=== 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: be1f5882581616708fdc25ee40722957c884746f ################# 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/01 job id: 53377635 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/02 job id: 53377636 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/03 job id: 53377637 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/04 job id: 53377638 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/05 job id: 53377639 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/06 job id: 53377640 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/07 job id: 53377641 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/08 job id: 53377642 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/09 job id: 53377643 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/10 job id: 53377644 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/11 job id: 53377645 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/12 job id: 53377646 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/13 job id: 53377647 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/14 job id: 53377648 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/15 job id: 53377649 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/16 job id: 53377651 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/17 job id: 53377652 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/18 job id: 53377653 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/19 job id: 53377654 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/20 job id: 53377655 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/21 job id: 53377656 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/22 job id: 53377657 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/23 job id: 53377658 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/24 job id: 53377659 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/25 job id: 53377660 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/26 job id: 53377661 --- 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/be1f5882581616708fdc25ee40722957c884746f_performance_tests/27 job id: 53377662 --- 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/be1f5882581616708fdc25ee40722957c884746f_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 0.053 0.081 128.966 128.967 farming_run 1 2.0 128.318 128.319 128.880 128.914 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.528766E+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 0.005 0.026 110.531 110.532 qs_energies 1 2.0 0.001 0.001 110.345 110.347 mp2_main 1 3.0 0.000 0.000 108.522 108.524 mp2_gpw_main 1 4.0 0.035 0.059 107.621 107.623 mp2_ri_gpw_compute_in 1 5.0 0.174 0.177 88.568 88.824 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 57.189 57.452 mp2_eri_3c_integrate_gpw 272 7.0 0.150 0.164 43.633 48.656 get_2c_integrals 1 6.0 0.008 0.009 30.727 31.206 integrate_v_rspace 273 8.0 0.438 0.454 25.559 30.085 fft_wrap_pw1pw2 5465 10.4 0.068 0.070 28.080 28.571 grid_integrate_task_list 273 9.0 20.802 26.277 20.802 26.277 fft_wrap_pw1pw2_100 2178 11.4 1.309 1.454 25.524 26.044 compute_2c_integrals 1 7.0 0.002 0.002 21.381 21.383 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 20.853 21.109 mp2_eri_2c_integrate_gpw 1 9.0 2.389 2.433 20.851 21.108 rpa_ri_compute_en 1 5.0 0.020 0.026 18.927 19.063 fft3d_s 5443 12.4 16.903 17.416 16.924 17.437 ao_to_mo_and_store_B_mult_1 272 7.0 10.765 15.304 10.765 15.304 calculate_wavefunction 272 8.0 5.280 5.428 12.738 13.638 calc_potential_gpw 544 9.5 0.004 0.004 10.978 11.742 rpa_num_int 1 6.0 0.000 0.006 10.514 10.514 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.455 10.513 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 9.820 10.509 potential_pw2rs 545 10.0 0.106 0.108 8.393 10.193 cp_fm_cholesky_decompose 12 