=== 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: 1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SPGLIB 2.3.1, LIBGRPP 20231215, # SPFFT 1.0.6, SPLA 1.5.5 # # 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: 28.03.2024 # # \ 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; \ rm -rf build; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-hdf5 --with-libvdwxc --with-pexsi --with-plumed --with-sirius=no --with-spfft --with-spla; \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 #USE_DEEPMD := 2.2.7 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 #USE_SIRIUS := 7.5.2 USE_SPFFT := 1.0.6 USE_SPGLIB := 2.3.1 USE_SPLA := 1.5.5 # Only needed for SIRIUS #LIBVDWXC_VER := 0.4.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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_DEEPMD),) USE_DEEPMD := $(strip $(USE_DEEPMD)) DEEPMD_INC := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/include DEEPMD_LIB := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/lib CFLAGS += -I$(DEEPMD_INC) DFLAGS += -D__DEEPMD LIBS += $(DEEPMD_LIB)/libdeepmd.so LIBS += $(DEEPMD_LIB)/libdeepmd_c.so LIBS += $(DEEPMD_LIB)/libdeepmd_cc.so LIBS += $(DEEPMD_LIB)/libdeepmd_dyn_cudart.so LIBS += $(DEEPMD_LIB)/libdeepmd_op.so LIBS += $(DEEPMD_LIB)/libdeepmd_op_cuda.so LIBS += $(DEEPMD_LIB)/libtensorflow_cc.so.2 LIBS += $(DEEPMD_LIB)/libtensorflow_framework.so.2 endif ifneq ($(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_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib ifeq ($(USE_ACC), yes) SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_SPFFT),) USE_SPFFT := $(strip $(USE_SPFFT)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/include ifeq ($(USE_ACC), yes) SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/lib endif CFLAGS += -I$(SPFFT_INC) DFLAGS += -D__SPFFT LIBS += $(SPFFT_LIB)/libspfft.a endif ifneq ($(USE_SPLA),) USE_SPLA := $(strip $(USE_SPLA)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPLA_LIB := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/lib/cuda else SPLA_LIB := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/lib endif CFLAGS += -I$(SPLA_INC) DFLAGS += -D__SPLA LIBS += $(SPLA_LIB)/libspla.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_f90cstub.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/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/01 job id: 52726360 --- 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/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/02 job id: 52726364 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/03 job id: 52726368 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/04 job id: 52726370 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/05 job id: 52726372 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/06 job id: 52726374 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/07 job id: 52726375 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/08 job id: 52726376 --- 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/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/09 job id: 52726377 --- 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/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/10 job id: 52726378 --- 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/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/11 job id: 52726379 --- 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/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/12 job id: 52726380 --- 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/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/13 job id: 52726381 --- 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/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/14 job id: 52726383 --- 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/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/15 job id: 52726387 --- 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/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/16 job id: 52726388 --- 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/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/17 job id: 52726389 --- 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/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/18 job id: 52726391 --- 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/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/19 job id: 52726392 --- 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/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/20 job id: 52726393 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/21 job id: 52726394 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/22 job id: 52726395 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/23 job id: 52726397 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/24 job id: 52726398 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/25 job id: 52726399 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/26 job id: 52726400 --- 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] = 20 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/27 job id: 52726401 --- 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/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 172669. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.036 128.457 128.458 farming_run 1 2.0 127.934 127.935 128.427 128.431 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.545056E+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 822089. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1131389. 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 0.007 0.028 109.892 109.893 qs_energies 1 2.0 0.001 0.001 109.702 109.704 mp2_main 1 3.0 0.000 0.000 107.886 107.887 mp2_gpw_main 1 4.0 0.032 0.053 106.989 106.991 mp2_ri_gpw_compute_in 1 5.0 0.174 0.176 88.373 88.560 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 57.301 57.490 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.172 43.742 48.667 get_2c_integrals 1 6.0 0.008 0.009 30.595 30.897 integrate_v_rspace 273 8.0 0.442 0.456 25.619 30.156 fft_wrap_pw1pw2 5465 10.4 0.068 0.070 27.982 28.561 grid_integrate_task_list 273 9.0 20.863 26.357 20.863 26.357 fft_wrap_pw1pw2_100 2178 11.4 1.275 1.438 25.432 25.984 compute_2c_integrals 1 7.0 0.002 0.002 21.331 21.333 compute_2c_integrals_loop_lm 1 8.0 0.001 0.003 20.710 21.066 mp2_eri_2c_integrate_gpw 1 9.0 2.393 2.431 20.708 21.063 rpa_ri_compute_en 1 5.0 0.019 0.023 18.491 18.733 fft3d_s 5443 12.4 16.976 17.428 16.997 17.448 ao_to_mo_and_store_B_mult_1 272 7.0 10.773 15.303 10.773 15.303 calculate_wavefunction 272 8.0 5.282 5.441 12.748 13.603 calc_potential_gpw 544 9.5 0.004 0.004 10.941 11.658 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 9.730 10.509 rpa_num_int 1 6.0 0.002 0.034 10.438 10.438 rpa_num_int_RPA_matrix_operati 8 7.0 0.005 0.007 10.424 10.431 cp_fm_cholesky_decompose 12 8.2 9.416 9.711 9.416 9.711 calc_mat_Q 8 8.0 0.000 0.000 9.412 9.490 potential_pw2rs 545 10.0 0.106 0.108 8.370 9.384 contract_S_to_Q 8 9.0 0.000 0.000 8.836 8.909 cholesky_decomp 1 7.0 0.000 0.000 8.456 8.753 collocate_single_gaussian 272 10.0 0.040 0.042 7.780 8.623 parallel_gemm_fm 14 9.1 0.000 0.000 8.461 8.580 parallel_gemm_fm_cosma 14 10.1 8.461 8.580 8.461 8.580 create_integ_mat 1 6.0 0.028 0.034 7.773 7.773 array2fm 1 7.0 0.000 0.000 6.693 7.202 pw_scatter_s 2720 12.7 4.507 4.868 4.507 4.868 pw_poisson_solve 545 10.5 0.010 0.010 4.656 4.836 pw_gather_s 2722 12.2 3.784 4.611 3.784 4.611 pw_poisson_set 548 11.5 0.017 0.018 3.366 3.559 array2fm_buffer_send 1 8.0 3.020 3.160 3.020 3.160 pw_copy 4911 11.6 2.624 2.732 2.624 2.732 ao_to_mo_and_store_B_E_Ex_1 272 7.0 1.077 1.230 2.021 2.230 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=106.988023, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2802.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 200775. 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 0.026 0.044 428.685 428.687 farming_run 1 2.0 427.756 427.763 428.632 428.634 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.269531E+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 713945. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 480173. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171443. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.033 223.377 223.378 qs_energies 1 2.0 0.000 0.001 223.090 223.122 scf_env_do_scf 1 3.0 0.000 0.000 115.186 115.186 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.265 114.272 rebuild_ks_matrix 4 6.0 0.000 0.000 114.264 114.271 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 114.264 114.271 hfx_ks_matrix 4 8.0 0.001 0.001 113.912 113.915 integrate_four_center 4 9.0 0.153 0.462 113.911 113.914 mp2_main 1 3.0 0.000 0.000 107.603 107.635 mp2_gpw_main 1 4.0 0.049 0.074 106.724 106.757 integrate_four_center_main 4 10.0 0.118 0.603 102.055 105.861 integrate_four_center_bin 263 11.0 101.937 105.833 101.937 105.833 init_scf_loop 1 4.0 0.000 0.000 96.952 96.952 mp2_ri_gpw_compute_in 1 5.0 0.065 0.068 79.853 80.958 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 56.789 57.892 mp2_eri_3c_integrate_gpw 91 7.0 0.146 0.162 44.588 49.464 integrate_v_rspace 95 8.0 0.397 0.560 28.253 32.972 fft_wrap_pw1pw2 1868 10.4 0.030 0.037 30.542 31.055 mp2_ri_gpw_compute_en 1 5.0 0.058 0.090 26.718 28.994 fft_wrap_pw1pw2_100 730 11.4 0.644 0.719 28.245 28.760 grid_integrate_task_list 95 9.0 23.551 28.483 23.551 28.483 ao_to_mo_and_store_B_mult_1 91 7.0 10.528 26.168 10.528 26.168 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.833 1.919 24.956 24.967 get_2c_integrals 1 6.0 0.000 0.000 22.971 22.998 compute_2c_integrals 1 7.0 0.002 0.003 22.134 22.144 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 21.752 22.003 mp2_eri_2c_integrate_gpw 1 9.0 1.736 1.891 21.751 22.002 fft3d_s 1823 12.4 19.191 20.077 19.205 20.090 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 18.232 18.232 calc_potential_gpw 182 9.5 0.002 0.002 12.363 13.328 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 10.702 11.376 calculate_wavefunction 91 8.0 1.982 2.025 10.006 10.788 potential_pw2rs 186 10.0 0.033 0.035 9.045 10.525 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.579 8.739 9.320 collocate_single_gaussian 91 10.0 0.017 0.024 8.326 8.970 local_gemm 172 8.0 8.183 8.764 8.183 8.764 mp2_ri_gpw_compute_en_comm 22 7.0 0.499 0.530 7.996 8.574 mp_sync 37 10.5 3.843 7.073 3.843 7.073 integrate_four_center_load 4 10.0 0.000 0.000 6.744 6.749 hfx_load_balance 1 11.0 0.000 0.000 6.744 6.749 mp_sendrecv_dm3 2068 8.0 6.033 6.589 6.033 6.589 mp2_ri_gpw_compute_en_ener 172 7.0 6.343 6.442 6.343 6.442 pw_poisson_solve 186 10.4 0.005 0.006 5.557 6.062 pw_gather_s 912 12.2 4.853 5.550 4.853 5.550 pw_scatter_s 910 12.7 4.110 4.784 4.110 4.784 pw_poisson_set 189 11.4 0.009 0.010 4.108 4.651 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=106.721924, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1525.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 445.435904E+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 3690 72628. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.051 0.074 57.416 57.417 qs_mol_dyn_low 1 2.0 0.004 0.009 57.099 57.114 qs_forces 11 3.9 0.002 0.003 57.035 57.037 qs_energies 11 4.9 0.002 0.003 55.471 55.497 scf_env_do_scf 11 5.9 0.000 0.001 48.568 48.568 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 46.551 46.551 dbcsr_multiply_generic 2286 12.5 0.095 0.097 36.969 37.548 qs_scf_new_mos 108 7.5 0.000 0.001 36.267 36.539 qs_scf_loop_do_ot 108 8.5 0.000 0.001 36.266 36.539 ot_scf_mini 108 9.5 0.002 0.002 34.492 34.662 multiply_cannon 2286 13.5 0.179 0.194 28.689 30.095 multiply_cannon_loop 2286 14.5 1.839 1.953 27.975 29.394 velocity_verlet 10 3.0 0.002 0.005 27.794 27.796 ot_mini 108 10.5 0.001 0.001 20.928 21.176 qs_ot_get_derivative 108 11.5 0.001 0.001 17.814 18.003 mp_waitall_1 245248 16.5 10.219 16.474 10.219 16.474 multiply_cannon_metrocomm3 54864 15.5 0.072 0.077 7.015 14.172 multiply_cannon_multrec 54864 15.5 3.546 5.489 8.432 12.066 qs_ot_get_p 119 10.4 0.008 0.016 8.604 8.878 rebuild_ks_matrix 119 8.3 0.000 0.000 8.087 8.238 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.039 8.087 8.237 mp_sum_l 7287 12.8 6.165 8.021 6.165 8.021 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.093 7.239 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.354 6.764 dbcsr_mm_accdrv_process 76910 16.1 1.893 2.996 4.801 6.725 multiply_cannon_sync_h2d 54864 15.5 5.065 6.493 5.065 6.493 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.560 5.685 init_scf_run 11 5.9 0.000 0.001 5.677 5.678 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.677 5.677 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.405 5.470 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.582 4.583 sum_up_and_integrate 119 10.3 0.001 0.002 4.524 4.551 integrate_v_rspace 119 11.3 0.002 0.002 4.513 4.540 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.373 4.374 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.220 4.366 calculate_rho_elec 119 8.7 0.011 0.016 4.219 4.365 cp_fm_redistribute_end 50 14.0 2.229 4.345 2.235 4.349 cp_fm_diag_elpa_base 50 14.0 2.107 4.233 2.111 4.240 calculate_dm_sparse 119 9.5 0.000 0.000 3.809 3.978 jit_kernel_multiply 13 15.8 2.843 3.774 2.843 3.774 calculate_first_density_matrix 1 7.0 0.000 0.000 3.732 3.740 multiply_cannon_metrocomm1 54864 15.5 0.056 0.062 2.465 3.652 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.983 3.202 apply_single 119 13.6 0.000 0.000 2.983 3.202 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.813 2.828 ot_diis_step 108 11.5 0.006 0.006 2.817 2.818 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.737 2.815 acc_transpose_blocks 54864 15.5 0.216 0.238 2.235 2.745 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.491 2.549 density_rs2pw 119 9.7 0.003 0.004 2.180 2.280 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.210 2.238 mp_sum_d 4139 12.0 1.467 2.115 1.467 2.115 grid_integrate_task_list 119 12.3 2.035 2.111 2.035 2.111 init_scf_loop 11 6.9 0.000 0.000 1.991 1.991 potential_pw2rs 119 12.3 0.003 0.004 1.882 1.916 wfi_extrapolate 11 7.9 0.001 0.001 1.868 1.868 make_m2s 4572 13.5 0.054 0.056 1.806 1.860 make_images 4572 14.5 0.132 0.138 1.724 1.777 fft_wrap_pw1pw2 1201 11.6 0.011 0.012 1.657 1.732 mp_alltoall_d11v 2130 13.8 1.403 1.587 1.403 1.587 transfer_rs2pw 487 10.6 0.006 0.006 1.421 1.509 fft3d_ps 1201 13.6 0.368 0.471 1.433 1.499 acc_transpose_blocks_sync 164592 16.5 1.210 1.453 1.210 1.453 mp_waitany 12084 13.8 1.306 1.452 1.306 1.452 transfer_pw2rs 487 13.2 0.005 0.006 1.409 1.431 grid_collocate_task_list 119 9.7 1.350 1.407 1.350 1.407 fft_wrap_pw1pw2_140 487 12.2 0.048 0.053 1.303 1.381 dbcsr_dot_sd 1205 11.9 0.047 0.059 0.808 1.223 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=57.417000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=424.545455, yerr=0.890724 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.366080E+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 3679 72937. