=== 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: 98431332714702b8c1de37aafb7c70fe7a47c205 ################# 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/01 job id: 53357938 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/02 job id: 53357939 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/03 job id: 53357940 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/04 job id: 53357941 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/05 job id: 53357942 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/06 job id: 53357943 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/07 job id: 53357944 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/08 job id: 53357946 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/09 job id: 53357947 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/10 job id: 53357948 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/11 job id: 53357949 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/12 job id: 53357950 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/13 job id: 53357952 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/14 job id: 53357953 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/15 job id: 53357954 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/16 job id: 53357955 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/17 job id: 53357956 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/18 job id: 53357957 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/19 job id: 53357958 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/20 job id: 53357959 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/21 job id: 53357960 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/22 job id: 53357961 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/23 job id: 53357962 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/24 job id: 53357963 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/25 job id: 53357964 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/26 job id: 53357965 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/27 job id: 53357966 --- 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/98431332714702b8c1de37aafb7c70fe7a47c205_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 0.056 0.109 129.530 129.531 farming_run 1 2.0 128.530 128.564 129.438 129.472 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.544413E+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 0.006 0.025 110.486 110.487 qs_energies 1 2.0 0.001 0.001 110.235 110.238 mp2_main 1 3.0 0.000 0.000 108.387 108.391 mp2_gpw_main 1 4.0 0.051 0.080 107.445 107.448 mp2_ri_gpw_compute_in 1 5.0 0.192 0.274 88.697 88.954 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 57.202 57.439 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.174 43.649 48.617 get_2c_integrals 1 6.0 0.008 0.009 30.873 31.241 integrate_v_rspace 273 8.0 0.442 0.457 25.571 30.165 fft_wrap_pw1pw2 5465 10.4 0.067 0.072 28.138 28.776 grid_integrate_task_list 273 9.0 20.802 26.358 20.802 26.358 fft_wrap_pw1pw2_100 2178 11.4 1.302 1.440 25.585 26.222 compute_2c_integrals 1 7.0 0.002 0.002 21.487 21.521 compute_2c_integrals_loop_lm 1 8.0 0.002 0.004 20.913 21.191 mp2_eri_2c_integrate_gpw 1 9.0 2.405 2.441 20.912 21.188 rpa_ri_compute_en 1 5.0 0.019 0.022 18.604 18.686 fft3d_s 5443 12.4 16.914 17.365 16.935 17.387 ao_to_mo_and_store_B_mult_1 272 7.0 10.767 15.298 10.767 15.298 calculate_wavefunction 272 8.0 5.265 5.418 12.742 13.630 calc_potential_gpw 544 9.5 0.004 0.004 10.885 11.758 rpa_num_int 1 6.0 0.001 0.008 10.540 10.541 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 9.822 10.528 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.494 10.517 potential_pw2rs 545 10.0 0.106 0.108 8.495 10.314 cp_fm_cholesky_decompose 12 8.2 9.574 9.990 9.574 9.990 calc_mat_Q 8 8.0 0.000 0.000 9.448 9.540 cholesky_decomp 1 7.0 0.000 0.000 8.577 8.989 contract_S_to_Q 8 9.0 0.000 0.000 8.871 8.959 collocate_single_gaussian 272 10.0 0.040 0.042 7.867 8.725 parallel_gemm_fm 14 9.1 0.000 0.000 8.497 8.618 parallel_gemm_fm_cosma 14 10.1 8.497 8.618 8.497 8.618 create_integ_mat 1 6.0 0.023 0.030 7.734 7.735 array2fm 1 7.0 0.000 0.000 6.798 7.152 pw_gather_s 2722 12.2 3.958 5.090 3.958 5.090 pw_poisson_solve 545 10.5 0.010 0.011 4.667 4.930 pw_scatter_s 2720 12.7 4.477 4.770 4.477 4.770 pw_poisson_set 548 11.5 0.019 0.019 3.364 3.586 array2fm_buffer_send 1 8.0 3.048 3.171 3.048 3.171 pw_copy 4911 11.6 2.610 2.768 2.610 2.768 ao_to_mo_and_store_B_E_Ex_1 272 7.0 1.081 1.228 2.023 2.220 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=107.447730, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2784.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_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 0.028 0.053 425.984 426.042 farming_run 1 2.0 424.938 424.976 425.743 425.768 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.273627E+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 0.033 222.301 222.302 qs_energies 1 2.0 0.003 0.008 222.042 222.074 scf_env_do_scf 1 3.0 0.000 0.000 114.086 114.086 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.238 113.245 rebuild_ks_matrix 4 6.0 0.000 0.000 113.237 113.244 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.060 113.237 113.244 hfx_ks_matrix 4 8.0 0.001 0.001 112.893 112.896 integrate_four_center 4 9.0 0.154 0.469 112.892 112.896 mp2_main 1 3.0 0.001 0.014 107.671 107.705 mp2_gpw_main 1 4.0 0.066 0.128 106.832 106.868 integrate_four_center_main 4 10.0 0.123 0.644 101.622 104.422 integrate_four_center_bin 264 11.0 101.499 104.329 101.499 104.329 init_scf_loop 1 4.0 0.000 0.000 95.969 95.969 mp2_ri_gpw_compute_in 1 5.0 0.072 0.108 79.735 80.828 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 56.708 57.810 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.164 44.672 49.460 integrate_v_rspace 95 8.0 0.399 0.572 28.343 32.944 fft_wrap_pw1pw2 1868 10.4 0.029 0.035 30.494 31.056 fft_wrap_pw1pw2_100 730 11.4 0.637 0.728 28.200 28.728 mp2_ri_gpw_compute_en 1 5.0 0.083 0.133 26.919 28.619 grid_integrate_task_list 95 9.0 23.641 28.456 23.641 28.456 ao_to_mo_and_store_B_mult_1 91 7.0 10.363 28.031 10.363 28.031 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.831 1.904 25.114 25.124 get_2c_integrals 1 6.0 0.001 0.006 22.933 22.982 compute_2c_integrals 1 7.0 0.002 0.003 22.101 22.132 compute_2c_integrals_loop_lm 1 8.0 0.002 0.006 21.725 21.955 mp2_eri_2c_integrate_gpw 1 9.0 1.745 1.862 21.723 21.954 fft3d_s 1823 12.4 19.243 19.768 19.256 19.782 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.116 18.116 calc_potential_gpw 182 9.5 0.002 0.003 12.433 13.479 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 10.683 11.308 calculate_wavefunction 91 8.0 1.977 2.015 9.995 10.785 potential_pw2rs 186 10.0 0.033 0.034 8.954 10.224 mp2_ri_gpw_compute_en_expansio 172 7.0 0.559 0.584 8.783 9.342 collocate_single_gaussian 91 10.0 0.018 0.022 8.299 8.861 local_gemm 172 8.0 8.224 8.758 8.224 8.758 mp2_ri_gpw_compute_en_comm 22 7.0 0.504 0.525 8.116 8.573 integrate_four_center_load 4 10.0 0.000 0.000 6.751 6.756 hfx_load_balance 1 11.0 0.000 0.000 6.751 6.756 mp_sync 37 10.5 3.249 6.697 3.249 6.697 mp_sendrecv_dm3 2068 8.0 6.137 6.561 6.137 6.561 mp2_ri_gpw_compute_en_ener 172 7.0 6.339 6.420 6.339 6.420 pw_poisson_solve 186 10.4 0.005 0.006 5.538 6.142 pw_gather_s 912 12.2 4.746 5.397 4.746 5.397 pw_scatter_s 910 12.7 4.148 4.854 4.148 4.854 pw_poisson_set 189 11.4 0.009 0.011 4.103 4.632 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=106.852061, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1527.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_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 445.706240E+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 635385. 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 0.024 0.065 53.984 53.985 qs_mol_dyn_low 1 2.0 0.003 0.004 53.711 53.723 qs_forces 11 3.9 0.002 0.003 53.639 53.640 qs_energies 11 4.9 0.022 0.107 52.118 52.147 scf_env_do_scf 11 5.9 0.000 0.001 45.692 45.692 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 43.848 43.848 dbcsr_multiply_generic 2286 12.5 0.109 0.134 33.977 34.535 qs_scf_new_mos 108 7.5 0.000 0.001 33.836 34.131 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.836 34.130 ot_scf_mini 108 9.5 0.002 0.002 32.069 32.263 multiply_cannon 2286 13.5 0.178 0.190 26.579 28.077 multiply_cannon_loop 2286 14.5 1.821 1.925 25.934 27.401 velocity_verlet 10 3.0 0.003 0.022 25.861 25.861 ot_mini 108 10.5 0.001 0.001 19.221 19.464 qs_ot_get_derivative 108 11.5 0.001 0.001 16.207 16.385 mp_waitall_1 245248 16.5 8.707 14.871 8.707 14.871 multiply_cannon_metrocomm3 54864 15.5 0.073 0.080 6.226 13.094 multiply_cannon_multrec 54864 15.5 3.601 5.690 7.593 10.962 qs_ot_get_p 119 10.4 0.001 0.001 8.077 8.386 rebuild_ks_matrix 119 8.3 0.000 0.000 7.829 7.965 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.829 7.965 mp_sum_l 7287 12.8 5.442 7.213 5.442 7.213 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.888 7.002 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.710 6.159 multiply_cannon_sync_h2d 54864 15.5 5.108 6.131 5.108 6.131 dbcsr_mm_accdrv_process 76910 16.1 1.850 2.909 3.907 5.537 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.201 5.305 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.199 5.238 init_scf_run 11 5.9 0.000 0.001 4.874 4.874 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.874 4.874 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.389 4.389 sum_up_and_integrate 119 10.3 0.001 0.002 4.371 4.378 integrate_v_rspace 119 11.3 0.002 0.002 4.360 4.369 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.104 4.301 calculate_rho_elec 119 8.7 0.011 0.016 4.104 4.300 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.181 4.182 cp_fm_redistribute_end 50 14.0 2.135 4.156 2.140 4.159 cp_fm_diag_elpa_base 50 14.0 2.013 4.048 2.017 4.056 calculate_dm_sparse 119 9.5 0.000 0.000 3.030 3.224 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.938 3.153 apply_single 119 13.6 0.000 0.000 2.938 3.153 acc_transpose_blocks 54864 15.5 0.223 0.245 2.414 3.130 calculate_first_density_matrix 1 7.0 0.000 0.000 2.996 3.006 multiply_cannon_metrocomm1 54864 15.5 0.056 0.061 1.843 2.858 jit_kernel_multiply 13 15.8 1.993 2.757 1.993 2.757 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.725 2.739 ot_diis_step 108 11.5 0.006 0.006 2.734 2.734 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.548 2.636 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.364 2.418 density_rs2pw 119 9.7 0.003 0.004 2.059 2.160 grid_integrate_task_list 119 12.3 2.023 2.110 2.023 2.110 mp_sum_d 4137 12.0 1.462 2.020 1.462 2.020 init_scf_loop 11 6.9 0.000 0.000 1.829 1.830 wfi_extrapolate 11 7.9 0.001 0.001 1.802 1.802 potential_pw2rs 119 12.3 0.004 0.004 1.745 1.758 make_m2s 4572 13.5 0.054 0.056 1.661 1.724 make_images 4572 14.5 0.134 0.139 1.579 1.641 mp_alltoall_d11v 2130 13.8 1.377 1.630 1.377 1.630 fft_wrap_pw1pw2 1201 11.6 0.011 0.012 1.471 1.542 transfer_rs2pw 487 10.6 0.006 0.006 1.390 1.501 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.429 1.462 acc_transpose_blocks_sync 164592 16.5 1.190 1.430 1.190 1.430 acc_transpose_blocks_kernels 54864 16.5 0.247 0.383 0.955 1.425 grid_collocate_task_list 119 9.7 1.351 1.407 1.351 1.407 mp_waitany 12084 13.8 1.273 1.378 1.273 1.378 transfer_pw2rs 487 13.2 0.005 0.006 1.322 1.333 fft3d_ps 1201 13.6 0.359 0.457 1.245 1.313 fft_wrap_pw1pw2_140 487 12.2 0.049 0.056 1.144 1.214 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.751 1.145 qs_energies_init_hamiltonians 11 5.9 0.031 0.111 1.029 1.112 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.985000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=425.272727, yerr=0.862439 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_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 486.539264E+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 1341781. 