8.2 9.513 10.000 9.513 10.000 calc_mat_Q 8 8.0 0.001 0.007 9.436 9.558 cholesky_decomp 1 7.0 0.000 0.000 8.541 9.026 contract_S_to_Q 8 9.0 0.000 0.000 8.859 8.979 parallel_gemm_fm 14 9.1 0.000 0.000 8.486 8.625 parallel_gemm_fm_cosma 14 10.1 8.486 8.625 8.486 8.625 collocate_single_gaussian 272 10.0 0.040 0.042 7.832 8.515 create_integ_mat 1 6.0 0.113 0.150 8.058 8.058 array2fm 1 7.0 0.000 0.000 6.895 7.371 pw_poisson_solve 545 10.5 0.010 0.010 4.669 4.906 pw_scatter_s 2720 12.7 4.545 4.902 4.545 4.902 pw_gather_s 2722 12.2 3.832 4.762 3.832 4.762 pw_poisson_set 548 11.5 0.018 0.019 3.378 3.606 array2fm_buffer_send 1 8.0 2.959 3.149 2.959 3.149 pw_copy 4911 11.6 2.615 2.728 2.615 2.728 ao_to_mo_and_store_B_E_Ex_1 272 7.0 1.082 1.231 2.025 2.232 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=107.620028, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2783.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 0.052 0.102 428.671 428.686 farming_run 1 2.0 426.881 426.904 428.526 428.549 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.278267E+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 0.030 222.801 222.802 qs_energies 1 2.0 0.005 0.018 222.455 222.472 scf_env_do_scf 1 3.0 0.000 0.000 114.684 114.684 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.825 113.831 rebuild_ks_matrix 4 6.0 0.000 0.000 113.824 113.829 qs_ks_build_kohn_sham_matrix 4 7.0 0.053 0.059 113.824 113.829 hfx_ks_matrix 4 8.0 0.001 0.001 113.478 113.481 integrate_four_center 4 9.0 0.156 0.618 113.477 113.481 mp2_main 1 3.0 0.001 0.007 107.481 107.498 mp2_gpw_main 1 4.0 0.071 0.151 106.589 106.605 integrate_four_center_main 4 10.0 0.098 0.414 102.083 104.393 integrate_four_center_bin 262 11.0 101.985 104.385 101.985 104.385 init_scf_loop 1 4.0 0.000 0.000 96.681 96.681 mp2_ri_gpw_compute_in 1 5.0 0.074 0.105 79.707 80.726 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.060 56.712 57.721 mp2_eri_3c_integrate_gpw 91 7.0 0.147 0.160 44.564 49.348 integrate_v_rspace 95 8.0 0.397 0.572 28.232 32.844 fft_wrap_pw1pw2 1868 10.4 0.029 0.035 30.513 31.001 fft_wrap_pw1pw2_100 730 11.4 0.639 0.747 28.215 28.755 mp2_ri_gpw_compute_en 1 5.0 0.058 0.079 26.603 28.687 grid_integrate_task_list 95 9.0 23.527 28.386 23.527 28.386 ao_to_mo_and_store_B_mult_1 91 7.0 10.479 27.111 10.479 27.111 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.851 1.900 24.951 24.962 get_2c_integrals 1 6.0 0.022 0.097 22.852 22.887 compute_2c_integrals 1 7.0 0.002 0.003 21.995 22.016 compute_2c_integrals_loop_lm 1 8.0 0.002 0.006 21.648 21.876 mp2_eri_2c_integrate_gpw 1 9.0 1.745 1.810 21.646 21.876 fft3d_s 1823 12.4 19.249 20.000 19.262 20.013 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.002 18.002 calc_potential_gpw 182 9.5 0.002 0.002 12.285 12.892 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 10.632 11.305 calculate_wavefunction 91 8.0 1.982 2.021 10.002 10.795 potential_pw2rs 186 10.0 0.033 0.034 9.052 10.310 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.591 8.773 9.207 collocate_single_gaussian 91 10.0 0.018 0.025 8.283 8.917 local_gemm 172 8.0 8.218 8.634 8.218 8.634 mp2_ri_gpw_compute_en_comm 22 7.0 0.504 0.531 7.948 8.340 mp_sync 37 10.5 3.265 7.291 3.265 7.291 integrate_four_center_load 4 10.0 0.000 0.000 6.761 6.766 hfx_load_balance 1 11.0 0.000 0.000 6.761 6.766 mp2_ri_gpw_compute_en_ener 172 7.0 6.334 6.421 6.334 6.421 mp_sendrecv_dm3 2068 8.0 5.974 6.376 5.974 6.376 pw_poisson_solve 186 10.4 0.005 0.006 5.490 6.060 pw_gather_s 912 12.2 4.831 5.412 4.831 5.412 pw_scatter_s 910 12.7 4.093 4.531 4.093 4.531 pw_poisson_set 189 11.4 0.009 0.011 4.056 4.514 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=106.588360, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1533.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 445.526016E+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 596931. 