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 923281. 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 0.013 0.032 40.004 40.005 qs_mol_dyn_low 1 2.0 0.005 0.021 39.671 39.678 qs_forces 11 3.9 0.002 0.003 39.608 39.611 qs_energies 11 4.9 0.002 0.003 37.922 37.928 scf_env_do_scf 11 5.9 0.000 0.001 32.437 32.437 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 30.141 30.141 dbcsr_multiply_generic 2286 12.5 0.102 0.105 22.918 23.247 qs_scf_new_mos 108 7.5 0.001 0.001 21.334 21.559 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.333 21.558 ot_scf_mini 108 9.5 0.002 0.003 20.403 20.561 multiply_cannon 2286 13.5 0.208 0.215 17.505 18.901 velocity_verlet 10 3.0 0.001 0.005 18.535 18.536 multiply_cannon_loop 2286 14.5 1.190 1.242 16.297 17.832 ot_mini 108 10.5 0.001 0.001 12.557 12.787 mp_waitall_1 200699 16.5 5.829 11.288 5.829 11.288 qs_ot_get_derivative 108 11.5 0.001 0.001 10.187 10.348 multiply_cannon_metrocomm3 27432 15.5 0.070 0.073 4.192 9.912 multiply_cannon_multrec 27432 15.5 1.838 4.106 6.512 9.385 rebuild_ks_matrix 119 8.3 0.000 0.000 6.919 7.058 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.919 7.058 dbcsr_mm_accdrv_process 47894 16.0 3.628 5.996 4.596 6.837 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.091 6.216 qs_ot_get_p 119 10.4 0.002 0.010 4.901 5.119 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.800 4.720 init_scf_run 11 5.9 0.000 0.001 4.273 4.274 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.273 4.273 mp_sum_l 7287 12.8 2.171 4.176 2.171 4.176 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.070 4.067 apply_single 119 13.6 0.000 0.000 3.070 4.067 sum_up_and_integrate 119 10.3 0.001 0.001 3.772 3.778 integrate_v_rspace 119 11.3 0.002 0.002 3.758 3.763 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.678 3.703 calculate_rho_elec 119 8.7 0.021 0.024 3.677 3.702 qs_ot_p2m_diag 50 11.0 0.008 0.013 3.149 3.167 make_m2s 4572 13.5 0.051 0.053 2.807 3.151 make_images 4572 14.5 0.204 0.241 2.717 3.065 calculate_first_density_matrix 1 7.0 0.000 0.000 2.737 2.739 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.738 2.738 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.390 2.475 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.383 2.384 cp_fm_redistribute_end 50 14.0 1.206 2.351 1.212 2.355 ot_diis_step 108 11.5 0.011 0.011 2.317 2.318 cp_fm_diag_elpa_base 50 14.0 1.108 2.254 1.138 2.296 calculate_dm_sparse 119 9.5 0.000 0.001 2.201 2.282 init_scf_loop 11 6.9 0.000 0.000 2.276 2.276 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.202 2.204 multiply_cannon_sync_h2d 27432 15.5 1.700 2.189 1.700 2.189 density_rs2pw 119 9.7 0.003 0.004 1.975 2.063 acc_transpose_blocks 27432 15.5 0.112 0.118 1.687 2.056 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.997 2.036 grid_integrate_task_list 119 12.3 1.840 1.934 1.840 1.934 jit_kernel_multiply 8 16.3 0.908 1.876 0.908 1.876 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.736 1.769 make_images_data 4572 15.5 0.047 0.053 1.297 1.625 potential_pw2rs 119 12.3 0.006 0.006 1.502 1.510 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.140 1.503 fft3d_ps 1201 13.6 0.511 0.562 1.457 1.484 wfi_extrapolate 11 7.9 0.001 0.001 1.479 1.479 prepare_preconditioner 11 7.9 0.000 0.000 1.368 1.395 make_preconditioner 11 8.9 0.000 0.000 1.368 1.395 mp_alltoall_d11v 2130 13.8 1.261 1.392 1.261 1.392 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.383 1.391 fft_wrap_pw1pw2_140 487 12.2 0.047 0.049 1.341 1.377 grid_collocate_task_list 119 9.7 1.287 1.353 1.287 1.353 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.268 1.311 transfer_rs2pw 487 10.6 0.005 0.006 1.205 1.297 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.227 1.275 mp_allgather_i34 2286 14.5 0.633 1.073 0.633 1.073 mp_sum_d 4139 12.0 0.560 1.041 0.560 1.041 transfer_pw2rs 487 13.2 0.004 0.005 0.989 0.997 acc_transpose_blocks_kernels 27432 16.5 0.189 0.278 0.732 0.981 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.944 0.947 acc_transpose_blocks_sync 82296 16.5 0.817 0.939 0.817 0.939 make_images_sizes 4572 15.5 0.005 0.005 0.633 0.908 mp_alltoall_i44 4572 16.5 0.629 0.903 0.629 0.903 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.885 0.900 mp_alltoall_z22v 1201 15.6 0.728 0.801 0.728 0.801 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.005000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.000000, yerr=1.858641 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 520.503296E+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 3629 73909. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.054 38.541 38.541 qs_mol_dyn_low 1 2.0 0.004 0.008 36.625 38.271 qs_forces 11 3.9 0.002 0.003 32.915 32.917 qs_energies 11 4.9 0.002 0.002 31.331 31.333 scf_env_do_scf 11 5.9 0.000 0.001 26.207 26.207 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 23.871 23.872 velocity_verlet 10 3.0 0.002 0.005 18.911 19.025 dbcsr_multiply_generic 2286 12.5 0.096 0.098 17.513 17.581 qs_scf_new_mos 108 7.5 0.001 0.001 15.942 15.967 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.942 15.967 ot_scf_mini 108 9.5 0.002 0.002 15.162 15.183 multiply_cannon 2286 13.5 0.191 0.196 14.006 14.775 multiply_cannon_loop 2286 14.5 0.866 0.909 13.198 13.991 ot_mini 108 10.5 0.001 0.001 9.270 9.287 qs_ot_get_derivative 108 11.5 0.001 0.001 7.760 7.780 multiply_cannon_multrec 18288 15.5 1.859 2.890 7.202 7.431 rebuild_ks_matrix 119 8.3 0.000 0.000 6.142 6.155 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 6.141 6.155 dbcsr_mm_accdrv_process 38222 16.0 5.191 6.044 5.251 6.112 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.408 5.420 md_write_output 11 3.9 0.012 0.347 0.169 5.334 md_output 10 3.0 0.000 0.000 0.171 5.333 write_trajectory 44 4.9 0.001 0.007 0.156 4.987 write_particle_coordinates 11 5.9 0.156 4.980 0.156 4.980 init_scf_run 11 5.9 0.000 0.001 3.908 3.908 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.908 3.908 mp_waitall_1 158411 16.6 2.811 3.743 2.811 3.743 mp_sum_dm 438 4.9 3.548 3.666 3.548 3.666 sum_up_and_integrate 119 10.3 0.001 0.002 3.647 3.653 integrate_v_rspace 119 11.3 0.002 0.003 3.634 3.642 update_particle_set 20 4.0 0.000 0.000 3.522 3.638 qs_ot_get_p 119 10.4 0.001 0.002 3.610 3.628 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.850 3.485 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.398 3.402 calculate_rho_elec 119 8.7 0.031 0.032 3.397 3.402 calculate_first_density_matrix 1 7.0 0.000 0.000 2.642 2.643 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.414 2.420 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.049 2.349 apply_single 119 13.6 0.000 0.000 2.049 2.349 init_scf_loop 11 6.9 0.000 0.000 2.313 2.314 multiply_cannon_metrocomm3 18288 15.5 0.047 0.048 1.413 2.175 make_m2s 4572 13.5 0.044 0.045 2.013 2.162 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.111 2.112 density_rs2pw 119 9.7 0.003 0.004 1.943 2.085 make_images 4572 14.5 0.192 0.204 1.927 2.076 calculate_dm_sparse 119 9.5 0.000 0.001 2.043 2.052 grid_integrate_task_list 119 12.3 1.801 1.887 1.801 1.887 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.876 1.877 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.835 1.849 cp_fm_diag_elpa_base 50 14.0 1.809 1.823 1.833 1.848 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.798 1.805 acc_transpose_blocks 18288 15.5 0.077 0.079 1.704 1.779 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.727 1.741 mp_max_l 33 2.8 1.648 1.703 1.648 1.703 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.685 1.691 prepare_preconditioner 11 7.9 0.000 0.000 1.537 1.539 make_preconditioner 11 8.9 0.000 0.000 1.537 1.539 ot_diis_step 108 11.5 0.011 0.011 1.488 1.488 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.417 1.473 potential_pw2rs 119 12.3 0.007 0.007 1.432 1.437 fft3d_ps 1201 13.6 0.536 0.556 1.413 1.428 mp_sum_l 7287 12.8 1.107 1.404 1.107 1.404 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.363 1.367 fft_wrap_pw1pw2_140 487 12.2 0.060 0.062 1.345 1.359 grid_collocate_task_list 119 9.7 1.233 1.308 1.233 1.308 transfer_rs2pw 487 10.6 0.005 0.005 1.145 1.280 multiply_cannon_sync_h2d 18288 15.5 1.034 1.218 1.034 1.218 wfi_extrapolate 11 7.9 0.001 0.003 1.214 1.214 make_images_data 4572 15.5 0.047 0.051 0.865 1.054 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.964 0.980 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.751 0.964 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.956 0.957 transfer_pw2rs 487 13.2 0.004 0.004 0.924 0.931 mp_alltoall_d11v 2130 13.8 0.770 0.912 0.770 0.912 acc_transpose_blocks_kernels 18288 16.5 0.219 0.226 0.863 0.872 acc_transpose_blocks_sync 54864 16.5 0.744 0.814 0.744 0.814 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.805 0.806 mp_alltoall_z22v 1201 15.6 0.720 0.788 0.720 0.788 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=38.541000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=493.727273, yerr=3.413633 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 557.371392E+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 3679 72932. MP_Allreduce 10304 342. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.087 38.322 38.326 qs_mol_dyn_low 1 2.0 0.004 0.009 37.901 37.909 qs_forces 11 3.9 0.009 0.021 37.808 37.809 qs_energies 11 4.9 0.002 0.005 36.074 36.079 scf_env_do_scf 11 5.9 0.002 0.011 29.237 29.238 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.315 26.315 dbcsr_multiply_generic 2286 12.5 0.102 0.105 21.712 21.802 qs_scf_new_mos 108 7.5 0.001 0.001 18.285 18.338 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.284 18.337 multiply_cannon 2286 13.5 0.216 0.223 17.841 18.224 velocity_verlet 10 3.0 0.002 0.005 18.139 18.143 multiply_cannon_loop 2286 14.5 1.562 1.623 16.882 17.352 ot_scf_mini 108 9.5 0.002 0.003 17.234 17.289 multiply_cannon_multrec 27432 15.5 2.482 3.154 10.244 10.662 ot_mini 108 10.5 0.001 0.001 10.569 10.632 dbcsr_mm_accdrv_process 47916 15.9 6.862 8.971 7.661 9.245 qs_ot_get_derivative 108 11.5 0.001 0.001 8.687 8.743 rebuild_ks_matrix 119 8.3 0.000 0.000 6.316 6.365 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 6.315 6.365 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.590 5.636 init_scf_run 11 5.9 0.000 0.001 5.376 5.377 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.376 5.377 calculate_first_density_matrix 1 7.0 0.000 0.002 3.959 3.963 qs_ot_get_p 119 10.4 0.001 0.002 3.787 3.867 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.101 3.563 sum_up_and_integrate 119 10.3 0.001 0.003 3.463 3.470 integrate_v_rspace 119 11.3 0.003 0.003 3.451 3.459 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.331 3.381 calculate_rho_elec 119 8.7 0.040 0.046 3.331 3.381 calculate_dm_sparse 119 9.5 0.000 0.000 2.990 3.046 mp_waitall_1 137007 16.6 2.168 2.905 2.168 2.905 init_scf_loop 11 6.9 0.001 0.003 2.891 2.892 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.203 2.672 apply_single 119 13.6 0.000 0.000 2.202 2.672 acc_transpose_blocks 27432 15.5 0.115 0.117 2.538 2.632 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.620 2.622 jit_kernel_multiply 8 15.9 0.732 2.500 0.732 2.500 make_m2s 4572 13.5 0.054 0.056 2.348 2.461 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.421 2.434 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.391 2.419 make_images 4572 14.5 0.272 0.331 2.241 2.352 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.324 2.332 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 2.073 2.086 prepare_preconditioner 11 7.9 0.000 0.000 2.050 2.058 make_preconditioner 11 8.9 0.000 0.001 2.050 2.058 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.783 1.990 cp_dbcsr_syevd 50 12.0 0.003 0.005 1.915 1.916 density_rs2pw 119 9.7 0.004 0.005 1.782 1.909 grid_integrate_task_list 119 12.3 1.816 1.898 1.816 1.898 ot_diis_step 108 11.5 0.012 0.012 1.839 1.840 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.737 1.769 mp_sum_l 7287 12.8 1.138 1.702 1.138 1.702 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.614 1.615 cp_fm_redistribute_end 50 14.0 0.808 1.589 0.814 1.591 acc_transpose_blocks_sync 82296 16.5 1.481 1.560 1.481 1.560 cp_fm_diag_elpa_base 50 14.0 0.750 1.515 0.774 1.556 multiply_cannon_metrocomm3 27432 15.5 0.039 0.041 0.909 1.465 fft_wrap_pw1pw2_140 487 12.2 0.071 0.073 1.421 1.455 fft3d_ps 1201 13.6 0.577 0.637 1.398 1.429 wfi_extrapolate 11 7.9 0.001 0.003 1.352 1.352 grid_collocate_task_list 119 9.7 1.249 1.342 1.249 1.342 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.315 1.332 potential_pw2rs 119 12.3 0.008 0.009 1.266 1.270 transfer_rs2pw 487 10.6 0.005 0.005 1.009 1.154 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.144 1.148 make_images_data 4572 15.5 0.047 0.051 0.886 1.022 mp_alltoall_d11v 2130 13.8 0.854 1.013 0.854 1.013 hybrid_alltoall_any 4725 16.4 0.066 0.155 0.800 0.983 acc_transpose_blocks_kernels 27432 16.5 0.273 0.280 0.912 0.927 dbcsr_complete_redistribute 329 12.2 0.112 0.140 0.711 0.878 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.800 0.877 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.814 0.817 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=38.326000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.090909, yerr=3.449757 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 617.975808E+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 62672. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3629 73907. 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 0.010 0.025 29.022 29.022 qs_mol_dyn_low 1 2.0 0.003 0.004 28.771 28.803 qs_forces 11 3.9 0.002 0.002 28.710 28.710 qs_energies 11 4.9 0.002 0.002 27.006 27.009 scf_env_do_scf 11 5.9 0.000 0.001 22.039 22.039 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.676 19.676 velocity_verlet 10 3.0 0.002 0.002 14.701 14.703 dbcsr_multiply_generic 2286 12.5 0.095 0.098 13.453 13.525 qs_scf_new_mos 108 7.5 0.001 0.001 12.114 12.137 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.113 12.137 ot_scf_mini 108 9.5 0.002 0.002 11.392 11.415 multiply_cannon 2286 13.5 0.225 0.233 10.794 11.250 multiply_cannon_loop 2286 14.5 0.646 0.660 9.866 10.014 ot_mini 108 10.5 0.001 0.001 6.539 6.567 multiply_cannon_multrec 9144 15.5 1.726 1.926 6.145 6.375 rebuild_ks_matrix 119 8.3 0.000 0.000 5.757 5.779 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.756 5.778 qs_ot_get_derivative 108 11.5 0.001 0.001 5.225 5.247 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.115 5.136 dbcsr_mm_accdrv_process 12550 15.8 3.715 4.299 4.311 4.392 init_scf_run 11 5.9 0.000 0.001 3.503 3.503 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.503 3.503 sum_up_and_integrate 119 10.3 0.001 0.001 3.403 3.407 integrate_v_rspace 119 11.3 0.003 0.003 3.393 3.397 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.280 3.288 calculate_rho_elec 119 8.7 0.060 0.061 3.280 3.287 qs_ot_get_p 119 10.4 0.001 0.001 2.951 2.993 init_scf_loop 11 6.9 0.000 0.000 2.339 2.340 calculate_first_density_matrix 1 7.0 0.000 0.000 2.314 2.315 mp_waitall_1 115863 16.7 1.573 2.075 1.573 2.075 make_m2s 4572 13.5 0.033 0.035 1.855 2.010 grid_integrate_task_list 119 12.3 1.874 1.928 1.874 1.928 make_images 4572 14.5 0.268 0.301 1.765 1.919 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.902 1.904 calculate_dm_sparse 119 9.5 0.000 0.000 1.868 1.890 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.808 1.816 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.791 1.813 density_rs2pw 119 9.7 0.003 0.003 1.695 1.757 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.657 1.658 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.654 1.655 prepare_preconditioner 11 7.9 0.000 0.000 1.603 1.607 make_preconditioner 11 8.9 0.000 0.000 1.603 1.607 acc_transpose_blocks 9144 15.5 0.041 0.042 1.461 1.541 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.491 1.518 fft_wrap_pw1pw2_140 487 12.2 0.094 0.096 1.474 1.484 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.465 1.477 fft3d_ps 1201 13.6 0.638 0.651 1.421 1.430 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.393 1.404 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.377 1.385 cp_fm_diag_elpa_base 50 14.0 1.350 1.366 1.375 1.383 grid_collocate_task_list 119 9.7 1.298 1.363 1.298 1.363 ot_diis_step 108 11.5 0.013 0.013 1.300 1.301 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.242 1.278 apply_single 119 13.6 0.000 0.000 1.242 1.278 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.261 1.269 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.231 1.232 potential_pw2rs 119 12.3 0.010 0.010 1.197 1.199 wfi_extrapolate 11 7.9 0.001 0.001 1.136 1.136 jit_kernel_multiply 6 15.6 0.556 1.131 0.556 1.131 make_images_data 4572 15.5 0.042 0.046 0.804 0.994 hybrid_alltoall_any 4725 16.4 0.065 0.178 0.753 0.969 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.873 0.923 transfer_rs2pw 487 10.6 0.004 0.005 0.834 0.896 mp_alltoall_d11v 2130 13.8 0.796 0.872 0.796 0.872 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.837 0.843 cp_fm_cholesky_invert 11 10.9 0.785 0.791 0.785 0.791 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.378 0.783 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.762 0.764 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.684 0.753 acc_transpose_blocks_sync 27432 16.5 0.721 0.738 0.721 0.738 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.680 0.728 mp_alltoall_z22v 1201 15.6 0.658 0.694 0.658 0.694 transfer_pw2rs 487 13.2 0.003 0.004 0.680 0.683 mp_allgather_i34 2286 14.5 0.242 0.668 0.242 0.668 jit_kernel_transpose 5 15.6 0.565 0.635 0.565 0.635 dbcsr_complete_redistribute 329 12.2 0.150 0.157 0.607 0.634 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.022000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=587.727273, yerr=4.350359 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 771.715072E+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 3629 74142. 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 0.012 0.032 42.757 42.757 qs_mol_dyn_low 1 2.0 0.004 0.015 42.513 42.520 qs_forces 11 3.9 0.014 0.017 40.658 40.661 qs_energies 11 4.9 0.002 0.002 38.662 38.666 scf_env_do_scf 11 5.9 0.000 0.001 32.718 32.718 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.842 26.843 velocity_verlet 10 3.0 0.002 0.002 23.734 23.964 dbcsr_multiply_generic 2286 12.5 0.118 0.121 19.491 19.615 qs_scf_new_mos 108 7.5 0.001 0.001 17.543 17.635 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.543 17.634 ot_scf_mini 108 9.5 0.002 0.002 16.374 16.473 multiply_cannon 2286 13.5 0.293 0.303 15.438 16.355 multiply_cannon_loop 2286 14.5 0.866 0.881 14.157 15.060 ot_mini 108 10.5 0.001 0.001 10.035 10.150 multiply_cannon_multrec 9144 15.5 3.523 4.856 9.019 9.083 qs_ot_get_derivative 108 11.5 0.001 0.001 7.946 8.043 rebuild_ks_matrix 119 8.3 0.000 0.000 7.071 7.213 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 7.070 7.212 dbcsr_mm_accdrv_process 12550 15.8 4.618 6.613 5.362 6.646 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.372 6.501 init_scf_loop 11 6.9 0.000 0.000 5.849 5.850 prepare_preconditioner 11 7.9 0.000 0.000 4.882 4.896 make_preconditioner 11 8.9 0.000 0.000 4.882 4.896 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.912 4.777 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.936 3.941 calculate_rho_elec 119 8.7 0.118 0.121 3.935 3.940 init_scf_run 11 5.9 0.000 0.001 3.895 3.895 scf_env_initial_rho_setup 11 6.9 0.002 0.002 3.895 3.895 sum_up_and_integrate 119 10.3 0.001 0.001 3.748 3.754 integrate_v_rspace 119 11.3 0.003 0.003 3.738 3.745 qs_ot_get_p 119 10.4 0.001 0.002 3.366 3.507 mp_waitall_1 94719 16.7 2.415 3.419 2.415 3.419 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.774 3.214 cp_fm_upper_to_full 72 14.2 2.077 2.953 2.077 2.953 make_m2s 4572 13.5 0.037 0.037 2.417 2.586 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.274 2.539 apply_single 119 13.6 0.000 0.000 2.274 2.539 make_images 4572 14.5 0.351 0.384 2.297 2.466 calculate_first_density_matrix 1 7.0 0.000 0.000 2.454 2.456 fft_wrap_pw1pw2 1201 11.6 0.014 0.014 2.405 2.413 calculate_dm_sparse 119 9.5 0.000 0.000 2.330 2.350 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.400 2.259 grid_integrate_task_list 119 12.3 2.092 2.147 2.092 2.147 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.062 2.111 ot_diis_step 108 11.5 0.014 0.014 2.059 2.059 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.058 2.059 dbcsr_complete_redistribute 329 12.2 0.263 0.267 1.529 2.054 fft_wrap_pw1pw2_140 487 12.2 0.179 0.180 2.035 2.042 density_rs2pw 119 9.7 0.003 0.003 1.974 2.001 qs_ot_p2m_diag 50 11.0 0.042 0.042 1.950 1.951 acc_transpose_blocks 9144 15.5 0.043 0.043 1.889 1.921 mp_sum_dm 438 4.9 1.629 1.873 1.629 1.873 fft3d_ps 1201 13.6 0.860 0.878 1.857 1.863 md_write_output 11 3.9 0.005 0.036 0.231 1.839 md_output 10 3.0 0.000 0.000 0.232 1.837 update_particle_set 20 4.0 0.000 0.000 1.606 1.836 mp_sum_l 7287 12.8 1.076 1.811 1.076 1.811 write_trajectory 44 4.9 0.005 0.039 0.226 1.803 qs_energies_init_hamiltonians 11 5.9 0.005 0.007 1.773 1.773 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.742 1.764 write_particle_coordinates 11 5.9 0.220 1.764 0.220 1.764 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.220 1.744 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.595 1.596 grid_collocate_task_list 119 9.7 1.529 1.571 1.529 1.571 mp_alltoall_i22 627 13.8 1.003 1.549 1.003 1.549 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 0.965 1.487 wfi_extrapolate 11 7.9 0.001 0.001 1.376 1.376 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.328 1.347 hybrid_alltoall_any 4725 16.4 0.091 0.153 1.076 1.342 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.317 1.325 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.294 1.294 cp_fm_diag_elpa_base 50 14.0 1.207 1.221 1.292 1.292 make_images_data 4572 15.5 0.045 0.048 1.056 1.278 potential_pw2rs 119 12.3 0.013 0.014 1.226 1.228 mp_alltoall_d11v 2130 13.8 1.166 1.210 1.166 1.210 acc_transpose_blocks_sync 27432 16.5 1.106 1.138 1.106 1.138 cp_fm_cholesky_invert 11 10.9 1.119 1.126 1.119 1.126 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.099 1.113 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.990 1.028 jit_kernel_multiply 5 15.3 0.715 0.981 0.715 0.981 qs_create_task_list 11 7.9 0.000 0.001 0.935 0.944 generate_qs_task_list 11 8.9 0.373 0.393 0.934 0.944 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.906 0.920 mp_alltoall_z22v 1201 15.6 0.862 0.888 0.862 0.888 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.757000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=727.545455, yerr=13.186520 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 496.553984E+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 3528 76079. MP_Allreduce 9920 483. MP_Sync 52 MP_Alltoall 1938 1555453. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.032 86.487 86.489 qs_mol_dyn_low 1 2.0 0.005 0.019 86.036 86.105 qs_forces 11 3.9 0.003 0.003 85.944 85.945 qs_energies 11 4.9 0.002 0.004 83.021 83.038 scf_env_do_scf 11 5.9 0.000 0.001 72.855 72.855 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 67.517 67.518 dbcsr_multiply_generic 2055 12.4 0.108 0.111 55.157 55.552 qs_scf_new_mos 99 7.5 0.000 0.001 50.583 50.744 qs_scf_loop_do_ot 99 8.5 0.000 0.001 50.583 50.744 ot_scf_mini 99 9.5 0.002 0.002 48.109 48.263 multiply_cannon 2055 13.4 0.175 0.180 45.248 46.838 multiply_cannon_loop 2055 14.4 1.831 1.892 44.298 45.841 velocity_verlet 10 3.0 0.002 0.006 43.766 43.767 ot_mini 99 10.5 0.001 0.001 29.074 29.211 qs_ot_get_derivative 99 11.5 0.001 0.002 22.102 22.242 multiply_cannon_multrec 49320 15.4 10.997 11.898 18.429 19.480 rebuild_ks_matrix 110 8.3 0.000 0.001 14.562 14.763 qs_ks_build_kohn_sham_matrix 110 9.3 0.017 0.062 14.562 14.763 mp_waitall_1 220248 16.4 12.605 14.135 12.605 14.135 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.740 12.917 qs_ot_get_p 110 10.4 0.001 0.002 10.242 10.390 multiply_cannon_sync_h2d 49320 15.4 9.278 9.882 9.278 9.882 multiply_cannon_metrocomm3 49320 15.4 0.080 0.087 7.394 8.649 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.028 8.591 init_scf_run 11 5.9 0.000 0.003 8.099 8.100 scf_env_initial_rho_setup 11 6.9 0.003 0.003 8.099 8.099 dbcsr_mm_accdrv_process 87628 16.1 3.292 3.498 7.304 7.829 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.259 7.802 apply_single 110 13.6 0.000 0.000 7.259 7.802 sum_up_and_integrate 110 10.3 0.002 0.004 6.921 6.959 integrate_v_rspace 110 11.3 0.002 0.003 6.895 6.940 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.735 6.770 ot_diis_step 99 11.5 0.005 0.006 6.654 6.655 mp_sum_l 6594 12.7 5.475 6.540 5.475 6.540 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.369 6.457 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.152 6.279 calculate_rho_elec 110 8.6 0.020 0.024 6.152 6.279 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.762 5.763 init_scf_loop 11 6.9 0.000 0.001 5.305 5.305 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.265 5.267 cp_fm_redistribute_end 48 14.0 2.675 5.240 2.684 5.244 cp_fm_diag_elpa_base 48 14.0 2.549 5.114 2.556 5.130 calculate_dm_sparse 110 9.5 0.001 0.001 4.534 4.634 multiply_cannon_metrocomm1 49320 15.4 0.063 0.066 3.340 4.302 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.257 4.259 jit_kernel_multiply 13 15.9 3.729 4.223 3.729 4.223 make_m2s 4110 13.4 0.061 0.066 3.980 4.070 calculate_first_density_matrix 1 7.0 0.000 0.000 4.007 4.011 wfi_extrapolate 11 7.9 0.003 0.017 3.988 3.988 make_images 4110 14.4 0.178 0.190 3.884 3.978 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.812 3.858 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.467 3.524 grid_integrate_task_list 110 12.3 3.237 3.381 3.237 3.381 density_rs2pw 110 9.6 0.004 0.004 3.152 3.306 prepare_preconditioner 11 7.9 0.000 0.000 3.115 3.138 make_preconditioner 11 8.9 0.000 0.000 3.115 3.138 acc_transpose_blocks 49320 15.4 0.211 0.225 2.743 2.898 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.845 2.884 fft_wrap_pw1pw2 1111 11.6 0.014 0.017 2.687 2.792 mp_alltoall_d11v 2046 13.8 2.294 2.656 2.294 2.656 potential_pw2rs 110 12.3 0.005 0.006 2.509 2.555 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.350 2.375 fft_wrap_pw1pw2_140 451 12.1 0.092 0.097 2.267 2.364 fft3d_ps 1111 13.6 0.732 0.815 2.227 2.326 grid_collocate_task_list 110 9.6 2.158 2.296 2.158 2.296 mp_sum_d 3893 11.9 1.575 2.109 1.575 2.109 transfer_rs2pw 451 10.6 0.006 0.006 1.904 2.084 mp_waitany 14300 13.8 1.874 2.058 1.874 2.058 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.960 1.984 make_images_data 4110 15.4 0.043 0.046 1.791 1.911 transfer_pw2rs 451 13.1 0.005 0.007 1.789 1.817 hybrid_alltoall_any 4261 16.3 0.084 0.483 1.557 1.816 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=86.489000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=472.272727, yerr=2.799055 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 589.316096E+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 3528 76294. MP_Allreduce 9919 558. MP_Sync 52 MP_Alltoall 1717 2955794. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.027 69.691 69.692 qs_mol_dyn_low 1 2.0 0.004 0.010 69.427 69.436 qs_forces 11 3.9 0.003 0.003 69.359 69.360 qs_energies 11 4.9 0.002 0.002 66.018 66.022 scf_env_do_scf 11 5.9 0.000 0.001 57.075 57.075 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.982 49.982 dbcsr_multiply_generic 2055 12.4 0.116 0.121 38.689 38.889 velocity_verlet 10 3.0 0.002 0.005 35.810 35.812 qs_scf_new_mos 99 7.5 0.001 0.001 34.080 34.243 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.080 34.242 multiply_cannon 2055 13.4 0.223 0.243 31.884 32.940 ot_scf_mini 99 9.5 0.003 0.003 32.386 32.544 multiply_cannon_loop 2055 14.4 1.159 1.180 30.555 31.450 ot_mini 99 10.5 0.001 0.001 18.786 18.936 multiply_cannon_multrec 24660 15.4 6.896 8.219 14.174 15.684 rebuild_ks_matrix 110 8.3 0.000 0.000 13.675 13.799 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.675 13.798 qs_ot_get_derivative 99 11.5 0.001 0.001 12.952 13.105 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.039 12.146 mp_waitall_1 176588 16.5 7.905 10.441 7.905 10.441 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 5.357 7.978 dbcsr_mm_accdrv_process 52282 16.1 5.554 6.510 7.111 7.420 multiply_cannon_sync_h2d 24660 15.4 6.334 7.321 6.334 7.321 qs_ot_get_p 110 10.4 0.001 0.002 6.899 7.057 init_scf_loop 11 6.9 0.000 0.000 7.051 7.051 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.486 7.014 apply_single 110 13.6 0.000 0.001 6.486 7.014 init_scf_run 11 5.9 0.000 0.001 6.489 6.489 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.488 6.489 sum_up_and_integrate 110 10.3 0.001 0.003 6.362 6.374 integrate_v_rspace 110 11.3 0.002 0.003 6.335 6.348 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.807 5.819 calculate_rho_elec 110 8.6 0.039 0.047 5.807 5.818 ot_diis_step 99 11.5 0.010 0.010 5.782 5.782 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.791 5.494 prepare_preconditioner 11 7.9 0.000 0.000 5.017 5.036 make_preconditioner 11 8.9 0.000 0.000 5.017 5.036 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.913 4.935 make_m2s 4110 13.4 0.057 0.059 4.295 4.763 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.623 4.724 make_images 4110 14.4 0.403 0.461 4.183 4.647 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.433 4.434 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.861 3.861 cp_fm_redistribute_end 48 14.0 1.945 3.827 1.952 3.830 cp_fm_diag_elpa_base 48 14.0 1.813 3.678 1.870 3.761 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.583 3.584 wfi_extrapolate 11 7.9 0.001 0.001 3.484 3.484 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.377 3.457 fft_wrap_pw1pw2 1111 11.6 0.014 0.017 3.218 3.360 grid_integrate_task_list 110 12.3 3.157 3.311 3.157 3.311 density_rs2pw 110 9.6 0.004 0.004 3.037 3.232 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.172 3.216 calculate_dm_sparse 110 9.5 0.001 0.001 3.046 3.076 calculate_first_density_matrix 1 7.0 0.000 0.000 2.907 2.909 fft_wrap_pw1pw2_140 451 12.1 0.110 0.116 2.681 2.818 make_images_data 4110 15.4 0.048 0.052 2.326 2.798 hybrid_alltoall_any 4261 16.3 0.105 0.452 2.048 2.790 mp_sum_l 6594 12.7 1.882 2.740 1.882 2.740 fft3d_ps 1111 13.6 1.071 1.273 2.598 2.729 cp_fm_cholesky_invert 11 10.9 2.475 2.490 2.475 2.490 grid_collocate_task_list 110 9.6 2.166 2.306 2.166 2.306 acc_transpose_blocks 24660 15.4 0.113 0.117 2.053 2.288 potential_pw2rs 110 12.3 0.008 0.008 2.222 2.234 mp_alltoall_d11v 2046 13.8 1.919 2.120 1.919 2.120 jit_kernel_multiply 10 16.3 1.191 2.092 1.191 2.092 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.999 2.019 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.952 1.953 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.799 1.815 transfer_rs2pw 451 10.6 0.007 0.007 1.551 1.688 mp_allgather_i34 2055 14.4 0.638 1.680 0.638 1.680 multiply_cannon_metrocomm4 22605 15.4 0.077 0.082 0.791 1.647 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.593 1.605 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.386 1.490 mp_waitany 10164 13.8 1.272 1.434 1.272 1.434 mp_irecv_dv 57340 16.2 0.661 1.422 0.661 1.422 dbcsr_complete_redistribute 325 12.2 0.261 0.321 1.236 1.422 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.692000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=558.090909, yerr=6.775015 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 666.955776E+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 3480 77281. 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 0.051 0.070 59.711 59.712 qs_mol_dyn_low 1 2.0 0.006 0.026 59.312 59.362 qs_forces 11 3.9 0.003 0.003 59.169 59.170 qs_energies 11 4.9 0.002 0.002 55.977 55.980 scf_env_do_scf 11 5.9 0.001 0.001 47.330 47.331 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.788 39.788 velocity_verlet 10 3.0 0.008 0.058 31.428 31.430 dbcsr_multiply_generic 2055 12.4 0.111 0.116 29.745 29.983 qs_scf_new_mos 99 7.5 0.001 0.001 25.345 25.448 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.345 25.448 multiply_cannon 2055 13.4 0.210 0.219 23.597 24.679 ot_scf_mini 99 9.5 0.002 0.002 24.093 24.217 multiply_cannon_loop 2055 14.4 0.811 0.869 22.437 23.412 ot_mini 99 10.5 0.001 0.001 13.879 14.002 rebuild_ks_matrix 110 8.3 0.000 0.000 12.161 12.296 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.161 12.296 multiply_cannon_multrec 16440 15.4 3.676 4.937 10.482 11.474 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.698 10.821 mp_waitall_1 139946 16.5 7.378 10.173 7.378 10.173 qs_ot_get_derivative 99 11.5 0.001 0.001 9.470 9.589 init_scf_loop 11 6.9 0.000 0.000 7.507 7.507 dbcsr_mm_accdrv_process 34862 16.1 6.020 6.831 6.651 7.250 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.537 7.193 init_scf_run 11 5.9 0.000 0.001 6.278 6.278 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.278 6.278 sum_up_and_integrate 110 10.3 0.001 0.002 6.211 6.225 integrate_v_rspace 110 11.3 0.003 0.003 6.185 6.198 prepare_preconditioner 11 7.9 0.000 0.000 5.760 5.774 make_preconditioner 11 8.9 0.000 0.000 5.760 5.774 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.525 5.536 calculate_rho_elec 110 8.6 0.058 0.058 5.524 5.535 qs_ot_get_p 110 10.4 0.001 0.001 5.383 5.514 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.211 5.454 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.823 5.363 apply_single 110 13.6 0.000 0.000 4.823 5.363 make_m2s 4110 13.4 0.049 0.051 4.183 4.550 make_images 4110 14.4 0.397 0.514 4.067 4.431 ot_diis_step 99 11.5 0.011 0.011 4.378 4.379 multiply_cannon_sync_h2d 16440 15.4 3.288 4.003 3.288 4.003 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.250 3.810 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.756 3.760 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.685 3.686 grid_integrate_task_list 110 12.3 3.225 3.408 3.225 3.408 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.384 3.385 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.303 3.353 calculate_first_density_matrix 1 7.0 0.000 0.000 3.296 3.297 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.077 3.086 density_rs2pw 110 9.6 0.004 0.004 2.768 2.958 wfi_extrapolate 11 7.9 0.001 0.001 2.892 2.892 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.830 2.837 cp_fm_diag_elpa_base 48 14.0 2.788 2.802 2.828 2.836 make_images_data 4110 15.4 0.045 0.049 2.328 2.831 hybrid_alltoall_any 4261 16.3 0.108 0.380 2.057 2.827 fft_wrap_pw1pw2_140 451 12.1 0.126 0.129 2.631 2.640 calculate_dm_sparse 110 9.5 0.001 0.001 2.540 2.568 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.460 2.528 multiply_cannon_metrocomm4 14385 15.4 0.048 0.053 0.872 2.438 cp_fm_cholesky_invert 11 10.9 2.401 2.414 2.401 2.414 fft3d_ps 1111 13.6 1.085 1.094 2.391 2.405 grid_collocate_task_list 110 9.6 2.217 2.401 2.217 2.401 mp_irecv_dv 48980 15.7 0.797 2.298 0.797 2.298 mp_alltoall_d11v 2046 13.8 1.835 2.087 1.835 2.087 potential_pw2rs 110 12.3 0.009 0.010 2.040 2.046 mp_sum_l 6594 12.7 1.337 1.995 1.337 1.995 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.948 1.949 dbcsr_complete_redistribute 325 12.2 0.319 0.372 1.435 1.741 acc_transpose_blocks 16440 15.4 0.075 0.077 1.581 1.700 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.660 1.673 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.480 1.495 mp_allgather_i34 2055 14.4 0.473 1.495 0.473 1.495 transfer_rs2pw 451 10.6 0.005 0.006 1.304 1.487 build_core_hamiltonian_matrix_ 11 4.9 0.002 0.015 1.364 1.479 cp_fm_cholesky_decompose 22 10.9 1.341 1.368 1.341 1.368 mp_waitany 17072 13.8 1.137 1.319 1.137 1.319 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.278 1.295 multiply_cannon_metrocomm1 16440 15.4 0.029 0.030 0.571 1.223 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.913 1.207 transfer_pw2rs 451 13.1 0.005 0.005 1.195 1.200 cp_fm_upper_to_full 70 14.2 0.949 1.197 0.949 1.197 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.712000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.727273, yerr=9.420393 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 735.604736E+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 3480 77279. 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 0.056 0.081 68.731 68.731 qs_mol_dyn_low 1 2.0 0.003 0.004 66.148 68.465 qs_forces 11 3.9 0.003 0.003 63.706 63.707 qs_energies 11 4.9 0.001 0.002 60.305 60.309 scf_env_do_scf 11 5.9 0.000 0.001 51.581 51.586 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.163 42.163 velocity_verlet 10 3.0 0.001 0.002 37.292 37.389 dbcsr_multiply_generic 2055 12.4 0.130 0.134 31.421 31.622 qs_scf_new_mos 99 7.5 0.001 0.001 27.709 27.809 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.708 27.808 ot_scf_mini 99 9.5 0.002 0.003 26.006 26.103 multiply_cannon 2055 13.4 0.236 0.254 24.276 24.992 multiply_cannon_loop 2055 14.4 1.431 1.475 22.843 23.437 ot_mini 99 10.5 0.001 0.001 15.007 15.123 multiply_cannon_multrec 24660 15.4 4.112 6.922 13.496 14.652 rebuild_ks_matrix 110 8.3 0.000 0.000 12.012 12.102 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.012 12.101 qs_ot_get_derivative 99 11.5 0.001 0.001 10.835 10.935 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.618 10.694 dbcsr_mm_accdrv_process 52304 16.0 8.041 9.502 9.226 10.126 init_scf_loop 11 6.9 0.000 0.000 9.377 9.378 prepare_preconditioner 11 7.9 0.000 0.000 7.630 7.643 make_preconditioner 11 8.9 0.000 0.000 7.630 7.643 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.486 7.308 sum_up_and_integrate 110 10.3 0.001 0.002 6.167 6.182 integrate_v_rspace 110 11.3 0.003 0.003 6.141 6.155 mp_waitall_1 121746 16.5 4.366 6.152 4.366 6.152 qs_ot_get_p 110 10.4 0.001 0.001 5.837 5.980 init_scf_run 11 5.9 0.000 0.001 5.977 5.978 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.977 5.977 make_m2s 4110 13.4 0.059 0.061 5.492 5.831 make_images 4110 14.4 0.573 0.694 5.351 5.686 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.638 5.650 calculate_rho_elec 110 8.6 0.077 0.081 5.638 5.649 md_output 10 3.0 0.000 0.000 0.201 4.737 md_write_output 11 3.9 0.019 0.435 0.198 4.708 write_trajectory 44 4.9 0.001 0.006 0.179 4.273 write_particle_coordinates 11 5.9 0.178 4.267 0.178 4.267 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.057 4.142 apply_single 110 13.6 0.000 0.000 4.056 4.141 ot_diis_step 99 11.5 0.011 0.011 4.133 4.133 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.936 3.952 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.629 3.681 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.417 3.473 grid_integrate_task_list 110 12.3 3.281 3.393 3.281 3.393 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.369 3.369 fft_wrap_pw1pw2 1111 11.6 0.013 0.017 3.230 3.253 multiply_cannon_metrocomm3 24660 15.4 0.037 0.039 1.411 3.210 calculate_dm_sparse 110 9.5 0.001 0.001 3.152 3.189 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.178 3.180 hybrid_alltoall_any 4261 16.3 0.121 0.459 2.366 3.102 make_images_data 4110 15.4 0.048 0.051 2.731 3.073 wfi_extrapolate 11 7.9 0.001 0.001 2.947 2.947 cp_fm_upper_to_full 70 14.2 2.075 2.938 2.075 2.938 calculate_first_density_matrix 1 7.0 0.000 0.000 2.933 2.936 density_rs2pw 110 9.6 0.004 0.004 2.761 2.929 dbcsr_complete_redistribute 325 12.2 0.385 0.449 2.113 2.821 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.782 2.821 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.808 2.819 cp_fm_diag_elpa_base 48 14.0 2.743 2.768 2.805 2.815 fft_wrap_pw1pw2_140 451 12.1 0.145 0.148 2.737 2.762 acc_transpose_blocks 24660 15.4 0.107 0.109 2.554 2.647 cp_fm_cholesky_invert 11 10.9 2.535 2.546 2.535 2.546 fft3d_ps 1111 13.6 1.126 1.165 2.501 2.520 multiply_cannon_sync_h2d 24660 15.4 2.352 2.509 2.352 2.509 grid_collocate_task_list 110 9.6 2.266 2.436 2.266 2.436 mp_max_l 33 2.8 2.317 2.417 2.317 2.417 mp_sum_dm 438 4.9 2.270 2.384 2.270 2.384 update_particle_set 20 4.0 0.000 0.000 2.229 2.326 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.254 2.255 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.564 2.250 mp_alltoall_d11v 2046 13.8 1.878 2.040 1.878 2.040 potential_pw2rs 110 12.3 0.012 0.013 1.964 1.971 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.830 1.864 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.135 1.807 mp_alltoall_i22 605 13.7 1.033 1.761 1.033 1.761 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.706 1.724 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.612 1.712 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.641 1.652 mp_sum_l 6594 12.7 0.953 1.566 0.953 1.566 acc_transpose_blocks_sync 73980 16.4 1.439 1.538 1.439 1.538 cp_fm_cholesky_decompose 22 10.9 1.368 1.425 1.368 1.425 multiply_cannon_metrocomm4 20550 15.4 0.058 0.061 0.834 1.424 transfer_rs2pw 451 10.6 0.005 0.006 1.207 1.387 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=68.731000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=698.545455, yerr=7.278407 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 864.133120E+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 258767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3480 77277. 