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 0.020 0.082 39.433 39.434 qs_mol_dyn_low 1 2.0 0.004 0.010 38.921 38.927 qs_forces 11 3.9 0.003 0.007 38.853 38.856 qs_energies 11 4.9 0.004 0.024 37.178 37.182 scf_env_do_scf 11 5.9 0.001 0.003 31.147 31.148 scf_env_do_scf_inner_loop 108 6.5 0.003 0.011 28.912 28.913 dbcsr_multiply_generic 2286 12.5 0.108 0.146 22.207 22.561 qs_scf_new_mos 108 7.5 0.001 0.001 20.499 20.731 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.499 20.730 ot_scf_mini 108 9.5 0.002 0.003 19.597 19.768 multiply_cannon 2286 13.5 0.206 0.215 17.089 18.506 velocity_verlet 10 3.0 0.002 0.008 17.890 17.891 multiply_cannon_loop 2286 14.5 1.187 1.239 15.953 17.447 ot_mini 108 10.5 0.001 0.001 12.105 12.344 mp_waitall_1 200699 16.5 5.579 10.947 5.579 10.947 qs_ot_get_derivative 108 11.5 0.001 0.001 9.759 9.931 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.086 9.760 multiply_cannon_multrec 27432 15.5 1.822 4.036 6.184 8.928 rebuild_ks_matrix 119 8.3 0.000 0.001 6.716 6.862 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.021 6.716 6.861 dbcsr_mm_accdrv_process 47894 16.0 3.506 5.496 4.284 6.369 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.920 6.053 qs_ot_get_p 119 10.4 0.001 0.001 4.588 4.816 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.646 4.537 init_scf_run 11 5.9 0.000 0.001 4.206 4.206 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.205 4.206 mp_sum_l 7287 12.8 2.107 4.103 2.107 4.103 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.037 4.031 apply_single 119 13.6 0.000 0.000 3.036 4.031 sum_up_and_integrate 119 10.3 0.001 0.003 3.586 3.590 integrate_v_rspace 119 11.3 0.002 0.003 3.572 3.576 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.466 3.498 calculate_rho_elec 119 8.7 0.021 0.024 3.466 3.498 qs_ot_p2m_diag 50 11.0 0.008 0.013 2.971 2.989 make_m2s 4572 13.5 0.052 0.055 2.595 2.941 make_images 4572 14.5 0.206 0.245 2.505 2.855 calculate_first_density_matrix 1 7.0 0.000 0.003 2.705 2.708 cp_dbcsr_syevd 50 12.0 0.003 0.005 2.570 2.571 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.286 2.376 ot_diis_step 108 11.5 0.011 0.011 2.298 2.298 cp_fm_diag_elpa 50 13.0 0.000 0.001 2.258 2.259 multiply_cannon_sync_h2d 27432 15.5 1.707 2.255 1.707 2.255 cp_fm_redistribute_end 50 14.0 1.145 2.234 1.148 2.236 init_scf_loop 11 6.9 0.002 0.014 2.215 2.215 cp_fm_diag_elpa_base 50 14.0 1.057 2.150 1.085 2.182 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.167 2.173 calculate_dm_sparse 119 9.5 0.001 0.001 2.059 2.140 acc_transpose_blocks 27432 15.5 0.113 0.120 1.710 2.127 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.951 1.993 grid_integrate_task_list 119 12.3 1.835 1.918 1.835 1.918 density_rs2pw 119 9.7 0.004 0.005 1.767 1.859 transfer_rs2pw 487 10.6 0.005 0.005 1.569 1.707 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.613 1.646 qs_energies_init_hamiltonians 11 5.9 0.002 0.014 1.549 1.552 jit_kernel_multiply 8 16.4 0.718 1.542 0.718 1.542 make_images_data 4572 15.5 0.048 0.054 1.215 1.535 hybrid_alltoall_any 4725 16.4 0.054 0.114 1.082 1.443 wfi_extrapolate 11 7.9 0.004 0.030 1.437 1.437 fft3d_ps 1201 13.6 0.508 0.563 1.331 1.359 potential_pw2rs 119 12.3 0.006 0.006 1.343 1.349 grid_collocate_task_list 119 9.7 1.286 1.333 1.286 1.333 fft_wrap_pw1pw2_140 487 12.2 0.048 0.050 1.293 1.325 mp_alltoall_d11v 2130 13.8 1.195 1.325 1.195 1.325 prepare_preconditioner 11 7.9 0.000 0.000 1.295 1.323 make_preconditioner 11 8.9 0.000 0.002 1.295 1.323 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.263 1.275 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.197 1.245 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.189 1.234 mp_waitany 5720 13.7 1.040 1.208 1.040 1.208 transfer_rs2pw_140 130 11.5 0.137 0.148 1.063 1.200 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.043 1.136 acc_transpose_blocks_kernels 27432 16.5 0.187 0.276 0.759 1.056 mp_sum_d 4137 12.0 0.576 0.996 0.576 0.996 mp_allgather_i34 2286 14.5 0.560 0.989 0.560 0.989 acc_transpose_blocks_sync 82296 16.5 0.812 0.930 0.812 0.930 transfer_pw2rs 487 13.2 0.004 0.005 0.881 0.886 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.862 0.876 build_core_hamiltonian_matrix 11 6.9 0.001 0.004 0.425 0.855 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.434000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.000000, yerr=1.537412 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_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 515.334144E+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 0.067 0.116 32.484 32.484 qs_mol_dyn_low 1 2.0 0.003 0.024 32.112 32.152 qs_forces 11 3.9 0.021 0.022 32.033 32.034 qs_energies 11 4.9 0.043 0.122 30.475 30.486 scf_env_do_scf 11 5.9 0.000 0.001 24.837 24.837 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.630 22.630 dbcsr_multiply_generic 2286 12.5 0.097 0.099 17.216 17.276 qs_scf_new_mos 108 7.5 0.001 0.001 15.204 15.223 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.204 15.222 multiply_cannon 2286 13.5 0.194 0.198 14.102 14.829 velocity_verlet 10 3.0 0.013 0.030 14.713 14.715 ot_scf_mini 108 9.5 0.002 0.002 14.448 14.464 multiply_cannon_loop 2286 14.5 0.859 0.886 13.344 14.082 ot_mini 108 10.5 0.001 0.001 8.946 8.966 qs_ot_get_derivative 108 11.5 0.001 0.001 7.467 7.484 multiply_cannon_multrec 18288 15.5 1.884 2.894 7.064 7.428 rebuild_ks_matrix 119 8.3 0.000 0.000 5.861 5.878 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 5.861 5.877 dbcsr_mm_accdrv_process 38222 16.0 5.025 5.804 5.085 5.871 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.156 5.170 init_scf_run 11 5.9 0.000 0.002 4.087 4.087 scf_env_initial_rho_setup 11 6.9 0.008 0.022 4.086 4.086 mp_waitall_1 158411 16.6 2.527 3.612 2.527 3.612 sum_up_and_integrate 119 10.3 0.001 0.001 3.422 3.428 integrate_v_rspace 119 11.3 0.003 0.003 3.410 3.416 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.807 3.412 qs_ot_get_p 119 10.4 0.001 0.002 3.291 3.317 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.089 3.094 calculate_rho_elec 119 8.7 0.031 0.031 3.088 3.093 calculate_first_density_matrix 1 7.0 0.000 0.000 2.909 2.917 multiply_cannon_metrocomm3 18288 15.5 0.047 0.050 1.425 2.377 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.015 2.372 apply_single 119 13.6 0.000 0.000 2.015 2.372 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.208 2.212 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.206 2.207 init_scf_loop 11 6.9 0.000 0.000 2.191 2.191 acc_transpose_blocks 18288 15.5 0.080 0.083 1.977 2.034 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.934 1.939 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.917 1.917 calculate_dm_sparse 119 9.5 0.000 0.001 1.900 1.906 grid_integrate_task_list 119 12.3 1.799 1.892 1.799 1.892 make_m2s 4572 13.5 0.043 0.044 1.730 1.877 make_images 4572 14.5 0.193 0.205 1.644 1.790 density_rs2pw 119 9.7 0.003 0.004 1.658 1.732 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.681 1.692 cp_fm_diag_elpa_base 50 14.0 1.658 1.674 1.680 1.691 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.671 1.679 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.558 1.566 ot_diis_step 108 11.5 0.011 0.011 1.466 1.466 prepare_preconditioner 11 7.9 0.000 0.000 1.454 1.456 make_preconditioner 11 8.9 0.000 0.000 1.453 1.456 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.337 1.391 mp_sum_l 7287 12.8 0.988 1.354 0.988 1.354 grid_collocate_task_list 119 9.7 1.233 1.320 1.233 1.320 fft_wrap_pw1pw2_140 487 12.2 0.060 0.062 1.264 1.272 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.246 1.258 fft3d_ps 1201 13.6 0.529 0.549 1.245 1.254 potential_pw2rs 119 12.3 0.007 0.007 1.225 1.228 qs_energies_init_hamiltonians 11 5.9 0.059 0.161 1.136 1.226 acc_transpose_blocks_kernels 18288 16.5 0.216 0.224 1.146 1.208 multiply_cannon_sync_h2d 18288 15.5 1.057 1.163 1.057 1.163 wfi_extrapolate 11 7.9 0.001 0.001 1.120 1.120 transfer_rs2pw 487 10.6 0.005 0.005 0.938 1.008 jit_kernel_transpose 5 15.6 0.930 0.992 0.930 0.992 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.932 0.948 make_images_data 4572 15.5 0.047 0.051 0.731 0.899 hybrid_alltoall_any 4725 16.4 0.059 0.116 0.627 0.815 transfer_pw2rs 487 13.2 0.004 0.004 0.788 0.790 acc_transpose_blocks_sync 54864 16.5 0.731 0.780 0.731 0.780 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.773 0.775 mp_alltoall_d11v 2130 13.8 0.662 0.758 0.662 0.758 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.670 0.731 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.484000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=490.545455, yerr=2.061052 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_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 554.999808E+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 0.048 0.117 37.155 37.189 qs_mol_dyn_low 1 2.0 0.021 0.130 36.157 36.199 qs_forces 11 3.9 0.004 0.018 35.942 35.945 qs_energies 11 4.9 0.007 0.026 34.189 34.231 scf_env_do_scf 11 5.9 0.001 0.002 28.700 28.701 scf_env_do_scf_inner_loop 108 6.5 0.005 0.025 25.903 25.903 dbcsr_multiply_generic 2286 12.5 0.121 0.126 20.159 20.285 qs_scf_new_mos 108 7.5 0.001 0.001 17.928 17.993 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.927 17.992 velocity_verlet 10 3.0 0.014 0.058 17.897 17.915 multiply_cannon 2286 13.5 0.217 0.227 16.440 17.166 ot_scf_mini 108 9.5 0.002 0.003 16.887 16.943 multiply_cannon_loop 2286 14.5 1.542 1.624 15.459 16.027 ot_mini 108 10.5 0.001 0.001 10.425 10.492 multiply_cannon_multrec 27432 15.5 2.456 3.149 9.128 9.437 qs_ot_get_derivative 108 11.5 0.001 0.001 8.553 8.612 dbcsr_mm_accdrv_process 47916 15.9 6.029 7.494 6.572 7.942 rebuild_ks_matrix 119 8.3 0.000 0.000 6.311 6.367 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.016 6.311 6.367 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.587 5.638 init_scf_run 11 5.9 0.000 0.001 3.997 3.998 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.997 3.997 qs_ot_get_p 119 10.4 0.001 0.002 3.594 3.665 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.223 3.538 sum_up_and_integrate 119 10.3 0.001 0.002 3.447 3.455 integrate_v_rspace 119 11.3 0.003 0.003 3.435 3.444 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.280 3.322 calculate_rho_elec 119 8.7 0.040 0.047 3.279 3.322 init_scf_loop 11 6.9 0.001 0.004 2.777 2.778 acc_transpose_blocks 27432 15.5 0.118 0.121 2.522 2.677 calculate_first_density_matrix 1 7.0 0.000 0.002 2.607 2.609 make_m2s 4572 13.5 0.054 0.056 2.351 2.484 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.402 2.433 make_images 4572 14.5 0.273 0.332 2.243 2.373 mp_waitall_1 137007 16.6 1.845 2.323 1.845 2.323 calculate_dm_sparse 119 9.5 0.000 0.000 2.185 2.237 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.036 2.233 apply_single 119 13.6 0.000 0.000 2.036 2.233 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.160 2.171 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.027 2.028 prepare_preconditioner 11 7.9 0.000 0.000 1.927 1.936 make_preconditioner 11 8.9 0.000 0.001 1.927 1.935 grid_integrate_task_list 119 12.3 1.824 1.917 1.824 1.917 jit_kernel_multiply 10 16.1 0.476 1.889 0.476 1.889 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.671 1.869 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.849 1.864 ot_diis_step 108 11.5 0.012 0.012 1.831 1.831 density_rs2pw 119 9.7 0.004 0.005 1.713 1.821 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.733 1.765 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.753 1.753 acc_transpose_blocks_sync 82296 16.5 1.444 1.548 1.444 1.548 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.504 1.505 cp_fm_redistribute_end 50 14.0 0.754 1.482 0.758 1.483 fft_wrap_pw1pw2_140 487 12.2 0.070 0.072 1.422 1.456 cp_fm_diag_elpa_base 50 14.0 0.699 1.413 0.723 1.454 mp_sum_l 7287 12.8 1.106 1.441 1.106 1.441 fft3d_ps 1201 13.6 0.573 0.634 1.396 1.422 wfi_extrapolate 11 7.9 0.001 0.002 1.328 1.328 grid_collocate_task_list 119 9.7 1.249 1.327 1.249 1.327 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.826 1.322 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.297 1.312 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.280 1.292 potential_pw2rs 119 12.3 0.008 0.010 1.252 1.254 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.214 1.219 transfer_rs2pw 487 10.6 0.005 0.005 0.996 1.125 make_images_data 4572 15.5 0.048 0.051 0.861 0.998 acc_transpose_blocks_kernels 27432 16.5 0.270 0.275 0.932 0.980 mp_alltoall_d11v 2130 13.8 0.837 0.973 0.837 0.973 hybrid_alltoall_any 4725 16.4 0.066 0.155 0.724 0.924 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.801 0.882 dbcsr_complete_redistribute 329 12.2 0.107 0.125 0.670 0.831 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.813 0.818 transfer_pw2rs 487 13.2 0.004 0.004 0.753 0.755 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.189000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.181818, yerr=3.157046 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_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 626.970624E+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 0.058 0.080 29.164 29.164 qs_mol_dyn_low 1 2.0 0.003 0.010 28.926 28.945 qs_forces 11 3.9 0.008 0.009 28.860 28.862 qs_energies 11 4.9 0.048 0.121 27.165 27.170 scf_env_do_scf 11 5.9 0.000 0.001 21.443 21.443 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.166 19.167 velocity_verlet 10 3.0 0.007 0.015 14.291 14.293 dbcsr_multiply_generic 2286 12.5 0.098 0.102 13.613 13.674 qs_scf_new_mos 108 7.5 0.001 0.001 11.766 11.789 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.766 11.788 multiply_cannon 2286 13.5 0.224 0.233 11.077 11.541 ot_scf_mini 108 9.5 0.002 0.002 11.052 11.075 multiply_cannon_loop 2286 14.5 0.641 0.660 10.161 10.284 multiply_cannon_multrec 9144 15.5 1.863 2.058 6.275 6.503 ot_mini 108 10.5 0.001 0.001 6.394 6.422 rebuild_ks_matrix 119 8.3 0.000 0.000 5.665 5.688 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.665 5.688 qs_ot_get_derivative 108 11.5 0.001 0.001 5.108 5.130 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.038 5.060 dbcsr_mm_accdrv_process 12550 15.8 3.565 4.331 4.304 4.422 init_scf_run 11 5.9 0.001 0.006 3.934 3.934 scf_env_initial_rho_setup 11 6.9 0.002 0.004 3.933 3.934 sum_up_and_integrate 119 10.3 0.001 0.001 3.328 3.332 integrate_v_rspace 119 11.3 0.003 0.003 3.318 3.321 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.172 3.180 calculate_rho_elec 119 8.7 0.060 0.061 3.172 3.179 calculate_first_density_matrix 1 7.0 0.000 0.000 2.823 2.825 qs_ot_get_p 119 10.4 0.001 0.002 2.774 2.811 init_scf_loop 11 6.9 0.000 0.000 2.255 2.256 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.144 2.145 grid_integrate_task_list 119 12.3 1.863 1.948 1.863 1.948 make_m2s 4572 13.5 0.034 0.035 1.775 1.938 acc_transpose_blocks 9144 15.5 0.042 0.043 1.766 1.903 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.856 1.865 make_images 4572 14.5 0.269 0.300 1.685 1.846 mp_waitall_1 115863 16.7 1.338 1.828 1.338 1.828 calculate_dm_sparse 119 9.5 0.000 0.000 1.790 1.806 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.794 1.796 