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 0.009 0.027 53.222 53.223 qs_mol_dyn_low 1 2.0 0.020 0.084 52.721 52.729 qs_forces 11 3.9 0.012 0.030 52.569 52.585 qs_energies 11 4.9 0.006 0.034 51.070 51.102 scf_env_do_scf 11 5.9 0.000 0.001 44.719 44.719 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 42.869 42.869 qs_scf_new_mos 108 7.5 0.000 0.000 33.130 33.434 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.130 33.433 dbcsr_multiply_generic 2286 12.5 0.094 0.100 33.039 33.418 ot_scf_mini 108 9.5 0.002 0.002 31.465 31.650 multiply_cannon 2286 13.5 0.177 0.186 26.121 27.659 multiply_cannon_loop 2286 14.5 1.812 1.910 25.498 27.050 velocity_verlet 10 3.0 0.011 0.053 25.880 25.883 ot_mini 108 10.5 0.001 0.001 18.912 19.159 qs_ot_get_derivative 108 11.5 0.001 0.001 15.908 16.109 mp_waitall_1 245248 16.5 8.317 14.156 8.317 14.156 multiply_cannon_metrocomm3 54864 15.5 0.072 0.077 6.020 12.701 multiply_cannon_multrec 54864 15.5 3.648 5.622 7.662 11.147 qs_ot_get_p 119 10.4 0.001 0.001 7.937 8.193 rebuild_ks_matrix 119 8.3 0.000 0.000 7.738 7.905 qs_ks_build_kohn_sham_matrix 119 9.3 0.015 0.051 7.738 7.905 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.808 6.959 mp_sum_l 7287 12.8 5.134 6.789 5.134 6.789 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.990 6.461 multiply_cannon_sync_h2d 54864 15.5 5.153 6.270 5.153 6.270 dbcsr_mm_accdrv_process 76910 16.1 1.838 2.899 3.929 5.675 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.106 5.148 init_scf_run 11 5.9 0.000 0.001 4.978 4.979 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.978 4.978 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.696 4.802 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.346 4.346 sum_up_and_integrate 119 10.3 0.001 0.003 4.295 4.301 integrate_v_rspace 119 11.3 0.002 0.003 4.284 4.292 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.162 4.162 cp_fm_redistribute_end 50 14.0 2.125 4.139 2.130 4.142 qs_rho_update_rho_low 119 7.7 0.000 0.001 3.925 4.051 calculate_rho_elec 119 8.7 0.011 0.016 3.924 4.051 cp_fm_diag_elpa_base 50 14.0 2.006 4.036 2.010 4.044 calculate_first_density_matrix 1 7.0 0.000 0.000 3.127 3.143 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.942 3.135 apply_single 119 13.6 0.000 0.000 2.941 3.135 calculate_dm_sparse 119 9.5 0.000 0.000 2.882 3.032 multiply_cannon_metrocomm1 54864 15.5 0.055 0.060 1.710 2.782 acc_transpose_blocks 54864 15.5 0.218 0.238 2.214 2.744 ot_diis_step 108 11.5 0.006 0.006 2.726 2.727 jit_kernel_multiply 13 15.8 2.028 2.710 2.028 2.710 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.407 2.452 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.311 2.325 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.052 2.111 grid_integrate_task_list 119 12.3 2.013 2.103 2.013 2.103 density_rs2pw 119 9.7 0.003 0.004 1.973 2.087 mp_sum_d 4137 12.0 1.345 1.912 1.345 1.912 init_scf_loop 11 6.9 0.000 0.000 1.834 1.835 wfi_extrapolate 11 7.9 0.003 0.024 1.772 1.772 potential_pw2rs 119 12.3 0.004 0.004 1.684 1.695 make_m2s 4572 13.5 0.054 0.056 1.558 1.608 make_images 4572 14.5 0.133 0.138 1.476 1.524 transfer_rs2pw 487 10.6 0.006 0.007 1.325 1.485 fft_wrap_pw1pw2 1201 11.6 0.011 0.013 1.419 1.484 mp_alltoall_d11v 2130 13.8 1.245 1.442 1.245 1.442 acc_transpose_blocks_sync 164592 16.5 1.199 1.442 1.199 1.442 mp_waitany 12084 13.8 1.260 1.440 1.260 1.440 grid_collocate_task_list 119 9.7 1.352 1.419 1.352 1.419 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.376 1.405 transfer_pw2rs 487 13.2 0.005 0.006 1.274 1.280 fft3d_ps 1201 13.6 0.358 0.458 1.193 1.249 fft_wrap_pw1pw2_140 487 12.2 0.048 0.053 1.102 1.169 parallel_gemm_fm 81 9.0 0.000 0.000 1.145 1.160 parallel_gemm_fm_cosma 81 10.0 1.145 1.160 1.145 1.160 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.735 1.104 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.223000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=424.818182, yerr=0.574960 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 484.696064E+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 1969194. 