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 0.055 0.077 55.438 55.439 qs_mol_dyn_low 1 2.0 0.003 0.006 55.071 55.098 qs_forces 11 3.9 0.003 0.003 54.961 54.962 qs_energies 11 4.9 0.002 0.003 51.309 51.315 scf_env_do_scf 11 5.9 0.000 0.001 42.302 42.302 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.117 35.118 velocity_verlet 10 3.0 0.004 0.005 30.490 30.496 dbcsr_multiply_generic 2055 12.4 0.142 0.177 24.152 24.351 qs_scf_new_mos 99 7.5 0.001 0.001 20.808 20.851 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.808 20.850 multiply_cannon 2055 13.4 0.239 0.250 18.447 19.731 ot_scf_mini 99 9.5 0.002 0.002 19.543 19.575 multiply_cannon_loop 2055 14.4 0.605 0.619 17.151 17.478 rebuild_ks_matrix 110 8.3 0.000 0.000 11.587 11.619 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.587 11.619 ot_mini 99 10.5 0.001 0.001 10.739 10.763 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.298 10.326 multiply_cannon_multrec 8220 15.4 3.155 4.417 8.265 9.311 mp_waitall_1 103326 16.6 5.974 7.677 5.974 7.677 init_scf_loop 11 6.9 0.000 0.000 7.135 7.136 qs_ot_get_derivative 99 11.5 0.001 0.001 7.039 7.071 sum_up_and_integrate 110 10.3 0.001 0.002 6.116 6.127 integrate_v_rspace 110 11.3 0.003 0.003 6.089 6.101 dbcsr_mm_accdrv_process 17442 15.9 3.188 4.250 4.971 5.940 init_scf_run 11 5.9 0.000 0.001 5.905 5.905 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.904 5.905 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.745 5.761 calculate_rho_elec 110 8.6 0.115 0.116 5.745 5.761 prepare_preconditioner 11 7.9 0.000 0.000 5.495 5.501 make_preconditioner 11 8.9 0.000 0.000 5.495 5.501 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.094 5.170 qs_ot_get_p 110 10.4 0.001 0.001 4.819 4.836 make_m2s 4110 13.4 0.037 0.039 4.309 4.616 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 2.953 4.490 make_images 4110 14.4 0.643 0.705 4.179 4.488 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.726 3.774 apply_single 110 13.6 0.000 0.000 3.726 3.774 ot_diis_step 99 11.5 0.012 0.012 3.676 3.677 grid_integrate_task_list 110 12.3 3.390 3.506 3.390 3.506 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.411 3.415 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.413 3.414 fft_wrap_pw1pw2 1111 11.6 0.014 0.015 3.395 3.401 calculate_first_density_matrix 1 7.0 0.000 0.000 3.232 3.233 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.067 3.067 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.931 2.954 density_rs2pw 110 9.6 0.003 0.004 2.761 2.933 fft_wrap_pw1pw2_140 451 12.1 0.192 0.195 2.906 2.918 hybrid_alltoall_any 4261 16.3 0.200 0.851 2.246 2.700 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.679 2.680 make_images_data 4110 15.4 0.041 0.047 2.300 2.678 cp_fm_cholesky_invert 11 10.9 2.634 2.643 2.634 2.643 calculate_dm_sparse 110 9.5 0.001 0.001 2.585 2.626 wfi_extrapolate 11 7.9 0.001 0.001 2.579 2.579 fft3d_ps 1111 13.6 1.269 1.274 2.555 2.568 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.520 2.529 cp_fm_diag_elpa_base 48 14.0 2.461 2.487 2.519 2.527 grid_collocate_task_list 110 9.6 2.369 2.501 2.369 2.501 multiply_cannon_sync_h2d 8220 15.4 2.373 2.494 2.373 2.494 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.040 2.068 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.782 1.987 jit_kernel_multiply 8 15.4 1.468 1.986 1.468 1.986 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.957 1.974 potential_pw2rs 110 12.3 0.015 0.015 1.896 1.900 mp_alltoall_d11v 2046 13.8 1.682 1.816 1.682 1.816 dbcsr_complete_redistribute 325 12.2 0.538 0.633 1.558 1.656 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.618 1.622 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.508 1.619 mp_allgather_i34 2055 14.4 0.461 1.570 0.461 1.570 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.530 1.539 cp_fm_cholesky_decompose 22 10.9 1.465 1.484 1.465 1.484 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.859 1.452 acc_transpose_blocks 8220 15.4 0.039 0.040 1.418 1.447 qs_create_task_list 11 7.9 0.000 0.001 1.234 1.334 generate_qs_task_list 11 8.9 0.395 0.465 1.234 1.334 transfer_rs2pw 451 10.6 0.005 0.005 1.153 1.326 mp_waitany 9240 13.8 1.056 1.226 1.056 1.226 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.465 1.200 copy_dbcsr_to_fm 151 11.3 0.002 0.003 1.155 1.169 mp_irecv_dv 24056 15.7 0.439 1.153 0.439 1.153 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.439000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=810.000000, yerr=15.960178 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 1.362940E+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 3469 77990. 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 0.033 0.078 79.771 79.772 qs_mol_dyn_low 1 2.0 0.003 0.003 79.371 79.421 qs_forces 11 3.9 0.003 0.004 79.290 79.291 qs_energies 11 4.9 0.002 0.002 75.184 75.184 scf_env_do_scf 11 5.9 0.000 0.001 64.775 64.775 velocity_verlet 10 3.0 0.002 0.003 48.760 48.767 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.969 43.970 dbcsr_multiply_generic 2055 12.4 0.122 0.126 30.039 30.132 qs_scf_new_mos 99 7.5 0.001 0.001 26.727 26.767 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.726 26.766 ot_scf_mini 99 9.5 0.002 0.002 24.922 24.946 multiply_cannon 2055 13.4 0.329 0.345 22.845 23.652 multiply_cannon_loop 2055 14.4 0.821 0.836 21.047 21.335 init_scf_loop 11 6.9 0.000 0.000 20.728 20.731 prepare_preconditioner 11 7.9 0.000 0.000 18.624 18.630 make_preconditioner 11 8.9 0.000 0.000 18.624 18.630 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.779 18.112 ot_mini 99 10.5 0.001 0.001 14.063 14.090 rebuild_ks_matrix 110 8.3 0.000 0.001 13.700 13.735 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.700 13.734 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.435 12.466 cp_fm_upper_to_full 70 14.2 8.269 11.718 8.269 11.718 multiply_cannon_multrec 8220 15.4 4.073 4.259 10.143 10.320 qs_ot_get_derivative 99 11.5 0.001 0.001 9.473 9.502 mp_waitall_1 84994 16.7 7.625 8.511 7.625 8.511 dbcsr_complete_redistribute 325 12.2 0.944 0.957 5.511 7.500 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.046 7.081 calculate_rho_elec 110 8.6 0.227 0.227 7.045 7.080 sum_up_and_integrate 110 10.3 0.002 0.002 6.713 6.730 integrate_v_rspace 110 11.3 0.004 0.004 6.685 6.702 copy_fm_to_dbcsr 174 11.2 0.001 0.001 4.431 6.414 dbcsr_mm_accdrv_process 11614 15.7 3.882 4.141 5.921 6.202 init_scf_run 11 5.9 0.000 0.001 6.139 6.139 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.139 6.139 make_m2s 4110 13.4 0.043 0.043 5.463 5.949 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 3.829 5.790 make_images 4110 14.4 0.883 0.940 5.273 5.758 qs_ot_get_p 110 10.4 0.001 0.001 5.535 5.569 mp_alltoall_i22 605 13.7 3.438 5.440 3.438 5.440 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.863 5.314 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.731 5.195 apply_single 110 13.6 0.000 0.000 4.731 5.194 fft_wrap_pw1pw2 1111 11.6 0.016 0.016 4.664 4.667 ot_diis_step 99 11.5 0.015 0.015 4.572 4.573 fft_wrap_pw1pw2_140 451 12.1 0.364 0.365 4.041 4.046 cp_fm_cholesky_invert 11 10.9 3.935 3.945 3.935 3.945 qs_ot_p2m_diag 48 11.0 0.150 0.155 3.919 3.926 grid_integrate_task_list 110 12.3 3.674 3.745 3.674 3.745 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.667 3.668 hybrid_alltoall_any 4261 16.3 0.262 0.563 2.863 3.629 make_images_data 4110 15.4 0.044 0.047 2.872 3.529 fft3d_ps 1111 13.6 1.827 1.839 3.523 3.528 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.060 3.517 density_rs2pw 110 9.6 0.004 0.004 3.496 3.504 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.398 3.399 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.346 3.346 wfi_extrapolate 11 7.9 0.001 0.001 3.311 3.311 calculate_dm_sparse 110 9.5 0.001 0.001 3.213 3.233 multiply_cannon_sync_h2d 8220 15.4 3.132 3.151 3.132 3.151 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.804 2.818 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.783 2.783 cp_fm_diag_elpa_base 48 14.0 2.481 2.558 2.781 2.781 calculate_first_density_matrix 1 7.0 0.000 0.000 2.715 2.715 grid_collocate_task_list 110 9.6 2.665 2.677 2.665 2.677 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.647 2.662 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.183 2.235 potential_pw2rs 110 12.3 0.021 0.021 2.230 2.233 mp_alltoall_d11v 2046 13.8 2.172 2.223 2.172 2.223 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.091 2.194 jit_kernel_multiply 10 15.2 1.844 2.175 1.844 2.175 qs_create_task_list 11 7.9 0.001 0.001 1.895 1.941 generate_qs_task_list 11 8.9 0.740 0.793 1.895 1.940 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.867 1.871 cp_fm_cholesky_decompose 22 10.9 1.806 1.833 1.806 1.833 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.808 1.827 acc_transpose_blocks 8220 15.4 0.040 0.041 1.724 1.766 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.721 1.737 qs_ot_get_orbitals 99 10.5 0.000 0.000 1.641 1.646 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=79.772000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1230.272727, yerr=52.810279 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 633.536512E+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 4027 66905. MP_Allreduce 11129 795. MP_Sync 87 MP_Alltoall 2226 2529087. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 205.897 205.898 qs_mol_dyn_low 1 2.0 0.003 0.004 205.470 205.483 qs_forces 11 3.9 0.005 0.005 205.374 205.376 qs_energies 11 4.9 0.002 0.003 199.717 199.743 scf_env_do_scf 11 5.9 0.001 0.001 182.677 182.694 scf_env_do_scf_inner_loop 117 6.6 0.003 0.007 163.752 163.754 dbcsr_multiply_generic 2507 12.6 0.179 0.183 126.285 126.971 qs_scf_new_mos 117 7.6 0.001 0.001 124.314 124.616 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.313 124.616 velocity_verlet 10 3.0 0.003 0.017 122.489 122.490 ot_scf_mini 117 9.6 0.003 0.003 117.610 117.910 multiply_cannon 2507 13.6 0.239 0.249 102.192 104.607 multiply_cannon_loop 2507 14.6 2.396 2.459 99.974 102.465 ot_mini 117 10.6 0.001 0.001 66.164 66.445 multiply_cannon_multrec 60168 15.6 31.698 33.660 41.842 43.857 qs_ot_get_derivative 117 11.6 0.001 0.001 41.405 41.699 rebuild_ks_matrix 128 8.3 0.001 0.001 33.864 34.380 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.016 33.863 34.379 mp_waitall_1 267128 16.5 29.229 32.422 29.229 32.422 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.454 30.919 qs_ot_get_p 128 10.4 0.001 0.001 29.161 29.644 multiply_cannon_sync_h2d 60168 15.6 26.411 27.938 26.411 27.938 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.260 25.130 apply_single 128 13.6 0.001 0.001 24.260 25.129 ot_diis_step 117 11.6 0.007 0.008 24.502 24.503 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.473 22.562 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.727 19.964 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.481 19.482 init_scf_loop 11 6.9 0.000 0.001 18.837 18.839 multiply_cannon_metrocomm3 60168 15.6 0.118 0.125 16.091 17.913 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.390 16.419 cp_fm_diag_elpa_base 83 14.4 16.336 16.374 16.384 16.414 make_m2s 5014 13.6 0.105 0.116 13.858 14.285 prepare_preconditioner 11 7.9 0.000 0.000 14.175 14.224 make_preconditioner 11 8.9 0.000 0.000 14.175 14.224 make_images 5014 14.6 0.400 0.418 13.673 14.110 sum_up_and_integrate 128 10.3 0.002 0.004 13.882 13.896 integrate_v_rspace 128 11.3 0.003 0.004 13.823 13.840 make_full_inverse_cholesky 11 9.9 0.000 0.000 13.445 13.612 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.849 12.981 calculate_rho_elec 128 8.7 0.045 0.064 12.848 12.981 init_scf_run 11 5.9 0.000 0.001 12.958 12.959 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.958 12.958 mp_sum_l 7950 12.9 8.977 10.732 8.977 10.732 dbcsr_mm_accdrv_process 124484 16.2 4.833 4.958 9.716 10.256 multiply_cannon_metrocomm1 60168 15.6 0.093 0.096 6.510 9.178 wfi_extrapolate 11 7.9 0.001 0.001 8.982 8.982 calculate_dm_sparse 128 9.5 0.001 0.001 8.581 8.685 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.247 8.385 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.199 8.264 cp_fm_cholesky_invert 11 10.9 8.079 8.097 8.079 8.097 make_images_data 5014 15.6 0.065 0.072 6.807 7.802 grid_integrate_task_list 128 12.3 7.064 7.471 7.064 7.471 hybrid_alltoall_any 5200 16.5 0.295 2.276 5.963 7.147 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.052 7.064 density_rs2pw 128 9.7 0.005 0.006 6.430 6.950 fft_wrap_pw1pw2 1291 11.7 0.018 0.025 6.520 6.794 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.085 6.231 mp_alltoall_d11v 2415 14.1 4.568 5.896 4.568 5.896 fft_wrap_pw1pw2_140 523 12.2 0.233 0.247 5.625 5.820 fft3d_ps 1291 13.7 2.098 2.604 5.345 5.564 grid_collocate_task_list 128 9.7 4.836 5.246 4.836 5.246 potential_pw2rs 128 12.3 0.009 0.010 4.375 4.399 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.898000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.636364, yerr=6.759138 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 848.973824E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453843137296 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4004 67737. MP_Allreduce 11071 960. MP_Sync 86 MP_Alltoall 1955 5113917. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.046 183.314 183.316 qs_mol_dyn_low 1 2.0 0.020 0.039 182.809 182.837 qs_forces 11 3.9 0.008 0.039 182.549 182.552 qs_energies 11 4.9 0.005 0.020 175.723 175.733 scf_env_do_scf 11 5.9 0.001 0.002 158.500 158.517 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 131.930 131.932 velocity_verlet 10 3.0 0.071 0.131 112.042 112.046 dbcsr_multiply_generic 2485 12.5 0.186 0.192 96.704 97.871 qs_scf_new_mos 116 7.6 0.001 0.001 94.028 94.556 qs_scf_loop_do_ot 116 8.6 0.001 0.001 94.028 94.555 ot_scf_mini 116 9.6 0.004 0.005 89.254 89.810 multiply_cannon 2485 13.5 0.501 0.568 76.676 81.038 multiply_cannon_loop 2485 14.5 1.549 1.614 73.696 76.602 ot_mini 116 10.6 0.002 0.009 49.886 50.475 mp_waitall_1 212858 16.6 23.969 37.816 23.969 37.816 multiply_cannon_multrec 29820 15.5 20.715 24.988 31.627 36.343 rebuild_ks_matrix 127 8.3 0.001 0.005 32.246 32.922 qs_ks_build_kohn_sham_matrix 127 9.3 0.020 0.045 32.244 32.922 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.005 29.648 multiply_cannon_metrocomm3 29820 15.5 0.093 0.098 15.618 28.661 qs_ot_get_derivative 116 11.6 0.001 0.002 28.077 28.625 init_scf_loop 11 6.9 0.003 0.024 26.473 26.474 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.945 23.004 apply_single 127 13.6 0.001 0.001 21.945 23.004 qs_ot_get_p 127 10.4 0.003 0.014 21.714 22.386 prepare_preconditioner 11 7.9 0.000 0.000 22.006 22.072 make_preconditioner 11 8.9 0.000 0.001 22.006 22.072 ot_diis_step 116 11.6 0.014 0.015 21.636 21.639 make_full_inverse_cholesky 11 9.9 0.000 0.001 20.867 21.270 multiply_cannon_sync_h2d 29820 15.5 17.779 19.775 17.779 19.775 qs_ot_p2m_diag 82 11.4 0.186 0.214 16.836 16.873 cp_dbcsr_syevd 82 12.4 0.008 0.023 15.622 15.624 make_m2s 4970 13.5 0.086 0.091 14.064 15.318 make_images 4970 14.5 1.145 1.342 13.850 15.105 sum_up_and_integrate 127 10.3 0.002 0.005 13.779 13.809 integrate_v_rspace 127 11.3 0.003 0.004 13.719 13.751 qs_rho_update_rho_low 127 7.7 0.002 0.015 12.872 12.903 calculate_rho_elec 127 8.7 0.088 0.113 12.870 12.903 init_scf_run 11 5.9 0.000 0.001 12.294 12.295 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.293 12.295 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.179 12.215 cp_fm_diag_elpa_base 82 14.4 11.991 12.048 12.161 12.192 cp_fm_cholesky_invert 11 10.9 12.118 12.154 12.118 12.154 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.061 11.469 dbcsr_mm_accdrv_process 61726 16.2 5.557 6.700 10.364 10.944 multiply_cannon_metrocomm4 27335 15.5 0.105 0.119 3.864 10.596 mp_irecv_dv 68888 16.3 3.656 10.197 3.656 10.197 make_images_data 4970 15.5 0.064 0.075 8.242 9.792 hybrid_alltoall_any 5155 16.4 0.345 1.539 6.964 9.316 wfi_extrapolate 11 7.9 0.004 0.020 8.238 8.238 fft_wrap_pw1pw2 1281 11.7 0.019 0.025 7.435 7.515 grid_integrate_task_list 127 12.3 7.139 7.480 7.139 7.480 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.265 7.044 density_rs2pw 127 9.7 0.006 0.008 6.591 7.011 calculate_dm_sparse 127 9.5 0.001 0.002 6.519 6.650 fft_wrap_pw1pw2_140 519 12.2 0.246 0.258 6.508 6.590 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.506 6.517 mp_sum_l 7884 12.9 4.192 6.306 4.192 6.306 fft3d_ps 1281 13.7 2.744 2.937 5.897 5.959 cp_fm_cholesky_decompose 22 10.9 5.721 5.800 5.721 5.800 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.518 5.682 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.360 5.437 grid_collocate_task_list 127 9.7 4.961 5.348 4.961 5.348 mp_alltoall_d11v 2401 14.1 4.499 5.133 4.499 5.133 mp_allgather_i34 2485 14.5 1.501 4.486 1.501 4.486 potential_pw2rs 127 12.3 0.019 0.059 4.321 4.337 mp_sum_d 4453 12.1 2.752 4.085 2.752 4.085 dbcsr_complete_redistribute 393 12.7 0.717 0.869 3.340 4.042 calculate_first_density_matrix 1 7.0 0.001 0.003 3.841 3.846 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.671 3.719 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=183.316000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=809.636364, yerr=1.149919 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 2.951022E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 4017216 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.3 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.034076E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1052064 MPI messages size (bytes): total size 2.737208E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.601750E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 284544 37295751168 131072 < size <= 4194304 665856 1004519030784 4194304 < size <= 16777216 66080 937890532432 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4050 66975. MP_Allreduce 11184 996. MP_Sync 88 MP_Alltoall 1724 9394234. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.037 170.536 170.537 qs_mol_dyn_low 1 2.0 0.005 0.012 170.018 170.056 qs_forces 11 3.9 0.004 0.004 169.579 169.582 qs_energies 11 4.9 0.002 0.006 163.042 163.055 scf_env_do_scf 11 5.9 0.001 0.002 147.414 147.415 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 118.513 118.514 velocity_verlet 10 3.0 0.018 0.079 107.673 107.685 dbcsr_multiply_generic 2529 12.6 0.186 0.191 83.234 84.295 qs_scf_new_mos 118 7.6 0.001 0.001 81.376 81.731 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.375 81.730 ot_scf_mini 118 9.6 0.004 0.004 77.106 77.511 multiply_cannon 2529 13.6 0.513 0.545 63.480 68.342 multiply_cannon_loop 2529 14.6 1.133 1.197 60.287 63.206 ot_mini 118 10.6 0.001 0.001 43.080 43.487 mp_waitall_1 172006 16.6 25.478 35.497 25.478 35.497 rebuild_ks_matrix 129 8.3 0.001 0.001 30.775 31.279 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.017 30.775 31.279 init_scf_loop 11 6.9 0.001 0.003 28.798 28.799 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.811 28.278 multiply_cannon_metrocomm3 20232 15.6 0.060 0.064 16.054 25.606 multiply_cannon_multrec 20232 15.6 12.957 15.936 22.605 25.592 prepare_preconditioner 11 7.9 0.000 0.000 24.728 24.774 make_preconditioner 11 8.9 0.000 0.001 24.728 24.773 make_full_inverse_cholesky 11 9.9 0.000 0.000 22.776 23.753 qs_ot_get_derivative 118 11.6 0.001 0.002 23.034 23.460 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.067 21.226 apply_single 129 13.6 0.001 0.001 20.067 21.226 ot_diis_step 118 11.6 0.018 0.018 19.933 19.933 qs_ot_get_p 129 10.4 0.001 0.002 19.121 19.561 make_m2s 5058 13.6 0.079 0.084 14.762 16.094 make_images 5058 14.6 1.142 1.256 14.524 15.854 multiply_cannon_sync_h2d 20232 15.6 13.658 15.113 13.658 15.113 qs_ot_p2m_diag 84 11.4 0.268 0.275 14.855 14.868 sum_up_and_integrate 129 10.3 0.002 0.003 13.837 13.865 integrate_v_rspace 129 11.3 0.003 0.003 13.777 13.805 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.752 13.752 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.085 13.121 calculate_rho_elec 129 8.7 0.133 0.147 13.085 13.120 cp_fm_cholesky_invert 11 10.9 11.830 11.849 11.830 11.849 make_images_data 5058 15.6 0.062 0.071 8.856 10.648 init_scf_run 11 5.9 0.000 0.001 10.644 10.644 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.644 10.644 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.365 10.385 cp_fm_diag_elpa_base 84 14.4 10.111 10.202 10.361 10.381 hybrid_alltoall_any 5245 16.5 0.456 2.063 7.719 9.728 multiply_cannon_metrocomm4 17703 15.6 0.066 0.076 3.593 9.553 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.166 9.469 mp_irecv_dv 50659 16.2 3.461 9.285 3.461 9.285 dbcsr_mm_accdrv_process 41846 16.2 5.918 6.277 9.105 9.257 grid_integrate_task_list 129 12.3 7.338 7.907 7.338 7.907 fft_wrap_pw1pw2 1301 11.7 0.019 0.024 7.631 7.738 wfi_extrapolate 11 7.9 0.001 0.001 7.357 7.357 fft_wrap_pw1pw2_140 527 12.2 0.286 0.295 6.709 6.824 density_rs2pw 129 9.7 0.006 0.008 6.491 6.812 cp_fm_cholesky_decompose 22 10.9 5.967 6.038 5.967 6.038 fft3d_ps 1301 13.7 2.788 3.050 5.925 6.003 calculate_dm_sparse 129 9.5 0.001 0.001 5.884 5.996 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.753 5.759 grid_collocate_task_list 129 9.7 5.188 5.636 5.188 5.636 dbcsr_complete_redistribute 397 12.7 1.037 1.108 4.341 5.548 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.647 5.246 mp_alltoall_d11v 2429 14.1 4.579 5.080 4.579 5.080 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.841 4.990 cp_fm_upper_to_full 106 14.8 3.937 4.984 3.937 4.984 mp_allgather_i34 2529 14.6 1.671 4.896 1.671 4.896 mp_sum_l 8016 12.9 3.166 4.751 3.166 4.751 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.139 4.333 potential_pw2rs 129 12.3 0.020 0.022 4.204 4.212 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.004 4.038 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.727 3.730 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.480 3.525 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=170.537000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=981.545455, yerr=5.581848 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410020794368 0.0% 0.0% 100.0% flops 32 x 32 x 32 1947767668736 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957875093504 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544039424 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714619674624 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 11444697722880 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019193401344 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019193401344 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624869085184 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.820223E+12 0.0% 0.0% 100.0% flops max/rank 4.321447E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705860672 0.0% 0.0% 100.0% number of processed stacks 5940480 0.0% 0.0% 100.0% average stack size 0.0 0.0 1128.8 marketing flops 143.532136E+12 ------------------------------------------------------------------------------- # multiplications 2496 max memory usage/rank 1.109250E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1138176 MPI messages size (bytes): total size 2.008880E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.765000E+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 318680 36009410560 131072 < size <= 4194304 711784 779363287040 4194304 < size <= 16777216 69840 660838110960 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3958 68502. MP_Allreduce 10955 1087. MP_Sync 86 MP_Alltoall 1700 12496250. 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 0.015 0.034 177.909 177.910 qs_mol_dyn_low 1 2.0 0.007 0.038 177.370 177.404 qs_forces 11 3.9 0.005 0.006 177.125 177.131 qs_energies 11 4.9 0.010 0.073 170.028 170.039 scf_env_do_scf 11 5.9 0.001 0.003 152.757 152.772 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 115.583 115.584 velocity_verlet 10 3.0 0.006 0.039 115.152 115.161 dbcsr_multiply_generic 2496 12.6 0.190 0.195 80.839 81.631 qs_scf_new_mos 116 7.6 0.001 0.001 79.993 80.305 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.992 80.305 ot_scf_mini 116 9.6 0.003 0.004 75.456 75.793 multiply_cannon 2496 13.6 0.548 0.576 55.833 58.798 multiply_cannon_loop 2496 14.6 1.852 1.922 52.266 54.058 ot_mini 116 10.6 0.001 0.001 42.907 43.210 init_scf_loop 11 6.9 0.001 0.003 37.048 37.049 prepare_preconditioner 11 7.9 0.000 0.000 32.903 32.947 make_preconditioner 11 8.9 0.000 0.001 32.903 32.947 make_full_inverse_cholesky 11 9.9 0.017 0.023 28.141 31.709 multiply_cannon_multrec 29952 15.6 13.529 18.553 26.733 31.562 rebuild_ks_matrix 127 8.3 0.001 0.001 29.258 29.556 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.021 29.257 29.556 mp_waitall_1 147208 16.7 17.389 27.038 17.389 27.038 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.379 26.648 qs_ot_get_derivative 116 11.6 0.001 0.002 23.297 23.629 make_m2s 4992 13.6 0.092 0.098 20.629 21.630 make_images 4992 14.6 1.979 2.235 20.323 21.327 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 19.091 19.642 apply_single 127 13.6 0.001 0.001 19.090 19.641 ot_diis_step 116 11.6 0.017 0.018 19.487 19.488 qs_ot_get_p 127 10.4 0.001 0.001 18.749 19.112 multiply_cannon_metrocomm3 29952 15.6 0.048 0.051 6.306 14.639 qs_ot_p2m_diag 82 11.4 0.338 0.384 14.501 14.552 sum_up_and_integrate 127 10.3 0.002 0.004 13.656 13.687 integrate_v_rspace 127 11.3 0.003 0.004 13.597 13.628 dbcsr_mm_accdrv_process 61880 16.2 8.427 9.239 12.774 13.294 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.144 13.179 calculate_rho_elec 127 8.7 0.173 0.188 13.144 13.179 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.109 13.110 cp_fm_upper_to_full 104 14.8 8.804 12.488 8.804 12.488 make_images_data 4992 15.6 0.064 0.069 11.094 12.486 cp_fm_cholesky_invert 11 10.9 12.048 12.071 12.048 12.071 hybrid_alltoall_any 5177 16.4 0.525 2.190 9.976 11.652 multiply_cannon_sync_h2d 29952 15.6 10.266 11.272 10.266 11.272 init_scf_run 11 5.9 0.000 0.001 11.105 11.106 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.105 11.106 dbcsr_complete_redistribute 393 12.7 1.440 1.549 7.615 10.465 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.774 9.997 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.882 9.904 cp_fm_diag_elpa_base 82 14.4 9.392 9.527 9.875 9.890 copy_fm_to_dbcsr 208 11.6 0.001 0.002 6.225 9.036 grid_integrate_task_list 127 12.3 7.467 7.871 7.467 7.871 fft_wrap_pw1pw2 1281 11.7 0.018 0.024 7.765 7.820 wfi_extrapolate 11 7.9 0.003 0.020 7.635 7.635 transfer_fm_to_dbcsr 11 9.9 0.001 0.005 4.739 7.467 multiply_cannon_metrocomm4 24960 15.6 0.078 0.086 2.748 7.129 mp_alltoall_i22 712 14.1 4.125 6.956 4.125 6.956 fft_wrap_pw1pw2_140 519 12.2 0.322 0.331 6.847 6.931 mp_irecv_dv 75698 16.2 2.596 6.855 2.596 6.855 density_rs2pw 127 9.7 0.005 0.007 6.234 6.587 calculate_dm_sparse 127 9.5 0.001 0.001 6.304 6.394 cp_fm_cholesky_decompose 22 10.9 6.053 6.101 6.053 6.101 fft3d_ps 1281 13.7 2.924 3.013 5.971 6.021 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.909 5.949 grid_collocate_task_list 127 9.7 5.246 5.565 5.246 5.565 mp_alltoall_d11v 2401 14.1 4.968 5.431 4.968 5.431 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.704 4.828 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.532 4.648 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 4.526 4.529 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.194 4.271 potential_pw2rs 127 12.3 0.021 0.022 3.982 3.995 dbcsr_special_finalize 12480 15.6 0.066 0.069 3.500 3.652 mp_allgather_i34 2496 14.6 1.661 3.612 1.661 3.612 dbcsr_merge_single_wm 12480 16.6 0.324 0.340 3.410 3.558 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=177.910000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1045.727273, yerr=12.483377 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.446670E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 759871. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3996 67859. MP_Allreduce 11050 1168. MP_Sync 86 MP_Alltoall 1700 18828174. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.073 165.326 165.327 qs_mol_dyn_low 1 2.0 0.003 0.004 164.705 164.753 qs_forces 11 3.9 0.004 0.004 164.583 164.586 qs_energies 11 4.9 0.002 0.002 157.183 157.188 scf_env_do_scf 11 5.9 0.001 0.001 139.531 139.543 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 108.573 108.574 velocity_verlet 10 3.0 0.012 0.017 106.447 106.452 dbcsr_multiply_generic 2485 12.5 0.183 0.191 72.832 73.285 qs_scf_new_mos 116 7.6 0.001 0.001 72.410 72.528 qs_scf_loop_do_ot 116 8.6 0.001 0.001 72.409 72.527 ot_scf_mini 116 9.6 0.003 0.004 68.025 68.125 multiply_cannon 2485 13.5 0.552 0.566 54.041 57.392 multiply_cannon_loop 2485 14.5 0.804 0.831 51.058 51.961 ot_mini 116 10.6 0.001 0.001 37.666 37.773 mp_waitall_1 124680 16.7 25.698 32.315 25.698 32.315 init_scf_loop 11 6.9 0.000 0.000 30.798 30.799 rebuild_ks_matrix 127 8.3 0.001 0.001 28.991 29.110 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.020 28.990 29.109 prepare_preconditioner 11 7.9 0.000 0.000 26.915 26.929 make_preconditioner 11 8.9 0.000 0.000 26.915 26.929 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.410 26.518 make_full_inverse_cholesky 11 9.9 0.022 0.029 25.017 25.271 multiply_cannon_multrec 9940 15.5 10.223 14.755 18.071 22.466 multiply_cannon_metrocomm3 9940 15.5 0.024 0.026 13.204 20.792 qs_ot_get_derivative 116 11.6 0.001 0.002 20.608 20.714 qs_ot_get_p 127 10.4 0.001 0.002 17.594 17.723 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.172 17.344 apply_single 127 13.6 0.001 0.001 17.172 17.344 ot_diis_step 116 11.6 0.019 0.020 16.986 16.986 make_m2s 4970 13.5 0.064 0.068 14.954 15.822 make_images 4970 14.5 2.135 2.561 14.649 15.521 cp_fm_cholesky_invert 11 10.9 14.830 14.841 14.830 14.841 qs_ot_p2m_diag 82 11.4 0.489 0.495 13.803 13.818 sum_up_and_integrate 127 10.3 0.002 0.002 13.758 13.814 integrate_v_rspace 127 11.3 0.004 0.004 13.698 13.757 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.483 13.517 calculate_rho_elec 127 8.7 0.255 0.266 13.482 13.517 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.635 12.636 multiply_cannon_sync_h2d 9940 15.5 10.624 11.350 10.624 11.350 init_scf_run 11 5.9 0.000 0.001 10.789 10.790 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.789 10.789 make_images_data 4970 15.5 0.054 0.066 8.694 9.990 hybrid_alltoall_any 5155 16.4 0.836 3.746 8.480 9.822 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.676 9.686 cp_fm_diag_elpa_base 82 14.4 9.439 9.523 9.672 9.682 dbcsr_mm_accdrv_process 20590 16.1 3.139 4.499 7.508 8.170 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.058 8.134 fft_wrap_pw1pw2 1281 11.7 0.017 0.020 8.099 8.121 grid_integrate_task_list 127 12.3 7.719 7.991 7.719 7.991 multiply_cannon_metrocomm1 9940 15.5 0.029 0.030 4.536 7.609 wfi_extrapolate 11 7.9 0.001 0.001 7.163 7.163 fft_wrap_pw1pw2_140 519 12.2 0.432 0.436 7.117 7.142 cp_fm_cholesky_decompose 22 10.9 6.549 6.600 6.549 6.600 density_rs2pw 127 9.7 0.005 0.005 6.208 6.438 calculate_dm_sparse 127 9.5 0.001 0.001 6.078 6.169 fft3d_ps 1281 13.7 3.094 3.202 6.086 6.114 grid_collocate_task_list 127 9.7 5.525 5.795 5.525 5.795 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.756 5.769 multiply_cannon_metrocomm4 7455 15.5 0.024 0.027 1.846 5.490 dbcsr_complete_redistribute 393 12.7 1.988 2.027 5.120 5.476 mp_irecv_dv 28618 15.9 1.807 5.405 1.807 5.405 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.307 5.308 mp_alltoall_d11v 2401 14.1 4.692 5.065 4.692 5.065 mp_allgather_i34 2485 14.5 1.183 4.748 1.183 4.748 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.645 4.715 potential_pw2rs 127 12.3 0.026 0.027 3.969 3.978 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.587 3.872 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.647 3.720 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.639 3.669 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.519 3.614 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.240 3.554 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.476 3.489 calculate_first_density_matrix 1 7.0 0.000 0.000 3.379 3.382 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=165.327000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1350.090909, yerr=33.451334 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/20/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 11.785520E+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 1979552 0.0% 0.0% 100.0% average stack size 0.0 0.0 3438.3 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 2.951115E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101080 MPI messages size (bytes): total size 1.144901E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.326685E+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 45568 35366371328 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726591885696 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4031 68213. MP_Allreduce 11137 1501. MP_Sync 87 MP_Alltoall 1724 36993716. 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 0.046 0.066 254.125 254.126 qs_mol_dyn_low 1 2.0 0.003 0.003 253.491 253.502 qs_forces 11 3.9 0.004 0.005 253.394 253.400 qs_energies 11 4.9 0.002 0.002 244.518 244.524 scf_env_do_scf 11 5.9 0.001 0.001 220.386 220.402 velocity_verlet 10 3.0 0.002 0.002 176.128 176.135 scf_env_do_scf_inner_loop 118 6.6 0.004 0.008 135.078 135.080 qs_scf_new_mos 118 7.6 0.001 0.001 91.440 91.564 qs_scf_loop_do_ot 118 8.6 0.001 0.001 91.439 91.563 dbcsr_multiply_generic 2527 12.6 0.287 0.296 86.254 86.933 ot_scf_mini 118 9.6 0.003 0.004 86.347 86.430 init_scf_loop 11 6.9 0.000 0.000 84.981 84.984 prepare_preconditioner 11 7.9 0.000 0.000 80.156 80.173 make_preconditioner 11 8.9 0.000 0.000 80.156 80.173 make_full_inverse_cholesky 11 9.9 0.038 0.039 63.361 77.364 multiply_cannon 2527 13.6 0.695 0.760 61.982 63.348 multiply_cannon_loop 2527 14.6 1.052 1.076 57.575 59.483 cp_fm_upper_to_full 105 14.8 34.930 50.172 34.930 50.172 ot_mini 118 10.6 0.001 0.001 45.801 45.901 mp_waitall_1 104500 16.8 30.140 35.504 30.140 35.504 rebuild_ks_matrix 129 8.3 0.001 0.001 34.404 34.536 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 34.404 34.536 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.015 32.126 dbcsr_complete_redistribute 395 12.7 3.863 3.998 23.126 31.440 copy_fm_to_dbcsr 209 11.7 0.001 0.001 19.396 27.674 qs_ot_get_derivative 118 11.6 0.002 0.002 25.122 25.210 qs_ot_get_p 129 10.4 0.002 0.002 24.980 25.056 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 16.752 24.948 mp_alltoall_i22 718 14.1 14.510 22.766 14.510 22.766 multiply_cannon_metrocomm3 10108 15.6 0.024 0.025 20.244 22.170 make_m2s 5054 13.6 0.074 0.076 19.362 21.115 qs_ot_p2m_diag 83 11.4 0.879 0.885 20.654 20.693 ot_diis_step 118 11.6 0.022 0.023 20.634 20.634 make_images 5054 14.6 3.103 3.307 18.876 20.632 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.821 20.120 apply_single 129 13.6 0.001 0.001 19.820 20.119 multiply_cannon_multrec 10108 15.6 10.376 12.245 19.147 19.390 cp_fm_cholesky_invert 11 10.9 19.055 19.068 19.055 19.068 cp_dbcsr_syevd 83 12.4 0.006 0.006 18.824 18.827 qs_rho_update_rho_low 129 7.7 0.001 0.001 16.119 16.138 calculate_rho_elec 129 8.7 0.486 0.487 16.119 16.137 sum_up_and_integrate 129 10.3 0.002 0.002 15.997 16.069 integrate_v_rspace 129 11.3 0.004 0.004 15.933 16.006 cp_fm_diag_elpa 83 13.4 0.000 0.000 15.516 15.517 cp_fm_diag_elpa_base 83 14.4 12.612 13.670 15.512 15.512 multiply_cannon_sync_h2d 10108 15.6 14.430 14.514 14.430 14.514 hybrid_alltoall_any 5240 16.5 1.321 3.086 11.298 14.003 make_images_data 5054 15.6 0.062 0.067 11.082 13.644 init_scf_run 11 5.9 0.000 0.001 13.529 13.529 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.529 13.529 dbcsr_mm_accdrv_process 20926 16.1 4.247 6.061 8.517 10.639 fft_wrap_pw1pw2 1301 11.7 0.020 0.020 10.325 10.332 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.998 10.059 fft_wrap_pw1pw2_140 527 12.2 0.830 0.833 9.095 9.112 wfi_extrapolate 11 7.9 0.001 0.001 9.063 9.063 grid_integrate_task_list 129 12.3 8.620 8.817 8.620 8.817 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.348 8.349 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.879 7.947 cp_fm_cholesky_decompose 22 10.9 7.647 7.698 7.647 7.698 mp_alltoall_d11v 2423 14.1 7.532 7.680 7.532 7.680 fft3d_ps 1301 13.7 4.032 4.056 7.605 7.619 density_rs2pw 129 9.7 0.005 0.005 7.107 7.158 calculate_dm_sparse 129 9.5 0.001 0.001 7.032 7.082 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.454 6.623 grid_collocate_task_list 129 9.7 6.459 6.536 6.459 6.536 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.741 5.805 qs_env_update_s_mstruct 11 6.9 0.000 0.000 5.183 5.348 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.829 5.235 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=254.126000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2653.363636, yerr=151.672058 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 1.270686E+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 1033 290947. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.054 85.074 85.076 qs_energies 1 2.0 0.000 0.000 84.510 84.516 ls_scf 1 3.0 0.000 0.000 83.630 83.638 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.650 72.819 multiply_cannon 111 7.7 0.017 0.021 55.895 57.470 multiply_cannon_loop 111 8.7 0.228 0.241 52.451 54.186 ls_scf_main 1 4.0 0.000 0.000 52.380 52.381 density_matrix_trs4 2 5.0 0.002 0.003 46.841 46.933 ls_scf_init_scf 1 4.0 0.000 0.000 28.239 28.240 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.082 27.141 mp_waitall_1 11031 10.9 22.569 25.041 22.569 25.041 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.002 25.021 multiply_cannon_multrec 2664 9.7 8.123 8.919 15.515 17.260 multiply_cannon_sync_h2d 2664 9.7 13.442 15.231 13.442 15.231 make_m2s 222 7.7 0.009 0.011 13.081 13.653 make_images 222 8.7 0.099 0.109 13.059 13.633 multiply_cannon_metrocomm1 2664 9.7 0.010 0.010 9.705 12.638 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.525 8.393 make_images_data 222 9.7 0.004 0.005 7.672 8.238 hybrid_alltoall_any 227 10.6 0.216 1.828 6.570 8.128 dbcsr_mm_accdrv_process 4760 10.4 0.594 0.711 7.011 7.967 dbcsr_mm_accdrv_process_sort 4760 11.4 6.214 7.101 6.214 7.101 calculate_norms 4752 9.8 5.485 6.097 5.485 6.097 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.050 5.160 mp_sum_l 887 5.1 3.172 4.614 3.172 4.614 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.058 3.536 mp_irecv_dv 6231 10.9 2.042 3.512 2.042 3.512 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.364 3.509 make_images_sizes 222 9.7 0.000 0.000 0.684 3.323 mp_alltoall_i44 222 10.7 0.684 3.322 0.684 3.322 arnoldi_extremal 4 6.8 0.000 0.000 3.204 3.233 arnoldi_normal_ev 4 7.8 0.001 0.003 3.204 3.233 build_subspace 16 8.4 0.009 0.012 3.116 3.117 ls_scf_post 1 4.0 0.000 0.000 3.012 3.018 ls_scf_store_result 1 5.0 0.000 0.000 2.837 2.873 dbcsr_special_finalize 555 9.7 0.005 0.006 2.366 2.806 dbcsr_merge_single_wm 555 10.7 0.464 0.599 2.357 2.797 make_images_pack 222 9.7 2.211 2.632 2.213 2.634 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.308 2.547 dbcsr_sort_data 658 11.4 2.151 2.511 2.151 2.511 dbcsr_matrix_vector_mult_local 304 10.0 2.062 2.450 2.064 2.452 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.281 2.350 buffer_matrices_ensure_size 222 8.7 1.744 2.055 1.744 2.055 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.809 1.810 rebuild_ks_matrix 3 7.3 0.000 0.000 1.800 1.801 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.800 1.801 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.076000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1136.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 2.195878E+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 1033 301899. 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 0.033 0.085 90.161 90.162 qs_energies 1 2.0 0.001 0.007 89.564 89.569 ls_scf 1 3.0 0.000 0.001 88.224 88.228 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.148 74.407 multiply_cannon 111 7.7 0.027 0.040 52.774 56.593 ls_scf_main 1 4.0 0.002 0.034 54.408 54.410 multiply_cannon_loop 111 8.7 0.135 0.150 50.051 53.133 density_matrix_trs4 2 5.0 0.003 0.034 48.645 48.846 mp_waitall_1 9105 10.9 21.062 30.267 21.062 30.267 ls_scf_init_scf 1 4.0 0.000 0.003 30.167 30.179 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.870 28.967 multiply_cannon_multrec 1332 9.7 13.144 17.062 22.409 27.462 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.459 26.467 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.780 20.801 make_m2s 222 7.7 0.006 0.008 15.088 15.626 make_images 222 8.7 1.371 1.692 15.057 15.596 dbcsr_mm_accdrv_process 4041 10.4 0.327 0.501 8.863 10.431 dbcsr_mm_accdrv_process_sort 4041 11.4 8.379 9.897 8.379 9.897 make_images_data 222 9.7 0.004 0.005 8.754 9.597 hybrid_alltoall_any 227 10.6 0.542 2.567 8.147 9.546 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.198 7.955 mp_irecv_dv 3311 11.0 3.177 7.900 3.177 7.900 mp_sum_l 887 5.1 4.836 7.871 4.836 7.871 calculate_norms 2376 9.8 6.072 6.898 6.072 6.898 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.613 6.200 multiply_cannon_sync_h2d 1332 9.7 4.903 6.164 4.903 6.164 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.957 5.173 arnoldi_extremal 4 6.8 0.000 0.000 4.708 4.724 arnoldi_normal_ev 4 7.8 0.001 0.005 4.708 4.724 build_subspace 16 8.4 0.014 0.020 4.456 4.459 ls_scf_post 1 4.0 0.000 0.001 3.650 3.655 ls_scf_store_result 1 5.0 0.000 0.000 3.402 3.474 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.144 3.384 dbcsr_matrix_vector_mult_local 304 10.0 2.735 3.209 2.737 3.211 mp_allgather_i34 111 8.7 0.705 2.871 0.705 2.871 ls_scf_dm_to_ks 2 5.0 0.000 0.013 2.648 2.743 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.174 2.558 dbcsr_data_new 4174 10.1 2.115 2.389 2.115 2.389 make_images_pack 222 9.7 1.819 2.138 1.822 2.140 dbcsr_sort_data 436 11.2 1.829 2.082 1.829 2.082 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.975 1.985 rebuild_ks_matrix 3 7.3 0.000 0.000 1.962 1.973 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.014 1.962 1.973 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.162000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1841.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 2.918568E+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 1033 300682. 