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.728 1.736 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.710 1.727 density_rs2pw 119 9.7 0.003 0.003 1.580 1.657 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.551 1.551 prepare_preconditioner 11 7.9 0.000 0.000 1.532 1.535 make_preconditioner 11 8.9 0.000 0.000 1.532 1.535 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.427 1.451 fft_wrap_pw1pw2_140 487 12.2 0.094 0.095 1.427 1.437 qs_energies_init_hamiltonians 11 5.9 0.069 0.165 1.346 1.424 grid_collocate_task_list 119 9.7 1.296 1.366 1.296 1.366 fft3d_ps 1201 13.6 0.632 0.646 1.340 1.348 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.325 1.337 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.279 1.288 cp_fm_diag_elpa_base 50 14.0 1.253 1.270 1.278 1.286 ot_diis_step 108 11.5 0.013 0.013 1.275 1.275 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.213 1.251 apply_single 119 13.6 0.000 0.000 1.213 1.251 jit_kernel_multiply 6 15.3 0.699 1.222 0.699 1.222 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.177 1.182 potential_pw2rs 119 12.3 0.010 0.010 1.141 1.143 acc_transpose_blocks_kernels 9144 16.5 0.117 0.121 0.994 1.109 wfi_extrapolate 11 7.9 0.001 0.001 1.059 1.059 jit_kernel_transpose 5 15.6 0.876 0.992 0.876 0.992 hybrid_alltoall_any 4725 16.4 0.066 0.177 0.730 0.962 make_images_data 4572 15.5 0.042 0.046 0.764 0.941 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.876 0.927 mp_alltoall_d11v 2130 13.8 0.763 0.897 0.763 0.897 transfer_rs2pw 487 10.6 0.004 0.005 0.771 0.828 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.819 0.827 cp_fm_cholesky_invert 11 10.9 0.753 0.758 0.753 0.758 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.704 0.752 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.747 0.749 acc_transpose_blocks_sync 27432 16.5 0.715 0.736 0.715 0.736 dbcsr_complete_redistribute 329 12.2 0.218 0.279 0.657 0.681 transfer_pw2rs 487 13.2 0.003 0.004 0.651 0.654 mp_allgather_i34 2286 14.5 0.233 0.651 0.233 0.651 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.320 0.629 mp_alltoall_z22v 1201 15.6 0.585 0.615 0.585 0.615 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.164000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=592.545455, yerr=7.969985 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_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 794.386432E+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 0.082 0.108 41.236 41.237 qs_mol_dyn_low 1 2.0 0.003 0.003 40.816 40.972 qs_forces 11 3.9 0.007 0.008 40.733 40.734 qs_energies 11 4.9 0.006 0.039 38.753 38.760 scf_env_do_scf 11 5.9 0.001 0.001 32.192 32.192 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.313 26.315 velocity_verlet 10 3.0 0.002 0.002 21.843 21.848 dbcsr_multiply_generic 2286 12.5 0.143 0.156 19.647 19.760 qs_scf_new_mos 108 7.5 0.001 0.001 17.225 17.328 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.224 17.327 multiply_cannon 2286 13.5 0.297 0.300 15.710 16.509 ot_scf_mini 108 9.5 0.002 0.002 16.073 16.175 multiply_cannon_loop 2286 14.5 0.858 0.876 14.449 15.262 ot_mini 108 10.5 0.001 0.001 9.935 10.049 multiply_cannon_multrec 9144 15.5 3.488 5.081 9.010 9.138 qs_ot_get_derivative 108 11.5 0.001 0.001 7.872 7.975 rebuild_ks_matrix 119 8.3 0.000 0.000 6.913 7.055 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.913 7.054 dbcsr_mm_accdrv_process 12550 15.8 4.622 6.697 5.390 6.787 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.230 6.357 init_scf_loop 11 6.9 0.000 0.000 5.853 5.856 prepare_preconditioner 11 7.9 0.000 0.000 4.912 4.926 make_preconditioner 11 8.9 0.000 0.000 4.912 4.926 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.928 4.813 init_scf_run 11 5.9 0.000 0.001 4.232 4.232 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.232 4.232 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.860 3.868 calculate_rho_elec 119 8.7 0.118 0.122 3.859 3.868 sum_up_and_integrate 119 10.3 0.001 0.001 3.616 3.624 integrate_v_rspace 119 11.3 0.003 0.004 3.606 3.614 qs_ot_get_p 119 10.4 0.002 0.002 3.217 3.357 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.832 3.241 mp_waitall_1 94719 16.7 2.352 3.237 2.352 3.237 cp_fm_upper_to_full 72 14.2 2.098 2.998 2.098 2.998 calculate_first_density_matrix 1 7.0 0.000 0.000 2.814 2.815 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.234 2.571 apply_single 119 13.6 0.000 0.000 2.234 2.571 make_m2s 4572 13.5 0.037 0.037 2.368 2.536 make_images 4572 14.5 0.351 0.382 2.248 2.416 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.414 2.415 fft_wrap_pw1pw2 1201 11.6 0.014 0.014 2.330 2.338 calculate_dm_sparse 119 9.5 0.000 0.000 2.305 2.321 acc_transpose_blocks 9144 15.5 0.044 0.045 2.172 2.276 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.426 2.232 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.148 2.191 dbcsr_complete_redistribute 329 12.2 0.267 0.280 1.524 2.094 grid_integrate_task_list 119 12.3 2.068 2.080 2.068 2.080 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.025 2.078 qs_energies_init_hamiltonians 11 5.9 0.035 0.102 2.032 2.059 ot_diis_step 108 11.5 0.014 0.014 2.040 2.041 fft_wrap_pw1pw2_140 487 12.2 0.179 0.179 1.995 2.004 density_rs2pw 119 9.7 0.003 0.003 1.908 1.922 qs_ot_p2m_diag 50 11.0 0.042 0.043 1.848 1.851 fft3d_ps 1201 13.6 0.856 0.875 1.782 1.788 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.223 1.776 mp_sum_l 7287 12.8 0.983 1.758 0.983 1.758 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 0.979 1.530 mp_alltoall_i22 627 13.8 0.956 1.527 0.956 1.527 grid_collocate_task_list 119 9.7 1.513 1.526 1.513 1.526 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.500 1.500 wfi_extrapolate 11 7.9 0.001 0.001 1.356 1.356 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.311 1.321 hybrid_alltoall_any 4725 16.4 0.091 0.151 1.067 1.319 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.292 1.310 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.166 1.255 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.213 1.214 cp_fm_diag_elpa_base 50 14.0 1.128 1.141 1.212 1.212 make_images_data 4572 15.5 0.046 0.049 0.997 1.201 jit_kernel_multiply 6 15.8 0.738 1.192 0.738 1.192 potential_pw2rs 119 12.3 0.013 0.014 1.163 1.165 mp_alltoall_d11v 2130 13.8 1.129 1.149 1.129 1.149 acc_transpose_blocks_sync 27432 16.5 1.092 1.114 1.092 1.114 acc_transpose_blocks_kernels 9144 16.5 0.120 0.121 1.019 1.103 cp_fm_cholesky_invert 11 10.9 1.078 1.084 1.078 1.084 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.000 1.043 jit_kernel_transpose 5 15.6 0.900 0.982 0.900 0.982 qs_create_task_list 11 7.9 0.024 0.024 0.952 0.961 generate_qs_task_list 11 8.9 0.367 0.387 0.928 0.937 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.893 0.907 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.237000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=744.818182, yerr=13.703471 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_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 497.889280E+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 459019. 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 0.028 0.048 84.231 84.232 qs_mol_dyn_low 1 2.0 0.003 0.004 83.817 83.827 qs_forces 11 3.9 0.003 0.003 83.751 83.752 qs_energies 11 4.9 0.037 0.104 80.891 80.907 scf_env_do_scf 11 5.9 0.001 0.001 71.534 71.534 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 66.348 66.348 dbcsr_multiply_generic 2055 12.4 0.108 0.112 53.304 53.847 qs_scf_new_mos 99 7.5 0.000 0.001 49.694 49.801 qs_scf_loop_do_ot 99 8.5 0.000 0.001 49.694 49.801 ot_scf_mini 99 9.5 0.002 0.002 47.253 47.320 multiply_cannon 2055 13.4 0.174 0.179 43.652 45.149 multiply_cannon_loop 2055 14.4 1.835 1.904 42.727 44.287 velocity_verlet 10 3.0 0.001 0.002 42.192 42.193 ot_mini 99 10.5 0.001 0.001 28.392 28.473 qs_ot_get_derivative 99 11.5 0.001 0.001 21.486 21.585 multiply_cannon_multrec 49320 15.4 11.050 11.768 17.284 18.334 rebuild_ks_matrix 110 8.3 0.000 0.000 14.321 14.586 qs_ks_build_kohn_sham_matrix 110 9.3 0.010 0.012 14.320 14.585 mp_waitall_1 220248 16.4 11.852 13.038 11.852 13.038 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.559 12.793 qs_ot_get_p 110 10.4 0.002 0.016 10.217 10.345 multiply_cannon_sync_h2d 49320 15.4 9.285 9.938 9.285 9.938 multiply_cannon_metrocomm3 49320 15.4 0.081 0.085 7.278 8.796 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.502 8.050 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.203 7.653 apply_single 110 13.6 0.000 0.001 7.202 7.652 init_scf_run 11 5.9 0.000 0.002 7.072 7.072 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.072 7.072 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.857 6.890 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.802 6.865 sum_up_and_integrate 110 10.3 0.002 0.003 6.783 6.798 integrate_v_rspace 110 11.3 0.003 0.003 6.758 6.779 ot_diis_step 99 11.5 0.006 0.006 6.601 6.601 dbcsr_mm_accdrv_process 87628 16.1 3.216 3.422 6.107 6.539 mp_sum_l 6594 12.7 5.389 6.401 5.389 6.401 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.019 6.132 calculate_rho_elec 110 8.6 0.020 0.024 6.018 6.131 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.905 5.905 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.438 5.439 cp_fm_redistribute_end 48 14.0 2.759 5.413 2.769 5.417 cp_fm_diag_elpa_base 48 14.0 2.638 5.293 2.646 5.310 init_scf_loop 11 6.9 0.000 0.000 5.156 5.157 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.996 3.997 calculate_dm_sparse 110 9.5 0.000 0.001 3.827 3.952 wfi_extrapolate 11 7.9 0.001 0.001 3.932 3.932 make_m2s 4110 13.4 0.061 0.066 3.799 3.909 multiply_cannon_metrocomm1 49320 15.4 0.064 0.068 2.808 3.871 make_images 4110 14.4 0.178 0.190 3.703 3.819 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.525 3.603 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.389 3.447 grid_integrate_task_list 110 12.3 3.267 3.438 3.267 3.438 density_rs2pw 110 9.6 0.004 0.004 3.036 3.315 acc_transpose_blocks 49320 15.4 0.213 0.227 2.934 3.090 calculate_first_density_matrix 1 7.0 0.002 0.013 3.039 3.042 prepare_preconditioner 11 7.9 0.000 0.000 2.991 3.016 make_preconditioner 11 8.9 0.000 0.000 2.991 3.016 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.745 2.791 mp_alltoall_d11v 2046 13.8 2.141 2.757 2.141 2.757 jit_kernel_multiply 13 15.9 2.606 2.692 2.606 2.692 fft_wrap_pw1pw2 1111 11.6 0.013 0.018 2.550 2.613 potential_pw2rs 110 12.3 0.006 0.006 2.395 2.410 grid_collocate_task_list 110 9.6 2.159 2.269 2.159 2.269 fft_wrap_pw1pw2_140 451 12.1 0.092 0.098 2.170 2.231 fft3d_ps 1111 13.6 0.724 0.806 2.090 2.150 transfer_rs2pw 451 10.6 0.006 0.006 1.808 2.082 mp_sum_d 3891 11.9 1.565 2.081 1.565 2.081 mp_waitany 14300 13.8 1.801 2.040 1.801 2.040 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.900 1.935 make_images_data 4110 15.4 0.044 0.047 1.728 1.876 hybrid_alltoall_any 4261 16.3 0.084 0.481 1.503 1.810 transfer_pw2rs 451 13.1 0.006 0.007 1.702 1.709 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.669 1.703 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.232000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=472.636364, yerr=2.384523 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 591.147008E+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 2963866. 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 0.037 0.066 68.450 68.462 qs_mol_dyn_low 1 2.0 0.030 0.209 68.045 68.055 qs_forces 11 3.9 0.026 0.057 67.940 67.981 qs_energies 11 4.9 0.001 0.001 64.599 64.663 scf_env_do_scf 11 5.9 0.000 0.001 55.656 55.657 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.770 48.771 dbcsr_multiply_generic 2055 12.4 0.116 0.121 38.215 38.430 velocity_verlet 10 3.0 0.002 0.005 35.125 35.127 qs_scf_new_mos 99 7.5 0.001 0.001 33.140 33.275 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.140 33.274 multiply_cannon 2055 13.4 0.223 0.246 31.585 32.963 ot_scf_mini 99 9.5 0.003 0.003 31.474 31.598 multiply_cannon_loop 2055 14.4 1.159 1.178 30.321 31.272 ot_mini 99 10.5 0.001 0.001 18.412 18.546 multiply_cannon_multrec 24660 15.4 6.953 8.737 14.075 15.711 rebuild_ks_matrix 110 8.3 0.000 0.001 13.496 13.682 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.495 13.681 qs_ot_get_derivative 99 11.5 0.001 0.001 12.646 12.769 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.876 12.040 mp_waitall_1 176588 16.5 7.519 10.063 7.519 10.063 multiply_cannon_metrocomm3 24660 15.4 0.072 0.076 5.158 7.869 multiply_cannon_sync_h2d 24660 15.4 6.340 7.363 6.340 7.363 dbcsr_mm_accdrv_process 52282 16.1 5.504 6.284 6.952 7.356 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.456 7.041 apply_single 110 13.6 0.000 0.001 6.455 7.041 init_scf_loop 11 6.9 0.000 0.000 6.849 6.850 qs_ot_get_p 110 10.4 0.001 0.001 6.440 6.616 init_scf_run 11 5.9 0.000 0.001 6.281 6.281 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.280 6.281 sum_up_and_integrate 110 10.3 0.001 0.003 6.230 6.243 integrate_v_rspace 110 11.3 0.002 0.003 6.203 6.218 ot_diis_step 99 11.5 0.010 0.010 5.719 5.719 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.653 5.661 calculate_rho_elec 110 8.6 0.039 0.047 5.652 5.661 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.769 5.463 prepare_preconditioner 11 7.9 0.000 0.000 4.840 4.862 make_preconditioner 11 8.9 0.000 0.000 4.840 4.862 make_m2s 4110 13.4 0.057 0.061 4.188 4.647 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.535 4.556 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.452 4.554 make_images 4110 14.4 0.404 0.463 4.077 4.531 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.076 4.076 