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 0.068 0.157 40.291 40.293 qs_mol_dyn_low 1 2.0 0.004 0.007 39.356 39.365 qs_forces 11 3.9 0.003 0.008 39.249 39.251 qs_energies 11 4.9 0.003 0.012 37.585 37.591 scf_env_do_scf 11 5.9 0.001 0.002 31.835 31.836 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 29.296 29.297 dbcsr_multiply_generic 2286 12.5 0.106 0.135 22.506 22.862 qs_scf_new_mos 108 7.5 0.001 0.001 20.805 21.037 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.804 21.036 ot_scf_mini 108 9.5 0.003 0.003 19.889 20.052 multiply_cannon 2286 13.5 0.205 0.213 17.212 18.677 velocity_verlet 10 3.0 0.002 0.004 18.051 18.053 multiply_cannon_loop 2286 14.5 1.189 1.242 16.061 17.679 ot_mini 108 10.5 0.001 0.001 12.361 12.593 mp_waitall_1 200699 16.5 5.677 11.018 5.677 11.018 qs_ot_get_derivative 108 11.5 0.001 0.001 10.014 10.177 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.160 9.782 multiply_cannon_multrec 27432 15.5 1.830 4.105 6.289 9.151 rebuild_ks_matrix 119 8.3 0.000 0.000 6.786 6.934 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.016 6.786 6.933 dbcsr_mm_accdrv_process 47894 16.0 3.639 5.722 4.381 6.569 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.980 6.115 qs_ot_get_p 119 10.4 0.001 0.004 4.630 4.856 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.739 4.706 init_scf_run 11 5.9 0.000 0.001 4.459 4.459 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.458 4.459 mp_sum_l 7287 12.8 2.378 4.436 2.378 4.436 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.117 4.074 apply_single 119 13.6 0.000 0.000 3.117 4.074 sum_up_and_integrate 119 10.3 0.001 0.002 3.664 3.669 integrate_v_rspace 119 11.3 0.002 0.003 3.649 3.655 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.594 3.646 calculate_rho_elec 119 8.7 0.026 0.053 3.593 3.645 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.976 2.995 calculate_first_density_matrix 1 7.0 0.022 0.096 2.884 2.957 make_m2s 4572 13.5 0.052 0.054 2.583 2.925 make_images 4572 14.5 0.204 0.243 2.493 2.838 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.584 2.585 init_scf_loop 11 6.9 0.027 0.123 2.518 2.519 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.365 2.451 ot_diis_step 108 11.5 0.011 0.011 2.298 2.298 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.269 2.269 calculate_dm_sparse 119 9.5 0.001 0.001 2.185 2.265 cp_fm_redistribute_end 50 14.0 1.150 2.243 1.152 2.246 multiply_cannon_sync_h2d 27432 15.5 1.701 2.220 1.701 2.220 cp_fm_diag_elpa_base 50 14.0 1.061 2.153 1.089 2.195 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.058 2.060 acc_transpose_blocks 27432 15.5 0.112 0.119 1.657 2.013 grid_integrate_task_list 119 12.3 1.834 1.933 1.834 1.933 density_rs2pw 119 9.7 0.004 0.005 1.817 1.893 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.825 1.886 jit_kernel_multiply 9 16.3 0.684 1.831 0.684 1.831 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.634 1.663 make_images_data 4572 15.5 0.047 0.053 1.216 1.523 mp_alltoall_d11v 2130 13.8 1.294 1.437 1.294 1.437 hybrid_alltoall_any 4725 16.4 0.053 0.116 1.087 1.435 wfi_extrapolate 11 7.9 0.001 0.002 1.426 1.426 potential_pw2rs 119 12.3 0.006 0.006 1.415 1.421 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.378 1.385 fft3d_ps 1201 13.6 0.507 0.562 1.354 1.378 prepare_preconditioner 11 7.9 0.000 0.000 1.340 1.367 make_preconditioner 11 8.9 0.000 0.001 1.340 1.367 grid_collocate_task_list 119 9.7 1.287 1.348 1.287 1.348 fft_wrap_pw1pw2_140 487 12.2 0.047 0.050 1.295 1.325 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.238 1.282 