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 0.076 0.186 91.904 91.905 qs_energies 1 2.0 0.000 0.000 91.132 91.136 ls_scf 1 3.0 0.000 0.000 89.719 89.722 dbcsr_multiply_generic 111 6.7 0.016 0.018 74.416 74.776 multiply_cannon 111 7.7 0.030 0.061 51.525 56.324 ls_scf_main 1 4.0 0.000 0.000 55.893 55.897 multiply_cannon_loop 111 8.7 0.116 0.129 48.800 52.248 density_matrix_trs4 2 5.0 0.002 0.003 50.009 50.181 mp_waitall_1 7281 11.0 23.081 33.056 23.081 33.056 ls_scf_init_scf 1 4.0 0.000 0.000 30.187 30.189 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.988 29.080 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.634 26.643 multiply_cannon_multrec 888 9.7 12.638 15.117 21.223 24.273 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.762 22.465 make_m2s 222 7.7 0.006 0.007 16.552 17.237 make_images 222 8.7 1.583 1.851 16.514 17.199 make_images_data 222 9.7 0.004 0.004 9.612 10.748 hybrid_alltoall_any 227 10.6 0.641 2.955 9.075 10.442 dbcsr_mm_accdrv_process 3754 10.4 0.319 0.501 8.106 9.352 dbcsr_mm_accdrv_process_sort 3754 11.4 7.669 8.851 7.669 8.851 mp_sum_l 887 5.1 4.738 7.927 4.738 7.927 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.469 7.302 mp_irecv_dv 2335 11.1 2.453 7.242 2.453 7.242 multiply_cannon_sync_h2d 888 9.7 6.119 6.975 6.119 6.975 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.350 6.406 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.531 6.246 arnoldi_extremal 4 6.8 0.000 0.000 5.122 5.137 arnoldi_normal_ev 4 7.8 0.001 0.005 5.122 5.137 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.747 4.987 build_subspace 16 8.4 0.014 0.020 4.822 4.830 calculate_norms 1584 9.8 4.343 4.769 4.343 4.769 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.455 3.804 mp_allgather_i34 111 8.7 0.832 3.759 0.832 3.759 ls_scf_post 1 4.0 0.000 0.000 3.639 3.642 dbcsr_matrix_vector_mult_local 304 10.0 3.009 3.585 3.011 3.587 ls_scf_store_result 1 5.0 0.000 0.000 3.394 3.465 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.825 2.915 dbcsr_data_new 4116 9.9 2.120 2.486 2.120 2.486 make_images_sizes 222 9.7 0.000 0.000 1.091 2.215 mp_alltoall_i44 222 10.7 1.091 2.215 1.091 2.215 dbcsr_sort_data 325 11.1 1.888 2.153 1.888 2.153 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.914 1.916 dbcsr_finalize 304 7.8 0.026 0.032 1.608 1.905 rebuild_ks_matrix 3 7.3 0.000 0.000 1.896 1.898 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.896 1.898 make_images_pack 222 9.7 1.624 1.888 1.627 1.890 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=91.905000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2232.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 3.426214E+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 1033 301899. 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 0.143 0.155 98.488 98.489 qs_energies 1 2.0 0.000 0.000 97.118 97.124 ls_scf 1 3.0 0.000 0.000 95.445 95.450 dbcsr_multiply_generic 111 6.7 0.017 0.018 79.128 79.419 ls_scf_main 1 4.0 0.000 0.000 59.223 59.224 multiply_cannon 111 7.7 0.047 0.086 51.729 56.260 density_matrix_trs4 2 5.0 0.002 0.003 53.093 53.247 multiply_cannon_loop 111 8.7 0.154 0.172 46.743 50.371 ls_scf_init_scf 1 4.0 0.000 0.000 32.972 32.974 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.767 31.852 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.136 29.146 mp_waitall_1 6369 11.0 23.128 28.773 23.128 28.773 multiply_cannon_multrec 1332 9.7 14.055 17.234 21.978 25.476 make_m2s 222 7.7 0.007 0.008 21.502 22.885 make_images 222 8.7 3.141 3.604 21.451 22.836 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.277 17.463 make_images_data 222 9.7 0.004 0.004 11.996 13.686 hybrid_alltoall_any 227 10.6 0.796 3.757 11.338 13.120 dbcsr_mm_accdrv_process 3641 10.4 0.299 0.485 7.561 9.086 dbcsr_mm_accdrv_process_sort 3641 11.4 7.116 8.585 7.116 8.585 mp_sum_l 887 5.1 4.322 7.840 4.322 7.840 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.346 6.284 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.088 5.958 multiply_cannon_sync_h2d 1332 9.7 5.443 5.955 5.443 5.955 mp_irecv_dv 3229 10.9 2.064 5.888 2.064 5.888 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.685 5.599 arnoldi_extremal 4 6.8 0.000 0.000 5.268 5.281 arnoldi_normal_ev 4 7.8 0.001 0.004 5.268 5.280 build_subspace 16 8.4 0.014 0.021 4.930 4.937 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.732 4.931 calculate_norms 2376 9.8 4.165 4.550 4.165 4.550 mp_allgather_i34 111 8.7 2.097 4.224 2.097 4.224 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.600 3.890 dbcsr_matrix_vector_mult_local 304 10.0 3.173 3.672 3.175 3.674 dbcsr_sort_data 658 11.4 3.056 3.456 3.056 3.456 dbcsr_special_finalize 555 9.7 0.006 0.007 2.806 3.265 dbcsr_merge_single_wm 555 10.7 0.535 0.660 2.797 3.257 ls_scf_post 1 4.0 0.000 0.000 3.250 3.256 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.997 3.071 ls_scf_store_result 1 5.0 0.000 0.000 2.985 3.051 dbcsr_data_release 10477 10.7 1.582 2.398 1.582 2.398 make_images_pack 222 9.7 1.672 2.098 1.675 2.100 dbcsr_finalize 304 7.8 0.049 0.061 1.793 1.985 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.489000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2774.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 4.832588E+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 1033 300769. 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 0.046 0.056 92.284 92.285 qs_energies 1 2.0 0.000 0.000 91.464 91.470 ls_scf 1 3.0 0.000 0.000 89.441 89.447 dbcsr_multiply_generic 111 6.7 0.017 0.018 71.031 71.278 ls_scf_main 1 4.0 0.000 0.000 56.685 56.685 multiply_cannon 111 7.7 0.096 0.180 52.893 56.200 multiply_cannon_loop 111 8.7 0.088 0.095 50.305 51.721 density_matrix_trs4 2 5.0 0.002 0.003 49.713 49.779 ls_scf_init_scf 1 4.0 0.000 0.000 29.319 29.324 mp_waitall_1 5436 11.0 24.711 28.529 24.711 28.529 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.078 28.112 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.974 25.987 multiply_cannon_multrec 444 9.7 13.555 16.640 20.658 22.192 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.147 15.760 make_m2s 222 7.7 0.005 0.005 13.486 14.414 make_images 222 8.7 2.037 2.473 13.419 14.344 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.169 12.542 hybrid_alltoall_any 227 10.6 0.801 3.801 8.158 9.676 make_images_data 222 9.7 0.003 0.004 8.210 9.616 dbcsr_mm_accdrv_process 3003 10.4 0.343 0.416 6.783 7.918 multiply_cannon_sync_h2d 444 9.7 6.790 7.889 6.790 7.889 dbcsr_mm_accdrv_process_sort 3003 11.4 6.417 7.514 6.417 7.514 arnoldi_extremal 4 6.8 0.000 0.000 5.822 5.836 arnoldi_normal_ev 4 7.8 0.002 0.005 5.822 5.836 build_subspace 16 8.4 0.015 0.019 5.431 5.436 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.427 4.655 mp_sum_l 887 5.1 2.641 4.471 2.641 4.471 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.149 4.356 dbcsr_matrix_vector_mult_local 304 10.0 3.667 4.140 3.669 4.142 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.562 3.994 mp_irecv_dv 1241 11.2 1.550 3.963 1.550 3.963 calculate_norms 792 9.8 3.627 3.750 3.627 3.750 mp_allgather_i34 111 8.7 1.133 3.699 1.133 3.699 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.529 3.610 ls_scf_post 1 4.0 0.000 0.000 3.438 3.443 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.842 3.331 ls_scf_store_result 1 5.0 0.000 0.000 3.225 3.263 make_images_sizes 222 9.7 0.000 0.000 0.894 3.199 mp_alltoall_i44 222 10.7 0.893 3.198 0.893 3.198 dbcsr_finalize 304 7.8 0.062 0.077 2.205 2.283 dbcsr_data_new 4608 9.7 1.778 2.218 1.778 2.218 dbcsr_merge_all 275 8.9 0.479 0.524 2.061 2.139 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.053 2.053 rebuild_ks_matrix 3 7.3 0.000 0.000 2.019 2.020 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.019 2.020 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 2.008 2.008 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.285000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3736.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_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 8.998183E+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 1033 319197. 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 0.229 0.285 106.056 106.057 qs_energies 1 2.0 0.000 0.000 103.969 103.978 ls_scf 1 3.0 0.000 0.000 101.018 101.026 dbcsr_multiply_generic 111 6.7 0.024 0.027 75.010 75.147 ls_scf_main 1 4.0 0.000 0.000 63.665 63.666 density_matrix_trs4 2 5.0 0.002 0.003 54.848 54.900 multiply_cannon 111 7.7 0.171 0.259 48.508 50.940 multiply_cannon_loop 111 8.7 0.097 0.099 45.370 45.968 ls_scf_init_scf 1 4.0 0.000 0.000 33.548 33.549 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.135 32.157 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.437 29.442 mp_waitall_1 4527 11.1 21.885 25.866 21.885 25.866 make_m2s 222 7.7 0.005 0.005 22.836 24.062 make_images 222 8.7 3.580 3.900 22.728 23.952 multiply_cannon_multrec 444 9.7 17.875 18.469 22.521 23.120 hybrid_alltoall_any 227 10.6 1.655 3.629 12.941 15.885 make_images_data 222 9.7 0.003 0.004 13.181 15.469 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.105 10.366 multiply_cannon_sync_h2d 444 9.7 8.805 8.927 8.805 8.927 arnoldi_extremal 4 6.8 0.000 0.000 7.360 7.368 arnoldi_normal_ev 4 7.8 0.003 0.009 7.360 7.368 build_subspace 16 8.4 0.026 0.036 6.792 6.812 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.402 5.575 dbcsr_matrix_vector_mult_local 304 10.0 4.970 5.291 4.972 5.294 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.883 5.130 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.930 5.026 dbcsr_mm_accdrv_process 1814 10.4 0.268 0.354 4.460 4.600 dbcsr_mm_accdrv_process_sort 1814 11.4 4.124 4.272 4.124 4.272 mp_allgather_i34 111 8.7 1.248 4.003 1.248 4.003 ls_scf_post 1 4.0 0.000 0.000 3.804 3.814 make_images_sizes 222 9.7 0.000 0.000 1.511 3.698 mp_alltoall_i44 222 10.7 1.511 3.697 1.511 3.697 ls_scf_store_result 1 5.0 0.000 0.000 3.526 3.565 calculate_norms 792 9.8 3.220 3.286 3.220 3.286 dbcsr_finalize 304 7.8 0.082 0.089 3.085 3.186 dbcsr_merge_all 275 8.9 0.886 0.920 2.870 2.965 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 2.920 2.920 dbcsr_complete_redistribute 5 7.6 1.424 1.470 2.738 2.848 matrix_ls_to_qs 2 6.0 0.000 0.000 2.406 2.529 dbcsr_sort_data 325 11.1 2.444 2.522 2.444 2.522 dbcsr_data_new 6591 9.6 1.834 2.395 1.834 2.395 dbcsr_new_transposed 4 7.5 0.242 0.253 2.301 2.318 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.293 2.295 dbcsr_frobenius_norm 74 6.6 2.056 2.135 2.212 2.246 rebuild_ks_matrix 3 7.3 0.000 0.000 2.227 2.229 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.227 2.229 dbcsr_add_d 103 6.2 0.000 0.000 2.125 2.207 dbcsr_data_release 12724 10.6 1.982 2.207 1.982 2.207 dbcsr_add_anytype 103 7.2 0.859 0.893 2.125 2.207 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.057000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7213.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 589.029376E+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 77937. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 993456946. 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 0.042 0.174 264.640 264.642 qs_mol_dyn_low 1 2.0 0.007 0.038 263.313 263.339 qs_forces 5 3.8 0.005 0.024 263.063 263.066 qs_energies 5 4.8 0.005 0.041 259.624 259.655 scf_env_do_scf 5 5.8 0.001 0.017 245.055 245.056 scf_env_do_scf_inner_loop 105 6.6 0.005 0.054 216.128 216.167 qs_scf_new_mos 105 7.6 0.000 0.001 167.053 167.181 qs_scf_loop_do_ot 105 8.6 0.001 0.001 167.053 167.181 ot_scf_mini 105 9.6 0.003 0.004 156.888 157.079 dbcsr_multiply_generic 1445 12.2 0.129 0.139 138.259 138.607 multiply_cannon 1445 13.2 0.276 0.287 116.729 119.118 velocity_verlet 4 3.0 0.024 0.119 116.120 116.126 multiply_cannon_loop 1445 14.2 2.866 2.993 114.362 115.787 qs_ot_get_p 112 10.4 0.001 0.001 63.134 63.530 ot_mini 105 10.6 0.001 0.001 62.956 63.080 qs_ot_p2m_diag 40 11.0 0.020 0.031 50.967 51.057 cp_dbcsr_syevd 40 12.0 0.002 0.011 47.533 47.536 mp_waitall_1 488190 16.1 39.795 47.197 39.795 47.197 multiply_cannon_multrec 69360 15.2 29.584 35.343 40.590 46.016 cp_fm_syevd 40 13.0 0.000 0.007 41.220 41.387 qs_ot_get_derivative 55 11.6 0.001 0.010 41.125 41.244 cp_fm_redistribute_end 40 14.0 18.117 36.108 18.147 36.117 cp_fm_syevd_base 40 14.0 17.948 35.935 17.948 35.935 multiply_cannon_metrocomm3 69360 15.2 0.200 0.212 26.976 35.595 multiply_cannon_sync_h2d 69360 15.2 28.673 33.333 28.673 33.333 rebuild_ks_matrix 110 8.4 0.000 0.000 32.798 32.996 qs_ks_build_kohn_sham_matrix 110 9.4 0.013 0.045 32.798 32.996 qs_ks_update_qs_env 112 7.6 0.001 0.001 30.030 30.214 init_scf_loop 7 6.6 0.000 0.004 28.844 28.850 qs_rho_update_rho_low 110 7.6 0.000 0.001 23.077 23.441 calculate_rho_elec 110 8.6 0.028 0.031 23.077 23.440 prepare_preconditioner 7 7.6 0.000 0.000 23.403 23.437 make_preconditioner 7 8.6 0.001 0.009 23.403 23.437 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.119 23.387 apply_single 62 13.6 0.000 0.000 23.119 23.387 ot_new_cg_direction 55 11.6 0.000 0.001 21.136 21.137 density_rs2pw 110 9.6 0.004 0.006 16.979 17.453 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.452 15.684 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.851 14.994 make_full_inverse_cholesky 7 9.6 0.000 0.001 14.841 14.911 transfer_rs2pw 445 10.6 0.008 0.010 13.391 13.881 mp_sum_l 4764 12.2 13.033 13.874 13.033 13.874 fft_wrap_pw1pw2 1425 12.5 0.019 0.024 12.196 12.430 init_scf_run 5 5.8 0.000 0.001 12.116 12.117 scf_env_initial_rho_setup 5 6.8 0.002 0.003 12.116 12.117 dbcsr_mm_accdrv_process 154766 15.8 6.389 6.570 10.875 12.106 calculate_dm_sparse 110 9.5 0.000 0.001 11.860 12.053 sum_up_and_integrate 60 10.3 0.001 0.003 11.693 11.705 integrate_v_rspace 60 11.3 0.002 0.003 11.675 11.689 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 11.010 11.076 fft_wrap_pw1pw2_240 915 14.0 0.196 0.226 10.564 10.780 qs_vxc_create 110 10.4 0.002 0.003 10.604 10.640 check_diag 80 13.5 8.652 8.949 10.417 10.586 cp_fm_cholesky_invert 7 10.6 9.905 9.930 9.905 9.930 fft3d_pb 915 15.0 2.235 2.429 9.549 9.789 make_m2s 2890 13.2 0.077 0.085 8.106 8.699 make_images 2890 14.2 0.239 0.259 7.999 8.591 make_full_single_inverse 7 9.6 0.002 0.036 8.191 8.225 transfer_rs2pw_30 110 11.6 1.255 1.332 6.874 7.850 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 7.754 7.776 multiply_cannon_metrocomm1 69360 15.2 0.095 0.102 5.238 7.738 potential_pw2rs 60 12.3 0.002 0.003 7.642 7.725 acc_transpose_blocks 69360 15.2 0.340 0.355 6.909 7.723 mp_sendrecv_dv 168740 12.6 7.402 7.662 7.402 7.662 xc_rho_set_and_dset_create 110 12.4 0.071 0.080 7.366 7.610 calculate_first_density_matrix 1 7.0 0.000 0.004 7.249 7.256 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.001 7.132 7.187 xc_vxc_pw_create 60 11.3 0.036 0.042 7.101 7.141 mp_alltoall_z22v 2340 16.7 6.819 7.128 6.819 7.128 transfer_pw2rs 245 13.2 0.003 0.004 6.448 6.532 xc_pw_derive 510 13.4 0.005 0.006 6.358 6.403 mp_alltoall_d11v 1300 13.8 4.569 5.605 4.569 5.605 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=264.642000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=558.200000, yerr=3.487119 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 1b0cf2987a3cc1c0f88b561a560f5440bd7b29f8 Summary: empty Status: OK