0.886 7.342 7.669 xc_rho_set_and_dset_create 110 12.4 0.078 0.101 7.322 7.584 xc_vxc_pw_create 60 11.3 0.039 0.050 6.943 6.990 make_m2s 2890 13.2 0.082 0.092 6.308 6.953 make_full_single_inverse 7 9.6 0.001 0.001 6.817 6.849 make_images 2890 14.2 0.240 0.261 6.200 6.846 jit_kernel_transpose 5 15.0 6.507 6.807 6.507 6.807 xc_pw_derive 510 13.4 0.005 0.007 6.147 6.251 mp_waitany 7680 13.5 4.351 5.612 4.351 5.612 mp_alltoall_z22v 2340 17.7 5.184 5.444 5.184 5.444 potential_pw2rs 60 12.3 0.003 0.003 4.727 4.769 multiply_cannon_metrocomm4 67915 15.2 0.182 0.199 2.005 4.761 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=227.664000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=559.600000, yerr=1.743560 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: bf5cdd7210383fb7939885873c6db45d8a334c42 Summary: empty Status: OK