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.549 3.549 cp_fm_redistribute_end 48 14.0 1.790 3.523 1.795 3.525 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.482 3.484 cp_fm_diag_elpa_base 48 14.0 1.669 3.384 1.725 3.465 wfi_extrapolate 11 7.9 0.001 0.001 3.423 3.423 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.283 3.359 grid_integrate_task_list 110 12.3 3.155 3.294 3.155 3.294 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.146 3.211 fft_wrap_pw1pw2 1111 11.6 0.014 0.017 3.071 3.210 density_rs2pw 110 9.6 0.004 0.004 2.906 3.118 calculate_dm_sparse 110 9.5 0.001 0.001 3.025 3.051 calculate_first_density_matrix 1 7.0 0.000 0.000 2.763 2.765 fft_wrap_pw1pw2_140 451 12.1 0.109 0.114 2.616 2.757 make_images_data 4110 15.4 0.049 0.054 2.268 2.752 hybrid_alltoall_any 4261 16.3 0.106 0.457 1.985 2.743 fft3d_ps 1111 13.6 1.067 1.272 2.451 2.577 mp_sum_l 6594 12.7 1.906 2.525 1.906 2.525 cp_fm_cholesky_invert 11 10.9 2.361 2.376 2.361 2.376 grid_collocate_task_list 110 9.6 2.166 2.304 2.166 2.304 acc_transpose_blocks 24660 15.4 0.116 0.118 2.191 2.267 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.179 2.240 potential_pw2rs 110 12.3 0.008 0.008 2.113 2.121 jit_kernel_multiply 10 16.4 1.076 2.029 1.076 2.029 mp_alltoall_d11v 2046 13.8 1.816 1.978 1.816 1.978 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.950 1.970 transfer_rs2pw 451 10.6 0.007 0.007 1.640 1.860 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.779 1.800 multiply_cannon_metrocomm4 22605 15.4 0.079 0.083 0.797 1.681 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.594 1.608 mp_allgather_i34 2055 14.4 0.574 1.574 0.574 1.574 mp_irecv_dv 57340 16.2 0.662 1.556 0.662 1.556 mp_waitany 10164 13.8 1.362 1.543 1.362 1.543 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.389 1.493 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.462000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=558.545455, yerr=7.152414 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 665.788416E+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 0.027 0.045 59.716 59.717 qs_mol_dyn_low 1 2.0 0.003 0.004 59.254 59.265 qs_forces 11 3.9 0.003 0.003 59.156 59.156 qs_energies 11 4.9 0.001 0.001 55.983 55.986 scf_env_do_scf 11 5.9 0.000 0.001 48.003 48.003 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.338 40.338 velocity_verlet 10 3.0 0.008 0.042 31.712 31.713 dbcsr_multiply_generic 2055 12.4 0.115 0.119 29.278 29.546 qs_scf_new_mos 99 7.5 0.001 0.001 25.775 25.883 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.774 25.883 ot_scf_mini 99 9.5 0.002 0.002 24.520 24.645 multiply_cannon 2055 13.4 0.212 0.225 22.856 23.992 multiply_cannon_loop 2055 14.4 0.816 0.841 21.752 22.763 ot_mini 99 10.5 0.001 0.001 14.094 14.222 rebuild_ks_matrix 110 8.3 0.000 0.000 12.217 12.373 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.217 12.372 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.773 10.914 multiply_cannon_multrec 16440 15.4 3.524 4.478 9.895 10.696 mp_waitall_1 139946 16.5 7.253 10.510 7.253 10.510 qs_ot_get_derivative 99 11.5 0.001 0.001 9.621 9.749 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.671 7.738 init_scf_loop 11 6.9 0.000 0.000 7.628 7.630 dbcsr_mm_accdrv_process 34862 16.1 5.475 5.876 6.218 6.361 sum_up_and_integrate 110 10.3 0.001 0.002 6.225 6.240 integrate_v_rspace 110 11.3 0.003 0.003 6.199 6.214 prepare_preconditioner 11 7.9 0.000 0.000 5.793 5.816 make_preconditioner 11 8.9 0.000 0.000 5.793 5.816 qs_ot_get_p 110 10.4 0.010 0.037 5.624 5.778 init_scf_run 11 5.9 0.000 0.001 5.587 5.587 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.586 5.587 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.574 5.586 calculate_rho_elec 110 8.6 0.058 0.058 5.573 5.586 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.976 5.474 apply_single 110 13.6 0.000 0.000 4.975 5.474 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.224 5.467 make_m2s 4110 13.4 0.049 0.050 4.203 4.524 ot_diis_step 99 11.5 0.011 0.011 4.435 4.435 make_images 4110 14.4 0.397 0.520 4.086 4.408 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.336 3.998 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.898 3.901 multiply_cannon_sync_h2d 16440 15.4 3.189 3.874 3.189 3.874 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.522 3.523 grid_integrate_task_list 110 12.3 3.199 3.362 3.199 3.362 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.121 3.128 density_rs2pw 110 9.6 0.004 0.004 2.809 3.011 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.944 2.953 cp_fm_diag_elpa_base 48 14.0 2.902 2.913 2.942 2.951 wfi_extrapolate 11 7.9 0.001 0.001 2.874 2.874 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.770 2.771 calculate_dm_sparse 110 9.5 0.001 0.001 2.723 2.749 make_images_data 4110 15.4 0.046 0.049 2.336 2.720 fft_wrap_pw1pw2_140 451 12.1 0.126 0.130 2.663 2.675 hybrid_alltoall_any 4261 16.3 0.109 0.384 2.048 2.647 calculate_first_density_matrix 1 7.0 0.000 0.000 2.612 2.613 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.505 2.578 multiply_cannon_metrocomm4 14385 15.4 0.048 0.053 0.921 2.512 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.433 2.491 cp_fm_cholesky_invert 11 10.9 2.431 2.444 2.431 2.444 fft3d_ps 1111 13.6 1.085 1.101 2.434 2.443 grid_collocate_task_list 110 9.6 2.219 2.391 2.219 2.391 mp_irecv_dv 48980 15.7 0.843 2.370 0.843 2.370 mp_sum_l 6594 12.7 1.584 2.186 1.584 2.186 mp_alltoall_d11v 2046 13.8 1.894 2.173 1.894 2.173 potential_pw2rs 110 12.3 0.009 0.010 2.066 2.073 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.982 1.982 dbcsr_complete_redistribute 325 12.2 0.307 0.345 1.414 1.716 acc_transpose_blocks 16440 15.4 0.078 0.080 1.644 1.681 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.657 1.671 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.644 1.660 transfer_rs2pw 451 10.6 0.005 0.007 1.329 1.547 mp_allgather_i34 2055 14.4 0.417 1.491 0.417 1.491 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.364 1.476 cp_fm_cholesky_decompose 22 10.9 1.348 1.374 1.348 1.374 mp_waitany 17072 13.8 1.147 1.336 1.147 1.336 jit_kernel_multiply 6 16.8 0.328 1.324 0.328 1.324 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.295 1.306 transfer_pw2rs 451 13.1 0.005 0.005 1.208 1.211 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.717000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=629.818182, yerr=8.747136 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_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 748.474368E+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 0.053 0.071 63.102 63.103 qs_mol_dyn_low 1 2.0 0.003 0.004 62.784 62.794 qs_forces 11 3.9 0.003 0.003 62.721 62.721 qs_energies 11 4.9 0.001 0.002 59.328 59.331 scf_env_do_scf 11 5.9 0.001 0.001 50.908 50.913 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.563 41.564 velocity_verlet 10 3.0 0.002 0.002 34.747 34.749 dbcsr_multiply_generic 2055 12.4 0.121 0.123 30.931 31.145 qs_scf_new_mos 99 7.5 0.001 0.001 27.258 27.365 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.258 27.365 ot_scf_mini 99 9.5 0.002 0.003 25.571 25.673 multiply_cannon 2055 13.4 0.239 0.261 23.819 24.826 multiply_cannon_loop 2055 14.4 1.371 1.443 22.284 22.955 ot_mini 99 10.5 0.001 0.001 14.808 14.940 multiply_cannon_multrec 24660 15.4 4.097 6.665 13.137 14.305 rebuild_ks_matrix 110 8.3 0.000 0.000 11.962 12.068 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.962 12.068 qs_ot_get_derivative 99 11.5 0.001 0.001 10.648 10.756 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.573 10.666 dbcsr_mm_accdrv_process 52304 16.0 7.737 9.204 8.885 9.848 init_scf_loop 11 6.9 0.000 0.000 9.305 9.305 prepare_preconditioner 11 7.9 0.000 0.000 7.483 7.497 make_preconditioner 11 8.9 0.000 0.000 7.483 7.497 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.350 7.144 mp_waitall_1 121746 16.5 4.228 6.249 4.228 6.249 sum_up_and_integrate 110 10.3 0.001 0.002 6.080 6.092 integrate_v_rspace 110 11.3 0.003 0.003 6.053 6.065 make_m2s 4110 13.4 0.059 0.061 5.433 5.914 qs_ot_get_p 110 10.4 0.017 0.031 5.676 5.839 make_images 4110 14.4 0.576 0.694 5.292 5.766 init_scf_run 11 5.9 0.000 0.001 5.722 5.723 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.722 5.722 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.552 5.568 calculate_rho_elec 110 8.6 0.077 0.081 5.551 5.568 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.043 4.139 apply_single 110 13.6 0.000 0.000 4.043 4.139 ot_diis_step 99 11.5 0.011 0.011 4.121 4.121 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.765 3.780 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.550 3.604 grid_integrate_task_list 110 12.3 3.303 3.417 3.303 3.417 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.312 3.376 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.207 3.207 hybrid_alltoall_any 4261 16.3 0.122 0.461 2.364 3.199 make_images_data 4110 15.4 0.048 0.052 2.669 3.185 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.146 3.167 calculate_dm_sparse 110 9.5 0.001 0.001 3.103 3.148 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.462 3.054 wfi_extrapolate 11 7.9 0.001 0.001 2.943 2.943 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.931 2.932 cp_fm_upper_to_full 70 14.2 2.049 2.885 2.049 2.885 density_rs2pw 110 9.6 0.004 0.004 2.670 2.802 dbcsr_complete_redistribute 325 12.2 0.377 0.438 2.045 2.728 fft_wrap_pw1pw2_140 451 12.1 0.144 0.147 2.693 2.720 calculate_first_density_matrix 1 7.0 0.000 0.000 2.680 2.683 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.663 2.674 cp_fm_diag_elpa_base 48 14.0 2.598 2.622 2.659 2.669 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.545 2.579 acc_transpose_blocks 24660 15.4 0.110 0.113 2.402 2.568 multiply_cannon_sync_h2d 24660 15.4 2.394 2.531 2.394 2.531 fft3d_ps 1111 13.6 1.127 1.166 2.416 2.429 cp_fm_cholesky_invert 11 10.9 2.411 2.421 2.411 2.421 grid_collocate_task_list 110 9.6 2.265 2.410 2.265 2.410 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.222 2.222 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.527 2.191 mp_alltoall_d11v 2046 13.8 1.869 2.022 1.869 2.022 potential_pw2rs 110 12.3 0.012 0.013 1.888 1.892 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.778 1.811 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.123 1.770 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.615 1.714 mp_alltoall_i22 605 13.7 1.006 1.712 1.006 1.712 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.665 1.686 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.634 1.645 multiply_cannon_metrocomm4 20550 15.4 0.059 0.063 0.858 1.539 mp_sum_l 6594 12.7 0.976 1.504 0.976 1.504 acc_transpose_blocks_sync 73980 16.4 1.345 1.454 1.345 1.454 mp_irecv_dv 62702 16.1 0.752 1.454 0.752 1.454 cp_fm_cholesky_decompose 22 10.9 1.359 1.404 1.359 1.404 mp_allgather_i34 2055 14.4 0.623 1.364 0.623 1.364 transfer_rs2pw 451 10.6 0.005 0.006 1.150 1.304 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=63.103000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=705.454545, yerr=11.880952 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_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 865.898496E+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 0.035 0.067 54.071 54.072 qs_mol_dyn_low 1 2.0 0.003 0.003 53.792 53.802 qs_forces 11 3.9 0.003 0.003 53.690 53.691 qs_energies 11 4.9 0.001 0.001 50.064 50.068 scf_env_do_scf 11 5.9 0.000 0.001 41.676 41.676 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.614 34.615 velocity_verlet 10 3.0 0.025 0.046 30.047 30.050 dbcsr_multiply_generic 2055 12.4 0.110 0.111 23.525 23.686 qs_scf_new_mos 99 7.5 0.001 0.001 20.521 20.599 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.521 20.598 ot_scf_mini 99 9.5 0.002 0.002 19.271 19.340 multiply_cannon 2055 13.4 0.239 0.253 17.971 19.085 multiply_cannon_loop 2055 14.4 0.604 0.625 16.731 17.043 rebuild_ks_matrix 110 8.3 0.000 0.000 11.442 11.509 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.442 11.509 ot_mini 99 10.5 0.001 0.001 10.774 10.845 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.180 10.242 multiply_cannon_multrec 8220 15.4 3.251 4.606 7.822 8.925 mp_waitall_1 103326 16.6 5.927 7.639 5.927 7.639 qs_ot_get_derivative 99 11.5 0.001 0.001 7.053 7.123 init_scf_loop 11 6.9 0.000 0.000 7.015 7.016 sum_up_and_integrate 110 10.3 0.001 0.002 6.009 6.021 integrate_v_rspace 110 11.3 0.003 0.003 5.982 5.994 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.628 5.640 calculate_rho_elec 110 8.6 0.115 0.116 5.627 5.640 dbcsr_mm_accdrv_process 17442 15.9 3.250 4.487 4.429 5.542 prepare_preconditioner 11 7.9 0.000 0.000 5.395 5.399 make_preconditioner 11 8.9 0.000 0.000 5.395 5.399 init_scf_run 11 5.9 0.000 0.001 5.252 5.252 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.251 5.251 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.999 5.067 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 3.106 4.633 qs_ot_get_p 110 10.4 0.002 0.002 4.559 4.629 make_m2s 4110 13.4 0.038 0.039 4.215 4.473 make_images 4110 14.4 0.647 0.707 4.085 4.343 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.797 3.856 apply_single 110 13.6 0.000 0.000 3.797 3.856 ot_diis_step 99 11.5 0.012 0.012 3.700 3.700 grid_integrate_task_list 110 12.3 3.396 3.496 3.396 3.496 fft_wrap_pw1pw2 1111 11.6 0.014 0.015 3.276 3.290 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.222 3.225 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.887 2.889 fft_wrap_pw1pw2_140 451 12.1 0.191 0.193 2.834 2.843 density_rs2pw 110 9.6 0.004 