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.196 1.242 transfer_rs2pw 487 10.6 0.005 0.006 1.130 1.225 mp_sum_d 4137 12.0 0.551 1.040 0.551 1.040 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.009 1.011 mp_allgather_i34 2286 14.5 0.569 0.978 0.569 0.978 transfer_pw2rs 487 13.2 0.004 0.005 0.947 0.950 acc_transpose_blocks_kernels 27432 16.5 0.191 0.312 0.702 0.940 acc_transpose_blocks_sync 82296 16.5 0.817 0.939 0.817 0.939 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.864 0.877 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.293000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=460.454545, yerr=1.876342 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 517.554176E+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 0.086 0.105 32.063 32.064 qs_mol_dyn_low 1 2.0 0.003 0.008 31.570 31.585 qs_forces 11 3.9 0.002 0.003 31.510 31.512 qs_energies 11 4.9 0.001 0.002 29.974 29.978 scf_env_do_scf 11 5.9 0.000 0.001 25.070 25.070 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.740 22.741 dbcsr_multiply_generic 2286 12.5 0.096 0.098 17.098 17.184 qs_scf_new_mos 108 7.5 0.001 0.001 15.306 15.320 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.306 15.320 velocity_verlet 10 3.0 0.003 0.017 14.708 14.710 multiply_cannon 2286 13.5 0.192 0.196 13.873 14.669 ot_scf_mini 108 9.5 0.002 0.002 14.552 14.566 multiply_cannon_loop 2286 14.5 0.861 0.912 13.106 13.917 ot_mini 108 10.5 0.001 0.001 9.054 9.072 qs_ot_get_derivative 108 11.5 0.001 0.001 7.561 7.575 multiply_cannon_multrec 18288 15.5 1.873 2.877 7.089 7.352 dbcsr_mm_accdrv_process 38222 16.0 5.039 5.940 5.124 6.005 rebuild_ks_matrix 119 8.3 0.000 0.000 5.926 5.973 qs_ks_build_kohn_sham_matrix 119 9.3 0.033 0.073 5.925 5.972 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.222 5.267 mp_waitall_1 158411 16.6 2.572 3.685 2.572 3.685 init_scf_run 11 5.9 0.000 0.001 3.637 3.637 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.637 3.637 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.853 3.483 sum_up_and_integrate 119 10.3 0.001 0.001 3.417 3.431 integrate_v_rspace 119 11.3 0.003 0.003 3.404 3.418 qs_ot_get_p 119 10.4 0.001 0.002 3.286 3.315 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.077 3.081 calculate_rho_elec 119 8.7 0.031 0.031 3.077 3.080 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.465 2.483 calculate_first_density_matrix 1 7.0 0.000 0.000 2.472 2.473 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.048 2.411 apply_single 119 13.6 0.000 0.000 2.048 2.411 init_scf_loop 11 6.9 0.000 0.000 2.314 2.315 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.192 2.198 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.901 1.901 make_m2s 4572 13.5 0.044 0.045 1.758 1.897 grid_integrate_task_list 119 12.3 1.805 1.887 1.805 1.887 calculate_dm_sparse 119 9.5 0.000 0.001 1.872 1.880 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.843 1.844 acc_transpose_blocks 18288 15.5 0.079 0.082 1.703 1.825 make_images 4572 14.5 0.192 0.204 1.672 1.810 density_rs2pw 119 9.7 0.003 0.003 1.652 1.778 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.701 1.706 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.690 1.697 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.658 1.670 cp_fm_diag_elpa_base 50 14.0 1.635 1.650 1.656 1.668 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.565 1.572 prepare_preconditioner 11 7.9 0.000 0.000 1.509 1.533 make_preconditioner 11 8.9 0.000 0.000 1.509 1.532 mp_sum_l 7287 12.8 1.082 1.517 1.082 1.517 ot_diis_step 108 11.5 0.011 0.011 1.480 1.480 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.390 1.470 grid_collocate_task_list 119 9.7 1.234 1.311 1.234 1.311 fft_wrap_pw1pw2_140 487 12.2 0.061 0.063 1.270 1.278 fft3d_ps 1201 13.6 0.529 0.549 1.249 1.259 potential_pw2rs 119 12.3 0.007 0.007 1.221 1.224 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.213 