0.004 2.633 2.737 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.721 2.722 calculate_dm_sparse 110 9.5 0.001 0.001 2.639 2.685 calculate_first_density_matrix 1 7.0 0.000 0.000 2.660 2.662 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.655 2.656 hybrid_alltoall_any 4261 16.3 0.200 0.849 2.212 2.620 make_images_data 4110 15.4 0.041 0.047 2.225 2.541 grid_collocate_task_list 110 9.6 2.368 2.533 2.368 2.533 cp_fm_cholesky_invert 11 10.9 2.522 2.530 2.522 2.530 wfi_extrapolate 11 7.9 0.001 0.001 2.498 2.498 multiply_cannon_sync_h2d 8220 15.4 2.359 2.480 2.359 2.480 fft3d_ps 1111 13.6 1.265 1.273 2.435 2.454 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.333 2.340 cp_fm_diag_elpa_base 48 14.0 2.279 2.302 2.331 2.339 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.290 2.302 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.029 2.120 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.791 1.995 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.892 1.927 mp_alltoall_d11v 2046 13.8 1.682 1.813 1.682 1.813 potential_pw2rs 110 12.3 0.015 0.015 1.806 1.809 qs_env_update_s_mstruct 11 6.9 0.022 0.063 1.518 1.666 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.595 1.607 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.591 1.596 jit_kernel_multiply 8 15.5 0.857 1.547 0.857 1.547 dbcsr_complete_redistribute 325 12.2 0.525 0.542 1.429 1.528 mp_allgather_i34 2055 14.4 0.411 1.513 0.411 1.513 cp_fm_cholesky_decompose 22 10.9 1.451 1.463 1.451 1.463 acc_transpose_blocks 8220 15.4 0.039 0.041 1.365 1.391 qs_create_task_list 11 7.9 0.016 0.023 1.227 1.330 generate_qs_task_list 11 8.9 0.374 0.441 1.211 1.315 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.473 1.276 mp_irecv_dv 24056 15.7 0.446 1.228 0.446 1.228 transfer_rs2pw 451 10.6 0.005 0.005 1.086 1.218 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.177 1.199 mp_waitany 9240 13.8 1.039 1.161 1.039 1.161 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.787 1.114 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.072000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=812.909091, yerr=12.915491 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_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 1.359061E+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 0.069 0.105 78.680 78.681 qs_mol_dyn_low 1 2.0 0.003 0.004 78.168 78.177 qs_forces 11 3.9 0.003 0.004 77.665 77.666 qs_energies 11 4.9 0.001 0.002 73.580 73.581 scf_env_do_scf 11 5.9 0.001 0.001 63.242 63.242 velocity_verlet 10 3.0 0.017 0.051 48.176 48.232 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.963 42.966 dbcsr_multiply_generic 2055 12.4 0.138 0.191 29.698 29.795 qs_scf_new_mos 99 7.5 0.001 0.001 26.158 26.211 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.157 26.210 ot_scf_mini 99 9.5 0.002 0.002 24.351 24.395 multiply_cannon 2055 13.4 0.334 0.352 22.743 23.404 multiply_cannon_loop 2055 14.4 0.834 0.860 20.990 21.285 init_scf_loop 11 6.9 0.000 0.000 20.208 20.213 prepare_preconditioner 11 7.9 0.000 0.000 18.227 18.234 make_preconditioner 11 8.9 0.000 0.000 18.227 18.234 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.349 17.741 ot_mini 99 10.5 0.001 0.001 13.962 14.007 rebuild_ks_matrix 110 8.3 0.000 0.000 13.357 13.401 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.356 13.401 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.120 12.161 cp_fm_upper_to_full 70 14.2 8.179 11.683 8.179 11.683 multiply_cannon_multrec 8220 15.4 4.049 4.208 9.903 10.020 qs_ot_get_derivative 99 11.5 0.001 0.001 9.360 9.404 mp_waitall_1 84994 16.7 7.471 8.303 7.471 8.303 dbcsr_complete_redistribute 325 12.2 0.951 0.975 5.435 7.539 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.795 6.832 calculate_rho_elec 110 8.6 0.227 0.227 6.795 6.832 sum_up_and_integrate 110 10.3 0.002 0.002 6.539 6.554 integrate_v_rspace 110 11.3 0.004 0.004 6.511 6.526 copy_fm_to_dbcsr 174 11.2 0.001 0.001 4.434 6.521 init_scf_run 11 5.9 0.000 0.001 5.957 5.957 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.957 5.957 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 3.863 5.935 dbcsr_mm_accdrv_process 11614 15.7 3.877 4.159 5.707 5.934 make_m2s 4110 13.4 0.043 0.043 5.220 5.691 mp_alltoall_i22 605 13.7 3.497 5.609 3.497 5.609 make_images 4110 14.4 0.884 0.935 5.031 5.499 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.977 5.289 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.743 5.220 apply_single 110 13.6 0.000 0.000 4.743 5.220 qs_ot_get_p 110 10.4 0.002 0.002 5.116 5.179 ot_diis_step 99 11.5 0.015 0.016 4.581 4.581 fft_wrap_pw1pw2 1111 11.6 0.016 0.016 4.427 4.448 fft_wrap_pw1pw2_140 451 12.1 0.364 0.364 3.880 3.904 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 3.788 3.789 grid_integrate_task_list 110 12.3 3.673 3.740 3.673 3.740 cp_fm_cholesky_invert 11 10.9 3.720 3.729 3.720 3.729 qs_ot_p2m_diag 48 11.0 0.151 0.155 3.570 3.576 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.042 3.498 hybrid_alltoall_any 4261 16.3 0.269 0.583 2.764 3.366 fft3d_ps 1111 13.6 1.830 1.869 3.287 3.310 density_rs2pw 110 9.6 0.004 0.004 3.267 3.286 calculate_dm_sparse 110 9.5 0.001 0.001 3.248 3.276 make_images_data 4110 15.4 0.045 0.047 2.671 3.255 wfi_extrapolate 11 7.9 0.001 0.001 3.189 3.189 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.158 3.161 multiply_cannon_sync_h2d 8220 15.4 3.112 3.133 3.112 3.133 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.055 3.056 grid_collocate_task_list 110 9.6 2.662 2.676 2.662 2.676 calculate_first_density_matrix 1 7.0 0.000 0.000 2.658 2.660 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.618 2.647 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.629 2.644 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.483 2.484 cp_fm_diag_elpa_base 48 14.0 2.187 2.262 2.482 2.482 qs_env_update_s_mstruct 11 6.9 0.033 0.055 2.276 2.353 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.117 2.190 potential_pw2rs 110 12.3 0.021 0.021 2.094 2.097 mp_alltoall_d11v 2046 13.8 2.002 2.048 2.002 2.048 qs_create_task_list 11 7.9 0.034 0.065 1.939 1.982 jit_kernel_multiply 10 15.1 1.631 1.939 1.631 1.939 generate_qs_task_list 11 8.9 0.731 0.784 1.905 1.936 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.838 1.843 acc_transpose_blocks 8220 15.4 0.041 0.041 1.785 1.835 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.739 1.754 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.719 1.747 cp_fm_cholesky_decompose 22 10.9 1.700 1.714 1.700 1.714 qs_ot_get_orbitals 99 10.5 0.000 0.000 1.630 1.637 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=78.681000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1224.181818, yerr=54.378836 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 631.312384E+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 3008934. 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 0.010 0.027 204.544 204.545 qs_mol_dyn_low 1 2.0 0.003 0.004 204.069 204.083 qs_forces 11 3.9 0.005 0.005 203.939 203.939 qs_energies 11 4.9 0.001 0.002 198.325 198.340 scf_env_do_scf 11 5.9 0.001 0.001 181.777 181.794 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 162.912 162.914 dbcsr_multiply_generic 2507 12.6 0.178 0.182 125.856 126.746 qs_scf_new_mos 117 7.6 0.001 0.001 123.775 124.021 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.774 124.020 velocity_verlet 10 3.0 0.005 0.028 121.670 121.671 ot_scf_mini 117 9.6 0.003 0.003 117.106 117.348 multiply_cannon 2507 13.6 0.235 0.241 102.009 104.419 multiply_cannon_loop 2507 14.6 2.412 2.465 99.848 102.301 ot_mini 117 10.6 0.001 0.001 66.281 66.542 multiply_cannon_multrec 60168 15.6 31.910 34.081 41.835 44.068 qs_ot_get_derivative 117 11.6 0.001 0.001 41.274 41.538 rebuild_ks_matrix 128 8.3 0.001 0.001 33.719 34.072 qs_ks_build_kohn_sham_matrix 128 9.3 0.014 0.016 33.719 34.072 mp_waitall_1 267128 16.5 28.926 32.552 28.926 32.552 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.308 30.634 qs_ot_get_p 128 10.4 0.001 0.002 28.551 29.000 multiply_cannon_sync_h2d 60168 15.6 26.414 28.808 26.414 28.808 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.486 25.034 apply_single 128 13.6 0.001 0.001 24.485 25.034 ot_diis_step 117 11.6 0.008 0.008 24.767 24.768 qs_ot_p2m_diag 83 11.4 0.079 0.091 21.879 21.988 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.648 19.886 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.090 19.091 init_scf_loop 11 6.9 0.000 0.000 18.782 18.784 multiply_cannon_metrocomm3 60168 15.6 0.118 0.126 16.154 18.700 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.059 16.090 cp_fm_diag_elpa_base 83 14.4 16.011 16.048 16.053 16.083 make_m2s 5014 13.6 0.105 0.114 13.902 14.243 prepare_preconditioner 11 7.9 0.000 0.000 14.114 14.159 make_preconditioner 11 8.9 0.000 0.000 14.114 14.159 make_images 5014 14.6 0.399 0.418 13.717 14.070 sum_up_and_integrate 128 10.3 0.002 0.004 13.715 13.731 integrate_v_rspace 128 11.3 0.003 0.004 13.657 13.672 make_full_inverse_cholesky 11 9.9 0.000 0.000 13.374 13.519 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.691 12.822 calculate_rho_elec 128 8.7 0.045 0.064 12.690 12.821 init_scf_run 11 5.9 0.000 0.001 12.444 12.445 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.444 12.444 mp_sum_l 7950 12.9 8.687 10.161 8.687 10.161 dbcsr_mm_accdrv_process 124484 16.2 4.753 4.940 9.494 10.101 wfi_extrapolate 11 7.9 0.001 0.001 8.914 8.914 calculate_dm_sparse 128 9.5 0.001 0.001 8.544 8.656 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.226 8.372 multiply_cannon_metrocomm1 60168 15.6 0.095 0.098 6.307 8.253 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.127 8.224 cp_fm_cholesky_invert 11 10.9 8.028 8.046 8.028 8.046 make_images_data 5014 15.6 0.066 0.074 6.783 7.796 grid_integrate_task_list 128 12.3 7.034 7.462 7.034 7.462 hybrid_alltoall_any 5200 16.5 0.293 2.248 5.955 7.259 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.813 6.828 fft_wrap_pw1pw2 1291 11.7 0.017 0.024 6.527 6.802 density_rs2pw 128 9.7 0.005 0.006 6.287 6.724 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.858 5.983 fft_wrap_pw1pw2_140 523 12.2 0.234 0.245 5.636 5.797 mp_alltoall_d11v 2415 14.1 4.526 5.692 4.526 5.692 fft3d_ps 1291 13.7 2.096 2.622 5.343 5.568 grid_collocate_task_list 128 9.7 4.845 5.246 4.845 5.246 mp_sum_d 4481 12.1 3.554 4.383 3.554 4.383 potential_pw2rs 128 12.3 0.009 0.010 4.209 4.224 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.545000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.181818, yerr=7.132744 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 2.200017E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 6024768 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.7 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 849.989632E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2427840 MPI messages size (bytes): total size 4.132588E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.702167E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71532 2339635200 32768 < size <= 131072 729952 56049532928 131072 < size <= 4194304 1387568 1410045313024 4194304 < size <= 16777216 155760 1473828901424 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4050 66982. MP_Allreduce 11185 956. MP_Sync 88 MP_Alltoall 1983 6138524. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.026 183.176 183.176 qs_mol_dyn_low 1 2.0 0.003 0.003 182.699 182.712 qs_forces 11 3.9 0.008 0.012 182.592 182.593 qs_energies 11 4.9 0.001 0.002 175.841 175.853 scf_env_do_scf 11 5.9 0.001 0.001 159.089 159.106 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 132.931 132.933 velocity_verlet 10 3.0 0.015 0.037 113.345 113.347 dbcsr_multiply_generic 2529 12.6 0.189 0.194 98.500 99.888 qs_scf_new_mos 118 7.6 0.001 0.001 94.493 94.997 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.492 94.996 ot_scf_mini 118 9.6 0.004 0.004 89.639 90.177 multiply_cannon 2529 13.6 0.507 0.563 78.291 82.852 multiply_cannon_loop 2529 14.6 1.576 1.641 74.902 77.898 ot_mini 118 10.6 0.001 0.001 50.482 51.044 mp_waitall_1 216598 16.6 23.935 39.019 23.935 39.019 multiply_cannon_multrec 30348 15.6 21.215 25.896 32.209 37.314 rebuild_ks_matrix 129 8.3 0.001 0.001 32.633 33.458 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.019 32.633 33.457 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.419 30.154 qs_ot_get_derivative 118 11.6 0.002 0.002 28.430 28.962 multiply_cannon_metrocomm3 30348 15.6 0.098 0.106 15.624 28.808 init_scf_loop 11 6.9 0.000 0.000 26.063 26.064 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.194 23.356 apply_single 129 13.6 0.001 0.001 22.194 23.355 ot_diis_step 118 11.6 0.014 0.015 21.876 21.878 qs_ot_get_p 129 10.4 0.007 0.029 21.062 21.793 prepare_preconditioner 11 7.9 0.000 0.000 21.641 21.708 make_preconditioner 11 8.9 0.000 0.000 21.641 21.708 multiply_cannon_sync_h2d 30348 15.6 18.254 21.432 18.254 21.432 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.527 20.945 qs_ot_p2m_diag 84 11.4 0.191 0.219 16.047 16.083 make_m2s 5058 13.6 0.088 0.092 14.137 15.849 make_images 5058 14.6 1.166 1.370 13.918 15.632 cp_dbcsr_syevd 84 12.4 0.006 0.006 14.837 14.838 sum_up_and_integrate 129 10.3 0.002 0.004 13.759 13.789 integrate_v_rspace 129 11.3 0.003 0.004 13.699 13.730 qs_rho_update_rho_low 129 7.7 0.001 0.001 12.875 12.912 calculate_rho_elec 129 8.7 0.089 0.107 12.874 12.911 cp_fm_cholesky_invert 11 10.9 11.881 11.916 11.881 11.916 init_scf_run 11 5.9 0.000 0.001 11.851 11.852 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.851 11.852 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 11.300 11.682 cp_fm_diag_elpa 84 13.4 0.000 0.000 11.461 11.492 cp_fm_diag_elpa_base 84 14.4 11.284 11.322 11.450 11.478 dbcsr_mm_accdrv_process 62758 16.2 5.634 6.337 10.431 11.029 multiply_cannon_metrocomm4 27819 15.6 0.106 0.123 3.837 10.898 mp_irecv_dv 70084 16.3 3.623 10.477 3.623 10.477 make_images_data 5058 15.6 0.066 0.076 8.324 10.371 hybrid_alltoall_any 5245 16.5 0.352 1.562 7.116 9.612 wfi_extrapolate 11 7.9 0.001 0.001 8.199 8.199 grid_integrate_task_list 129 12.3 7.211 7.585 7.211 7.585 fft_wrap_pw1pw2 1301 11.7 0.019 0.024 7.427 7.503 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.316 7.225 calculate_dm_sparse 129 9.5 0.001 0.001 6.694 6.871 density_rs2pw 129 9.7 0.006 0.006 6.545 6.867 fft_wrap_pw1pw2_140 527 12.2 0.251 0.261 6.525 6.596 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.486 6.496 mp_sum_l 8016 12.9 4.267 6.241 4.267 6.241 fft3d_ps 1301 13.7 2.787 2.980 5.864 5.918 cp_fm_cholesky_decompose 22 10.9 5.658 5.752 5.658 5.752 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.490 5.714 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.418 5.488 grid_collocate_task_list 129 9.7 5.039 5.427 5.039 5.427 mp_alltoall_d11v 2429 14.1 4.359 5.047 4.359 5.047 mp_allgather_i34 2529 14.6 1.899 4.951 1.899 4.951 potential_pw2rs 129 12.3 0.014 0.015 4.196 4.214 mp_sum_d 4511 12.1 2.705 4.138 2.705 4.138 dbcsr_complete_redistribute 397 12.7 0.741 0.905 3.218 3.895 dbcsr_dot_sd 1331 12.0 0.777 0.857 2.474 3.746 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.631 3.706 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=183.176000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=810.363636, yerr=1.493789 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/17/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 2.950886E+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 4015680 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.9 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.030169E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1051232 MPI messages size (bytes): total size 2.737039E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.603648E+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 284160 37245419520 131072 < size <= 4194304 665408 1004401590272 4194304 < size <= 16777216 66080 937889764224 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4037 67188. MP_Allreduce 11152 997. MP_Sync 87 MP_Alltoall 1724 9394172. 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 0.056 0.080 167.793 167.794 qs_mol_dyn_low 1 2.0 0.008 0.028 167.281 167.297 qs_forces 11 3.9 0.005 0.020 166.970 166.973 qs_energies 11 4.9 0.009 0.041 160.469 160.478 scf_env_do_scf 11 5.9 0.007 0.051 144.610 144.611 scf_env_do_scf_inner_loop 118 6.6 0.006 0.025 116.019 116.021 velocity_verlet 10 3.0 0.023 0.128 105.965 105.978 dbcsr_multiply_generic 2527 12.6 0.191 0.194 82.138 83.244 qs_scf_new_mos 118 7.6 0.001 0.001 79.403 79.802 qs_scf_loop_do_ot 118 8.6 0.001 0.001 79.402 79.801 ot_scf_mini 118 9.6 0.004 0.004 75.148 75.615 multiply_cannon 2527 13.6 0.514 0.535 62.560 67.165 multiply_cannon_loop 2527 14.6 1.137 1.206 59.677 62.141 ot_mini 118 10.6 0.001 0.001 42.541 42.991 mp_waitall_1 171878 16.6 24.643 34.968 24.643 34.968 rebuild_ks_matrix 129 8.3 0.001 0.001 30.293 30.851 qs_ks_build_kohn_sham_matrix 129 9.3 0.019 0.045 30.292 30.850 init_scf_loop 11 6.9 0.001 0.004 28.474 28.480 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.368 27.870 multiply_cannon_multrec 20216 15.6 13.113 16.819 22.614 26.347 multiply_cannon_metrocomm3 20216 15.6 0.061 0.063 15.497 25.021 prepare_preconditioner 11 7.9 0.000 0.000 24.449 24.503 make_preconditioner 11 8.9 0.000 0.001 24.449 24.503 make_full_inverse_cholesky 11 9.9 0.000 0.000 22.519 23.469 qs_ot_get_derivative 118 11.6 0.001 0.002 22.735 23.200 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.896 21.100 apply_single 129 13.6 0.001 0.001 19.896 21.100 ot_diis_step 118 11.6 0.018 0.018 19.697 19.697 qs_ot_get_p 129 10.4 0.002 0.002 17.936 18.511 multiply_cannon_sync_h2d 20216 15.6 13.698 16.073 13.698 16.073 make_m2s 5054 13.6 0.079 0.085 14.568 15.727 make_images 5054 14.6 1.149 1.244 14.331 15.487 qs_ot_p2m_diag 83 11.4 0.265 0.272 13.762 13.770 sum_up_and_integrate 129 10.3 0.006 0.030 13.626 13.669 integrate_v_rspace 129 11.3 0.003 0.004 13.563 13.595 qs_rho_update_rho_low 129 7.7 0.001 0.001 12.925 12.963 calculate_rho_elec 129 8.7 0.133 0.147 12.925 12.962 cp_dbcsr_syevd 83 12.4 0.006 0.008 12.718 12.718 cp_fm_cholesky_invert 11 10.9 11.697 11.717 11.697 11.717 init_scf_run 11 5.9 0.000 0.001 10.752 10.753 scf_env_initial_rho_setup 11 6.9 0.001 0.003 10.752 10.753 make_images_data 5054 15.6 0.063 0.072 8.704 10.202 multiply_cannon_metrocomm4 17689 15.6 0.068 0.078 3.512 9.721 hybrid_alltoall_any 5240 16.5 0.455 2.078 7.544 9.700 mp_irecv_dv 50625 16.2 3.377 9.449 3.377 9.449 cp_fm_diag_elpa 83 13.4 0.001 0.001 9.398 9.418 cp_fm_diag_elpa_base 83 14.4 9.154 9.219 9.392 9.412 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.886 9.234 dbcsr_mm_accdrv_process 41830 16.2 5.793 6.071 8.949 9.078 grid_integrate_task_list 129 12.3 7.371 7.777 7.371 7.777 fft_wrap_pw1pw2 1301 11.7 0.019 0.023 7.436 7.545 wfi_extrapolate 11 7.9 0.005 0.023 7.291 7.291 fft_wrap_pw1pw2_140 527 12.2 0.287 0.293 6.579 6.696 density_rs2pw 129 9.7 0.006 0.007 6.299 6.690 calculate_dm_sparse 129 9.5 0.009 0.062 5.955 6.066 cp_fm_cholesky_decompose 22 10.9 5.854 5.911 5.854 5.911 fft3d_ps 1301 13.7 2.772 2.995 5.733 5.814 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.609 5.620 dbcsr_complete_redistribute 395 12.7 1.029 1.104 4.411 5.582 grid_collocate_task_list 129 9.7 5.203 5.501 5.203 5.501 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.715 5.375 mp_alltoall_d11v 2423 14.1 4.589 5.096 4.589 5.096 cp_fm_upper_to_full 105 14.8 3.907 4.858 3.907 4.858 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.680 4.822 mp_allgather_i34 2527 14.6 1.370 4.560 1.370 4.560 mp_sum_l 8010 12.9 3.172 4.538 3.172 4.538 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.252 4.420 potential_pw2rs 129 12.3 0.020 0.023 4.029 4.040 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.980 4.006 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.756 3.758 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.440 3.496 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=167.794000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=978.272727, yerr=4.114076 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/18/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 4.353791E+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 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.103122E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+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 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4009 67645. MP_Allreduce 11082 1082. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.032 176.239 176.239 qs_mol_dyn_low 1 2.0 0.003 0.003 175.845 175.882 qs_forces 11 3.9 0.004 0.004 175.260 175.265 qs_energies 11 4.9 0.011 0.038 168.176 168.186 scf_env_do_scf 11 5.9 0.001 0.001 151.494 151.509 velocity_verlet 10 3.0 0.002 0.002 115.025 115.047 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 114.591 114.592 dbcsr_multiply_generic 2507 12.6 0.196 0.201 80.526 81.373 qs_scf_new_mos 117 7.6 0.001 0.001 79.052 79.382 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.051 79.382 ot_scf_mini 117 9.6 0.003 0.003 74.496 74.841 multiply_cannon 2507 13.6 0.551 0.592 55.806 58.557 multiply_cannon_loop 2507 14.6 1.849 1.954 52.320 54.263 ot_mini 117 10.6 0.001 0.001 42.714 43.062 init_scf_loop 11 6.9 0.000 0.000 36.778 36.779 prepare_preconditioner 11 7.9 0.000 0.000 32.692 32.718 make_preconditioner 11 8.9 0.000 0.000 32.692 32.718 make_full_inverse_cholesky 11 9.9 0.017 0.023 27.916 31.476 multiply_cannon_multrec 30084 15.6 13.512 18.740 26.553 31.418 rebuild_ks_matrix 128 8.3 0.001 0.001 29.173 29.471 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 29.172 29.471 mp_waitall_1 147882 16.7 17.070 27.046 17.070 27.046 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.322 26.584 qs_ot_get_derivative 117 11.6 0.001 0.002 22.963 23.313 make_m2s 5014 13.6 0.093 0.098 20.346 21.342 make_images 5014 14.6 1.989 2.269 20.037 21.031 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.203 19.895 apply_single 128 13.6 0.001 0.001 19.203 19.895 ot_diis_step 117 11.6 0.018 0.019 19.628 19.629 qs_ot_get_p 128 10.4 0.001 0.002 17.874 18.255 multiply_cannon_metrocomm3 30084 15.6 0.049 0.052 6.397 15.065 qs_ot_p2m_diag 83 11.4 0.343 0.392 13.802 13.854 sum_up_and_integrate 128 10.3 0.002 0.003 13.573 13.604 integrate_v_rspace 128 11.3 0.003 0.004 13.513 13.545 dbcsr_mm_accdrv_process 62264 16.2 8.471 9.271 12.611 13.130 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.053 13.117 calculate_rho_elec 128 8.7 0.174 0.189 13.052 13.117 cp_fm_upper_to_full 105 14.8 8.746 12.424 8.746 12.424 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.399 12.401 make_images_data 5014 15.6 0.065 0.070 10.849 12.146 cp_fm_cholesky_invert 11 10.9 11.948 11.970 11.948 11.970 hybrid_alltoall_any 5200 16.5 0.531 2.215 9.866 11.452 multiply_cannon_sync_h2d 30084 15.6 10.393 11.189 10.393 11.189 init_scf_run 11 5.9 0.000 0.002 10.681 10.682 scf_env_initial_rho_setup 11 6.9 0.002 0.002 10.681 10.682 dbcsr_complete_redistribute 395 12.7 1.416 1.499 7.565 10.380 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.672 9.919 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.187 9.211 cp_fm_diag_elpa_base 83 14.4 8.706 8.831 9.180 9.197 copy_fm_to_dbcsr 209 11.7 0.001 0.001 6.195 8.970 grid_integrate_task_list 128 12.3 7.518 7.877 7.518 7.877 fft_wrap_pw1pw2 1291 11.7 0.018 0.023 7.694 7.753 multiply_cannon_metrocomm4 25070 15.6 0.080 0.092 2.855 7.466 transfer_fm_to_dbcsr 11 9.9 0.002 0.006 4.753 7.465 wfi_extrapolate 11 7.9 0.001 0.001 7.440 7.440 mp_irecv_dv 76098 16.2 2.696 7.176 2.696 7.176 fft_wrap_pw1pw2_140 523 12.2 0.322 0.327 6.814 6.908 mp_alltoall_i22 716 14.1 4.100 6.905 4.100 6.905 density_rs2pw 128 9.7 0.005 0.005 6.156 6.496 calculate_dm_sparse 128 9.5 0.001 0.001 6.280 6.365 cp_fm_cholesky_decompose 22 10.9 5.949 5.998 5.949 5.998 fft3d_ps 1291 13.7 2.918 2.950 5.895 5.931 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 5.609 5.651 grid_collocate_task_list 128 9.7 5.285 5.622 5.285 5.622 mp_alltoall_d11v 2415 14.1 4.901 5.395 4.901 5.395 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.493 4.585 qs_energies_init_hamiltonians 11 5.9 0.027 0.065 4.499 4.525 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.417 4.521 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.246 4.308 potential_pw2rs 128 12.3 0.021 0.022 3.882 3.892 dbcsr_special_finalize 12535 15.6 0.066 0.070 3.529 3.701 dbcsr_merge_single_wm 12535 16.6 0.322 0.346 3.437 3.607 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.499 3.531 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=176.239000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1041.363636, yerr=17.374342 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.446048E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532237E+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 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 755071. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3999 67810. MP_Allreduce 11057 1167. MP_Sync 87 MP_Alltoall 1712 18838186. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.183 0.192 162.606 162.606 qs_mol_dyn_low 1 2.0 0.003 0.003 162.028 162.058 qs_forces 11 3.9 0.004 0.004 161.937 161.940 qs_energies 11 4.9 0.001 0.002 154.528 154.534 scf_env_do_scf 11 5.9 0.001 0.001 137.270 137.279 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 106.694 106.695 velocity_verlet 10 3.0 0.003 0.011 105.552 105.556 dbcsr_multiply_generic 2507 12.6 0.190 0.196 71.601 71.947 qs_scf_new_mos 117 7.6 0.001 0.001 70.575 70.711 qs_scf_loop_do_ot 117 8.6 0.001 0.001 70.574 70.711 ot_scf_mini 117 9.6 0.003 0.003 66.162 66.284 multiply_cannon 2507 13.6 0.560 0.594 53.172 56.525 multiply_cannon_loop 2507 14.6 0.806 0.840 50.295 51.013 ot_mini 117 10.6 0.001 0.001 37.120 37.238 mp_waitall_1 125778 16.7 24.402 30.559 24.402 30.559 init_scf_loop 11 6.9 0.000 0.000 30.421 30.423 rebuild_ks_matrix 128 8.3 0.001 0.001 28.994 29.096 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 28.994 29.095 prepare_preconditioner 11 7.9 0.000 0.000 26.571 26.593 make_preconditioner 11 8.9 0.000 0.000 26.571 26.593 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.438 26.536 make_full_inverse_cholesky 11 9.9 0.021 0.027 24.678 24.951 multiply_cannon_multrec 10028 15.6 10.380 14.456 18.110 21.193 qs_ot_get_derivative 117 11.6 0.002 0.002 20.198 20.321 multiply_cannon_metrocomm3 10028 15.6 0.025 0.027 12.457 19.494 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 16.910 17.007 apply_single 128 13.6 0.001 0.001 16.909 17.006 ot_diis_step 117 11.6 0.020 0.020 16.849 16.850 qs_ot_get_p 128 10.4 0.002 0.002 16.393 16.482 make_m2s 5014 13.6 0.063 0.067 14.588 15.381 make_images 5014 14.6 2.161 2.580 14.280 15.070 cp_fm_cholesky_invert 11 10.9 14.576 14.587 14.576 14.587 sum_up_and_integrate 128 10.3 0.002 0.002 13.752 13.801 integrate_v_rspace 128 11.3 0.004 0.004 13.691 13.743 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.379 13.414 calculate_rho_elec 128 8.7 0.257 0.268 13.379 13.413 qs_ot_p2m_diag 83 11.4 0.496 0.502 12.697 12.715 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.541 11.541 multiply_cannon_sync_h2d 10028 15.6 10.880 11.267 10.880 11.267 init_scf_run 11 5.9 0.000 0.002 10.365 10.365 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.365 10.365 hybrid_alltoall_any 5200 16.5 0.843 3.777 8.250 9.651 make_images_data 5014 15.6 0.055 0.065 8.407 9.635 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.678 8.689 cp_fm_diag_elpa_base 83 14.4 8.445 8.522 8.674 8.686 grid_integrate_task_list 128 12.3 7.785 8.176 7.785 8.176 dbcsr_mm_accdrv_process 20762 16.1 3.343 4.545 7.380 8.092 fft_wrap_pw1pw2 1291 11.7 0.017 0.019 8.001 8.030 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.923 8.016 multiply_cannon_metrocomm1 10028 15.6 0.030 0.031 4.321 7.169 wfi_extrapolate 11 7.9 0.001 0.001 7.152 7.152 fft_wrap_pw1pw2_140 523 12.2 0.433 0.440 7.062 7.098 cp_fm_cholesky_decompose 22 10.9 6.485 6.552 6.485 6.552 density_rs2pw 128 9.7 0.005 0.005 6.098 6.336 calculate_dm_sparse 128 9.5 0.001 0.001 6.090 6.160 fft3d_ps 1291 13.7 3.107 3.200 5.978 6.001 grid_collocate_task_list 128 9.7 5.573 5.947 5.573 5.947 dbcsr_complete_redistribute 395 12.7 2.030 2.068 5.096 5.391 qs_energies_init_hamiltonians 11 5.9 0.017 0.066 5.368 5.371 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.296 5.335 mp_alltoall_d11v 2415 14.1 4.539 4.927 4.539 4.927 mp_allgather_i34 2507 14.6 1.071 4.501 1.071 4.501 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.130 4.189 potential_pw2rs 128 12.3 0.026 0.027 3.916 3.920 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.592 3.894 multiply_cannon_metrocomm4 7521 15.6 0.025 0.028 1.747 3.750 mp_irecv_dv 28860 15.9 1.706 3.677 1.706 3.677 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.561 3.622 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.271 3.586 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.548 3.580 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.418 3.504 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.463 3.470 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=162.606000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1352.909091, yerr=29.705997 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 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 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 2.885829E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+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 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4048 67935. MP_Allreduce 11179 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.069 0.106 244.978 244.979 qs_mol_dyn_low 1 2.0 0.003 0.004 244.164 244.176 qs_forces 11 3.9 0.005 0.005 243.644 243.647 qs_energies 11 4.9 0.002 0.002 234.896 234.899 scf_env_do_scf 11 5.9 0.001 0.001 212.312 212.329 velocity_verlet 10 3.0 0.016 0.046 170.659 170.716 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 129.784 129.786 qs_scf_new_mos 118 7.6 0.001 0.001 86.979 87.012 qs_scf_loop_do_ot 118 8.6 0.001 0.001 86.978 87.011 dbcsr_multiply_generic 2535 12.6 0.224 0.230 82.307 82.755 init_scf_loop 11 6.9 0.000 0.000 82.219 82.222 ot_scf_mini 118 9.6 0.004 0.004 82.003 82.013 prepare_preconditioner 11 7.9 0.000 0.000 77.396 77.402 make_preconditioner 11 8.9 0.000 0.000 77.396 77.402 make_full_inverse_cholesky 11 9.9 0.038 0.039 61.123 74.663 multiply_cannon 2535 13.6 0.682 0.739 59.151 60.739 multiply_cannon_loop 2535 14.6 1.048 1.074 55.107 56.377 cp_fm_upper_to_full 106 14.8 33.969 48.668 33.969 48.668 ot_mini 118 10.6 0.001 0.001 43.687 43.697 rebuild_ks_matrix 129 8.3 0.001 0.001 33.855 33.865 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 33.854 33.864 mp_waitall_1 104820 16.8 27.920 32.580 27.920 32.580 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.549 31.567 dbcsr_complete_redistribute 397 12.7 3.858 4.006 22.231 30.799 copy_fm_to_dbcsr 210 11.7 0.001 0.001 18.709 27.291 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 16.231 24.691 qs_ot_get_derivative 118 11.6 0.002 0.002 24.229 24.238 qs_ot_get_p 129 10.4 0.021 0.047 23.156 23.181 mp_alltoall_i22 720 14.1 13.983 22.774 13.983 22.774 make_m2s 5070 13.6 0.075 0.078 18.644 20.153 multiply_cannon_metrocomm3 10140 15.6 0.025 0.025 18.778 19.979 make_images 5070 14.6 3.110 3.315 18.161 19.669 ot_diis_step 118 11.6 0.022 0.022 19.436 19.436 qs_ot_p2m_diag 84 11.4 0.891 0.897 19.065 19.091 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 18.723 18.805 apply_single 129 13.6 0.001 0.001 18.723 18.805 multiply_cannon_multrec 10140 15.6 10.362 12.217 18.285 18.357 cp_fm_cholesky_invert 11 10.9 18.080 18.093 18.080 18.093 cp_dbcsr_syevd 84 12.4 0.006 0.006 17.262 17.264 sum_up_and_integrate 129 10.3 0.002 0.002 15.731 15.817 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.760 15.778 calculate_rho_elec 129 8.7 0.486 0.487 15.759 15.777 integrate_v_rspace 129 11.3 0.005 0.005 15.667 15.753 multiply_cannon_sync_h2d 10140 15.6 14.432 14.457 14.432 14.457 cp_fm_diag_elpa 84 13.4 0.000 0.000 14.106 14.107 cp_fm_diag_elpa_base 84 14.4 11.267 12.290 14.102 14.102 hybrid_alltoall_any 5257 16.5 1.321 3.076 10.767 13.192 make_images_data 5070 15.6 0.064 0.068 10.483 12.732 init_scf_run 11 5.9 0.000 0.001 12.140 12.141 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.140 12.141 fft_wrap_pw1pw2 1301 11.7 0.020 0.020 9.899 9.910 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.745 9.756 dbcsr_mm_accdrv_process 20958 16.1 4.302 6.081 7.668 9.583 grid_integrate_task_list 129 12.3 8.615 8.806 8.615 8.806 wfi_extrapolate 11 7.9 0.001 0.001 8.797 8.798 fft_wrap_pw1pw2_140 527 12.2 0.831 0.831 8.785 8.796 qs_energies_init_hamiltonians 11 5.9 0.035 0.090 8.247 8.254 cp_fm_cholesky_decompose 22 10.9 7.332 7.374 7.332 7.374 fft3d_ps 1301 13.7 4.007 4.021 7.185 7.210 mp_alltoall_d11v 2429 14.1 7.058 7.193 7.058 7.193 calculate_dm_sparse 129 9.5 0.001 0.001 6.876 6.908 density_rs2pw 129 9.7 0.005 0.005 6.792 6.856 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.549 6.620 grid_collocate_task_list 129 9.7 6.457 6.528 6.457 6.528 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.153 6.380 qs_env_update_s_mstruct 11 6.9 0.001 0.001 5.005 5.200 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.568 4.955 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=244.979000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2579.181818, yerr=158.617790 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_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 1.270710E+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 24272141. 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 0.057 0.098 85.568 85.569 qs_energies 1 2.0 0.000 0.000 84.666 84.674 ls_scf 1 3.0 0.000 0.000 83.783 83.789 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.702 72.886 multiply_cannon 111 7.7 0.017 0.020 55.995 57.050 multiply_cannon_loop 111 8.7 0.230 0.247 52.585 53.820 ls_scf_main 1 4.0 0.000 0.000 52.404 52.405 density_matrix_trs4 2 5.0 0.002 0.003 46.829 46.909 ls_scf_init_scf 1 4.0 0.000 0.000 28.361 28.364 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.220 27.284 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.086 25.107 mp_waitall_1 11031 10.9 22.335 24.034 22.335 24.034 multiply_cannon_multrec 2664 9.7 8.149 8.879 15.655 17.392 multiply_cannon_sync_h2d 2664 9.7 13.666 15.587 13.666 15.587 make_m2s 222 7.7 0.009 0.011 13.129 13.612 make_images 222 8.7 0.099 0.109 13.107 13.592 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.612 12.534 make_images_data 222 9.7 0.004 0.005 7.708 8.288 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.360 8.106 dbcsr_mm_accdrv_process 4760 10.4 0.603 0.729 7.122 8.080 hybrid_alltoall_any 227 10.6 0.216 1.842 6.644 7.925 dbcsr_mm_accdrv_process_sort 4760 11.4 6.317 7.191 6.317 7.191 calculate_norms 4752 9.8 5.535 6.224 5.535 6.224 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.041 5.185 mp_sum_l 887 5.1 3.084 4.922 3.084 4.922 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.043 3.734 mp_irecv_dv 6231 10.9 2.026 3.707 2.026 3.707 make_images_sizes 222 9.7 0.000 0.000 0.721 3.560 mp_alltoall_i44 222 10.7 0.720 3.559 0.720 3.559 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.263 3.480 arnoldi_extremal 4 6.8 0.000 0.000 3.294 3.317 arnoldi_normal_ev 4 7.8 0.001 0.003 3.293 3.317 build_subspace 16 8.4 0.009 0.012 3.198 3.200 ls_scf_post 1 4.0 0.000 0.000 3.018 3.025 ls_scf_store_result 1 5.0 0.000 0.000 2.850 2.877 dbcsr_special_finalize 555 9.7 0.005 0.006 2.340 2.797 dbcsr_merge_single_wm 555 10.7 0.459 0.601 2.331 2.788 make_images_pack 222 9.7 2.213 2.630 2.215 2.632 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.318 2.556 dbcsr_sort_data 658 11.4 2.131 2.524 2.131 2.524 dbcsr_matrix_vector_mult_local 304 10.0 2.068 2.458 2.070 2.460 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.320 2.431 buffer_matrices_ensure_size 222 8.7 1.755 2.058 1.755 2.058 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.811 1.812 rebuild_ks_matrix 3 7.3 0.000 0.000 1.801 1.803 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.006 1.801 1.802 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.569000, 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/98431332714702b8c1de37aafb7c70fe7a47c205_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 2.196017E+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 0.058 0.252 90.497 90.498 qs_energies 1 2.0 0.000 0.001 89.515 89.554 ls_scf 1 3.0 0.001 0.008 88.083 88.123 dbcsr_multiply_generic 111 6.7 0.018 0.037 74.122 74.430 multiply_cannon 111 7.7 0.026 0.040 52.809 56.806 ls_scf_main 1 4.0 0.002 0.036 54.282 54.290 multiply_cannon_loop 111 8.7 0.136 0.146 49.974 52.899 density_matrix_trs4 2 5.0 0.002 0.004 48.651 48.839 ls_scf_init_scf 1 4.0 0.000 0.002 30.137 30.140 mp_waitall_1 9105 10.9 20.747 29.871 20.747 29.871 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.841 28.956 multiply_cannon_multrec 1332 9.7 13.215 16.894 22.582 27.457 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.350 26.365 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.595 20.587 make_m2s 222 7.7 0.006 0.007 14.971 15.561 make_images 222 8.7 1.375 1.694 14.940 15.531 dbcsr_mm_accdrv_process 4041 10.4 0.331 0.558 8.966 10.487 dbcsr_mm_accdrv_process_sort 4041 11.4 8.471 9.953 8.471 9.953 hybrid_alltoall_any 227 10.6 0.541 2.558 8.056 9.485 make_images_data 222 9.7 0.004 0.005 8.619 9.440 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.159 7.766 mp_sum_l 887 5.1 4.902 7.753 4.902 7.753 mp_irecv_dv 3311 11.0 3.139 7.711 3.139 7.711 calculate_norms 2376 9.8 6.057 6.837 6.057 6.837 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.689 6.047 multiply_cannon_sync_h2d 1332 9.7 4.892 6.016 4.892 6.016 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.878 5.052 arnoldi_extremal 4 6.8 0.000 0.000 4.582 4.604 arnoldi_normal_ev 4 7.8 0.001 0.005 4.582 4.604 build_subspace 16 8.4 0.014 0.021 4.333 4.336 ls_scf_post 1 4.0 0.006 0.045 3.663 3.697 ls_scf_store_result 1 5.0 0.000 0.000 3.376 3.475 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.121 3.348 dbcsr_matrix_vector_mult_local 304 10.0 2.743 3.218 2.746 3.220 mp_allgather_i34 111 8.7 0.815 3.029 0.815 3.029 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.586 2.662 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.183 2.644 dbcsr_data_new 4174 10.1 2.132 2.431 2.132 2.431 make_images_pack 222 9.7 1.828 2.125 1.831 2.127 dbcsr_sort_data 436 11.2 1.854 2.111 1.854 2.111 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.964 1.969 rebuild_ks_matrix 3 7.3 0.000 0.000 1.951 1.956 qs_ks_build_kohn_sham_matrix 3 8.3 0.009 0.022 1.951 1.956 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.498000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1796.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_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 2.871988E+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 0.199 0.297 92.616 92.621 qs_energies 1 2.0 0.002 0.019 91.609 91.634 ls_scf 1 3.0 0.000 0.001 90.016 90.043 dbcsr_multiply_generic 111 6.7 0.016 0.019 74.376 74.639 multiply_cannon 111 7.7 0.036 0.106 51.499 56.380 ls_scf_main 1 4.0 0.001 0.007 55.868 55.872 multiply_cannon_loop 111 8.7 0.118 0.132 48.781 52.251 density_matrix_trs4 2 5.0 0.003 0.013 49.984 50.137 mp_waitall_1 7281 11.0 22.970 32.025 22.970 32.025 ls_scf_init_scf 1 4.0 0.000 0.001 30.463 30.466 ls_scf_init_matrix_S 1 5.0 0.002 0.014 29.222 29.290 matrix_sqrt_Newton_Schulz 2 6.5 0.013 0.085 26.708 26.740 multiply_cannon_multrec 888 9.7 12.749 15.420 21.417 24.688 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.507 21.501 make_m2s 222 7.7 0.006 0.007 16.502 17.173 make_images 222 8.7 1.582 1.850 16.464 17.135 make_images_data 222 9.7 0.004 0.004 9.658 10.622 hybrid_alltoall_any 227 10.6 0.640 2.954 9.090 10.296 dbcsr_mm_accdrv_process 3754 10.4 0.305 0.494 8.194 9.480 dbcsr_mm_accdrv_process_sort 3754 11.4 7.759 8.986 7.759 8.986 mp_sum_l 887 5.1 4.728 7.935 4.728 7.935 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.449 7.483 multiply_cannon_sync_h2d 888 9.7 6.065 7.243 6.065 7.243 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.453 7.007 mp_irecv_dv 2335 11.1 2.437 6.953 2.437 6.953 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.543 6.070 arnoldi_extremal 4 6.8 0.000 0.000 5.171 5.216 arnoldi_normal_ev 4 7.8 0.034 0.091 5.171 5.216 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.712 4.901 calculate_norms 1584 9.8 4.399 4.880 4.399 4.880 build_subspace 16 8.4 0.014 0.020 4.733 4.741 mp_allgather_i34 111 8.7 0.835 3.795 0.835 3.795 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.419 3.753 ls_scf_post 1 4.0 0.000 0.003 3.684 3.711 dbcsr_matrix_vector_mult_local 304 10.0 3.018 3.593 3.020 3.596 ls_scf_store_result 1 5.0 0.000 0.000 3.391 3.476 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.846 2.945 dbcsr_data_new 4116 9.9 2.107 2.480 2.107 2.480 make_images_sizes 222 9.7 0.000 0.000 1.014 2.192 mp_alltoall_i44 222 10.7 1.014 2.192 1.014 2.192 dbcsr_sort_data 325 11.1 1.881 2.170 1.881 2.170 dbcsr_finalize 304 7.8 0.029 0.051 1.627 1.998 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.957 1.961 rebuild_ks_matrix 3 7.3 0.000 0.000 1.940 1.943 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.006 1.940 1.943 dbcsr_merge_all 275 8.9 0.250 0.342 1.541 1.887 make_images_pack 222 9.7 1.621 1.882 1.624 1.885 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.621000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2233.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_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 3.430392E+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 0.048 0.075 97.850 97.851 qs_energies 1 2.0 0.003 0.024 96.955 96.979 ls_scf 1 3.0 0.001 0.006 95.209 95.213 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.804 79.069 ls_scf_main 1 4.0 0.000 0.003 59.077 59.078 multiply_cannon 111 7.7 0.041 0.082 51.741 56.224 density_matrix_trs4 2 5.0 0.002 0.004 52.956 53.098 multiply_cannon_loop 111 8.7 0.153 0.167 46.656 49.858 ls_scf_init_scf 1 4.0 0.000 0.001 32.893 32.896 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.642 31.716 matrix_sqrt_Newton_Schulz 2 6.5 0.016 0.069 29.018 29.029 mp_waitall_1 6369 11.0 22.650 28.708 22.650 28.708 multiply_cannon_multrec 1332 9.7 14.235 16.965 22.246 25.907 make_m2s 222 7.7 0.007 0.008 21.269 22.656 make_images 222 8.7 3.158 3.634 21.219 22.608 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.051 17.261 make_images_data 222 9.7 0.004 0.004 11.883 13.621 hybrid_alltoall_any 227 10.6 0.808 3.836 11.381 12.995 dbcsr_mm_accdrv_process 3641 10.4 0.306 0.483 7.647 9.206 dbcsr_mm_accdrv_process_sort 3641 11.4 7.202 8.705 7.202 8.705 mp_sum_l 887 5.1 4.220 8.315 4.220 8.315 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.268 6.588 multiply_cannon_sync_h2d 1332 9.7 5.467 6.176 5.467 6.176 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.083 6.107 mp_irecv_dv 3229 10.9 2.058 6.019 2.058 6.019 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.547 5.290 arnoldi_extremal 4 6.8 0.000 0.000 5.208 5.241 arnoldi_normal_ev 4 7.8 0.019 0.071 5.208 5.240 build_subspace 16 8.4 0.014 0.021 4.832 4.840 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.550 4.767 calculate_norms 2376 9.8 4.162 4.519 4.162 4.519 mp_allgather_i34 111 8.7 2.162 4.515 2.162 4.515 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.565 3.862 dbcsr_matrix_vector_mult_local 304 10.0 3.182 3.685 3.184 3.687 dbcsr_sort_data 658 11.4 3.064 3.331 3.064 3.331 ls_scf_post 1 4.0 0.000 0.001 3.239 3.245 dbcsr_special_finalize 555 9.7 0.006 0.007 2.822 3.195 dbcsr_merge_single_wm 555 10.7 0.544 0.669 2.813 3.187 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.987 3.040 ls_scf_store_result 1 5.0 0.000 0.000 2.974 3.038 dbcsr_data_release 10477 10.7 1.587 2.445 1.587 2.445 dbcsr_finalize 304 7.8 0.049 0.061 1.815 2.020 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.964 1.965 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.851000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2827.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_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 4.845318E+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 0.044 0.058 92.231 92.231 qs_energies 1 2.0 0.000 0.000 91.318 91.333 ls_scf 1 3.0 0.000 0.000 89.329 89.348 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.929 71.168 ls_scf_main 1 4.0 0.000 0.000 56.647 56.648 multiply_cannon 111 7.7 0.079 0.158 52.942 56.351 multiply_cannon_loop 111 8.7 0.088 0.095 50.347 51.745 density_matrix_trs4 2 5.0 0.002 0.003 49.690 49.771 ls_scf_init_scf 1 4.0 0.000 0.000 29.327 29.332 mp_waitall_1 5436 11.0 24.525 28.598 24.525 28.598 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.017 28.047 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.935 25.952 multiply_cannon_multrec 444 9.7 13.726 16.948 20.922 22.576 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.687 15.525 make_m2s 222 7.7 0.005 0.005 13.473 14.411 make_images 222 8.7 2.037 2.478 13.406 14.343 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.456 13.415 make_images_data 222 9.7 0.003 0.004 8.204 9.688 hybrid_alltoall_any 227 10.6 0.804 3.822 8.119 9.652 dbcsr_mm_accdrv_process 3003 10.4 0.371 0.426 6.870 8.089 multiply_cannon_sync_h2d 444 9.7 6.752 7.719 6.752 7.719 dbcsr_mm_accdrv_process_sort 3003 11.4 6.499 7.679 6.499 7.679 arnoldi_extremal 4 6.8 0.000 0.000 5.762 5.780 arnoldi_normal_ev 4 7.8 0.002 0.005 5.762 5.780 build_subspace 16 8.4 0.015 0.019 5.369 5.381 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.381 4.590 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.138 4.329 dbcsr_matrix_vector_mult_local 304 10.0 3.676 4.132 3.678 4.135 mp_sum_l 887 5.1 2.596 4.058 2.596 4.058 mp_allgather_i34 111 8.7 1.169 3.745 1.169 3.745 calculate_norms 792 9.8 3.634 3.708 3.634 3.708 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.550 3.608 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.497 3.577 mp_irecv_dv 1241 11.2 1.534 3.554 1.534 3.554 ls_scf_post 1 4.0 0.000 0.000 3.355 3.371 ls_scf_store_result 1 5.0 0.000 0.000 3.142 3.178 make_images_sizes 222 9.7 0.000 0.000 0.873 3.145 mp_alltoall_i44 222 10.7 0.873 3.144 0.873 3.144 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.762 3.031 dbcsr_finalize 304 7.8 0.062 0.077 2.210 2.279 dbcsr_data_new 4608 9.7 1.808 2.201 1.808 2.201 dbcsr_merge_all 275 8.9 0.480 0.522 2.065 2.132 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.122 2.123 rebuild_ks_matrix 3 7.3 0.000 0.000 2.089 2.090 qs_ks_build_kohn_sham_matrix 3 8.3 0.009 0.028 2.089 2.090 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 1.972 1.973 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.231000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3764.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_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 9.023918E+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 0.098 0.123 111.681 111.682 qs_energies 1 2.0 0.000 0.000 110.283 110.291 ls_scf 1 3.0 0.000 0.000 107.340 107.348 dbcsr_multiply_generic 111 6.7 0.024 0.027 81.385 81.530 ls_scf_main 1 4.0 0.000 0.000 63.439 63.439 multiply_cannon 111 7.7 0.155 0.258 55.123 57.450 density_matrix_trs4 2 5.0 0.002 0.003 54.660 54.711 multiply_cannon_loop 111 8.7 0.098 0.099 51.996 52.612 ls_scf_init_scf 1 4.0 0.000 0.000 40.110 40.110 ls_scf_init_matrix_S 1 5.0 0.000 0.000 38.675 38.699 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 35.965 35.970 mp_waitall_1 4527 11.1 23.120 28.208 23.120 28.208 make_m2s 222 7.7 0.005 0.005 22.612 23.805 make_images 222 8.7 3.569 3.873 22.504 23.695 multiply_cannon_multrec 444 9.7 17.823 18.510 22.530 23.252 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.488 16.930 hybrid_alltoall_any 227 10.6 1.654 3.610 12.790 15.685 make_images_data 222 9.7 0.003 0.004 13.003 15.245 multiply_cannon_sync_h2d 444 9.7 8.785 8.846 8.785 8.846 arnoldi_extremal 4 6.8 0.000 0.000 7.296 7.306 arnoldi_normal_ev 4 7.8 0.003 0.009 7.296 7.306 compute_matrix_preconditioner 1 6.0 0.002 0.002 7.190 7.195 build_subspace 16 8.4 0.026 0.036 6.747 6.757 acc_transpose_blocks 444 9.7 0.003 0.003 5.429 6.653 acc_transpose_blocks_kernels 444 10.7 0.006 0.006 5.347 6.572 jit_kernel_transpose 1 13.0 5.341 6.566 5.341 6.566 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.379 5.538 dbcsr_matrix_vector_mult_local 304 10.0 4.981 5.283 4.984 5.286 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.865 5.117 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.908 5.000 dbcsr_mm_accdrv_process 1814 10.4 0.281 0.387 4.531 4.691 dbcsr_mm_accdrv_process_sort 1814 11.4 4.180 4.320 4.180 4.320 mp_allgather_i34 111 8.7 1.216 3.898 1.216 3.898 ls_scf_post 1 4.0 0.000 0.000 3.791 3.800 make_images_sizes 222 9.7 0.000 0.000 1.473 3.618 mp_alltoall_i44 222 10.7 1.473 3.617 1.473 3.617 ls_scf_store_result 1 5.0 0.000 0.000 3.511 3.548 calculate_norms 792 9.8 3.236 3.273 3.236 3.273 dbcsr_finalize 304 7.8 0.082 0.089 3.079 3.168 dbcsr_merge_all 275 8.9 0.886 0.921 2.864 2.946 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.913 2.913 dbcsr_complete_redistribute 5 7.6 1.434 1.479 2.745 2.862 matrix_ls_to_qs 2 6.0 0.000 0.000 2.396 2.527 dbcsr_sort_data 325 11.1 2.437 2.499 2.437 2.499 dbcsr_new_transposed 4 7.5 0.252 0.269 2.327 2.349 dbcsr_data_new 6591 9.6 1.844 2.328 1.844 2.328 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.311 2.312 rebuild_ks_matrix 3 7.3 0.000 0.000 2.245 2.246 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.245 2.246 dbcsr_frobenius_norm 74 6.6 2.056 2.138 2.208 2.243 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=111.682000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7152.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/98431332714702b8c1de37aafb7c70fe7a47c205_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 589.643776E+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 1346700736. 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 0.066 0.157 222.943 222.946 qs_mol_dyn_low 1 2.0 0.010 0.092 221.208 221.396 qs_forces 5 3.8 0.005 0.057 220.857 220.861 qs_energies 5 4.8 0.002 0.036 217.949 217.989 scf_env_do_scf 5 5.8 0.000 0.008 199.783 199.786 scf_env_do_scf_inner_loop 105 6.6 0.002 0.009 175.509 175.516 qs_scf_new_mos 105 7.6 0.000 0.001 137.614 137.792 qs_scf_loop_do_ot 105 8.6 0.001 0.001 137.613 137.791 dbcsr_multiply_generic 1445 12.2 0.130 0.139 136.135 136.513 ot_scf_mini 105 9.6 0.003 0.004 127.903 128.051 multiply_cannon 1445 13.2 0.273 0.287 117.774 119.608 multiply_cannon_loop 1445 14.2 2.857 2.986 116.165 117.305 velocity_verlet 4 3.0 0.018 0.138 102.105 102.108 ot_mini 105 10.6 0.001 0.002 59.362 59.493 multiply_cannon_multrec 69360 15.2 29.764 34.760 39.967 45.521 mp_waitall_1 488190 16.1 33.853 43.029 33.853 43.029 qs_ot_get_p 112 10.4 0.001 0.002 38.639 38.958 qs_ot_get_derivative 55 11.6 0.001 0.022 37.620 37.741 multiply_cannon_metrocomm3 69360 15.2 0.205 0.218 25.172 35.472 multiply_cannon_sync_h2d 69360 15.2 29.194 33.673 29.194 33.673 rebuild_ks_matrix 110 8.4 0.000 0.001 27.986 28.175 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.068 27.986 28.175 qs_ot_p2m_diag 40 11.0 0.020 0.031 27.885 27.958 qs_ks_update_qs_env 112 7.6 0.001 0.001 25.734 25.901 cp_dbcsr_syevd 40 12.0 0.002 0.002 24.713 24.713 init_scf_loop 7 6.6 0.001 0.034 24.240 24.242 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 22.963 23.189 apply_single 62 13.6 0.000 0.000 22.963 23.189 ot_new_cg_direction 55 11.6 0.001 0.003 20.973 20.974 cp_fm_syevd 40 13.0 0.000 0.000 19.634 19.798 prepare_preconditioner 7 7.6 0.000 0.000 19.285 19.316 make_preconditioner 7 8.6 0.000 0.003 19.285 19.316 qs_rho_update_rho_low 110 7.6 0.001 0.002 15.639 15.970 calculate_rho_elec 110 8.6 0.029 0.037 15.639 15.970 init_scf_run 5 5.8 0.000 0.001 15.639 15.640 scf_env_initial_rho_setup 5 6.8 0.002 0.022 15.639 15.640 cp_fm_redistribute_end 40 14.0 7.529 15.002 7.541 15.007 cp_fm_syevd_base 40 14.0 7.457 14.933 7.457 14.933 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.743 14.888 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.341 13.437 mp_sum_l 4764 12.2 12.190 13.009 12.190 13.009 make_full_inverse_cholesky 7 9.6 0.000 0.000 12.319 12.380 calculate_dm_sparse 110 9.5 0.000 0.001 12.145 12.284 acc_transpose_blocks 69360 15.2 0.348 0.366 11.304 11.976 calculate_first_density_matrix 1 7.0 0.001 0.043 11.406 11.417 dbcsr_mm_accdrv_process 154766 15.8 6.292 6.467 10.070 10.861 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 10.676 10.694 density_rs2pw 110 9.6 0.004 0.005 10.090 10.530 fft_wrap_pw1pw2 1425 12.5 0.019 0.025 10.227 10.427 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 10.274 10.327 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.171 10.253 qs_vxc_create 110 10.4 0.009 0.046 9.715 9.731 check_diag 80 13.5 8.653 8.936 9.321 9.463 multiply_cannon_metrocomm1 69360 15.2 0.098 0.103 4.585 9.265 acc_transpose_blocks_kernels 69360 16.2 0.847 0.887 8.565 9.140 fft_wrap_pw1pw2_240 915 14.0 0.199 0.217 8.942 9.105 cp_fm_cholesky_invert 7 10.6 8.258 8.276 8.258 8.276 jit_kernel_transpose 5 15.0 7.718 8.266 7.718 8.266 sum_up_and_integrate 60 10.3 0.002 0.053 8.045 8.098 fft3d_pb 915 15.0 2.247 2.400 7.892 8.069 integrate_v_rspace 60 11.3 0.001 0.002 8.027 8.036 transfer_rs2pw 445 10.6 0.008 0.009 7.247 7.696 xc_rho_set_and_dset_create 110 12.4 0.076 0.109 6.877 7.117 make_full_single_inverse 7 9.6 0.001 0.001 6.715 6.746 make_m2s 2890 13.2 0.079 0.089 6.091 6.659 make_images 2890 14.2 0.239 0.260 5.983 6.551 xc_vxc_pw_create 60 11.3 0.036 0.059 6.508 6.535 xc_pw_derive 510 13.4 0.005 0.006 5.607 5.669 mp_alltoall_z22v 2340 16.7 5.032 5.252 5.032 5.252 mp_waitany 7680 13.5 4.372 4.908 4.372 4.908 multiply_cannon_metrocomm4 67915 15.2 0.182 0.205 2.008 4.698 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=222.946000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=560.000000, yerr=2.449490 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 98431332714702b8c1de37aafb7c70fe7a47c205 Summary: empty Status: OK