1.216 multiply_cannon_sync_h2d 18288 15.5 1.048 1.191 1.048 1.191 wfi_extrapolate 11 7.9 0.001 0.001 1.115 1.115 transfer_rs2pw 487 10.6 0.005 0.005 0.933 1.069 qs_energies_init_hamiltonians 11 5.9 0.047 0.082 1.060 1.063 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.938 0.952 make_images_data 4572 15.5 0.047 0.051 0.748 0.917 acc_transpose_blocks_kernels 18288 16.5 0.216 0.222 0.866 0.877 acc_transpose_blocks_sync 54864 16.5 0.739 0.862 0.739 0.862 hybrid_alltoall_any 4725 16.4 0.058 0.117 0.658 0.849 mp_alltoall_d11v 2130 13.8 0.657 0.809 0.657 0.809 transfer_pw2rs 487 13.2 0.004 0.004 0.781 0.783 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.771 0.774 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.677 0.738 mp_waitany 9880 13.7 0.521 0.661 0.521 0.661 jit_kernel_transpose 5 15.6 0.651 0.660 0.651 0.660 mp_alltoall_z22v 1201 15.6 0.580 0.647 0.580 0.647 transfer_rs2pw_140 130 11.5 0.115 0.118 0.513 0.643 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.064000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=491.181818, yerr=2.081004 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 554.954752E+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 0.030 0.101 36.803 36.803 qs_mol_dyn_low 1 2.0 0.003 0.004 36.527 36.562 qs_forces 11 3.9 0.002 0.002 36.468 36.469 qs_energies 11 4.9 0.002 0.002 34.757 34.763 scf_env_do_scf 11 5.9 0.000 0.001 29.098 29.098 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 26.234 26.234 dbcsr_multiply_generic 2286 12.5 0.104 0.105 21.063 21.217 qs_scf_new_mos 108 7.5 0.001 0.001 18.297 18.360 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.297 18.359 velocity_verlet 10 3.0 0.001 0.002 17.818 17.819 multiply_cannon 2286 13.5 0.216 0.223 17.246 17.759 ot_scf_mini 108 9.5 0.002 0.003 17.269 17.320 multiply_cannon_loop 2286 14.5 1.529 1.633 16.296 16.710 ot_mini 108 10.5 0.001 0.001 10.823 10.881 multiply_cannon_multrec 27432 15.5 2.642 3.455 9.771 10.174 qs_ot_get_derivative 108 11.5 0.001 0.001 8.953 9.007 dbcsr_mm_accdrv_process 47916 15.9 6.414 8.257 7.031 8.596 rebuild_ks_matrix 119 8.3 0.000 0.000 6.308 6.368 qs_ks_build_kohn_sham_matrix 119 9.3 0.025 0.033 6.307 6.367 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.580 5.635 init_scf_run 11 5.9 0.000 0.001 4.332 4.333 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.332 4.332 qs_ot_get_p 119 10.4 0.011 0.017 3.612 3.700 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.174 3.646 sum_up_and_integrate 119 10.3 0.001 0.001 3.437 3.444 integrate_v_rspace 119 11.3 0.003 0.003 3.426 3.433 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.240 3.280 calculate_rho_elec 119 8.7 0.040 0.046 3.239 3.280 calculate_first_density_matrix 1 7.0 0.000 0.000 2.947 2.949 acc_transpose_blocks 27432 15.5 0.116 0.120 2.726 2.851 init_scf_loop 11 6.9 0.000 0.000 2.843 2.844 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.202 2.669 apply_single 119 13.6 0.000 0.000 2.201 2.669 mp_waitall_1 137007 16.6 1.839 2.607 1.839 2.607 make_m2s 4572 13.5 0.054 0.056 2.299 2.421 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.378 2.379 calculate_dm_sparse 119 9.5 0.000 0.000 2.313 2.366 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.324 2.351 make_images 4572 14.5 0.272 0.332 2.191 2.310 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.203 2.218 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.153 2.164 jit_kernel_multiply 8 15.9 0.551 2.116 0.551 2.116 prepare_preconditioner 11 7.9 0.000 0.000 1.956 1.975 make_preconditioner 11 8.9 0.000 0.000 1.956 1.975 grid_integrate_task_list 119 12.3 1.830 1.895 1.830 1.895 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.694 1.893 ot_diis_step 108 11.5 0.012 0.012 1.827 1.827 mp_sum_l 7287 12.8 1.134 1.821 1.134 1.821 density_rs2pw 119 9.7 0.003 0.004 1.706 1.811 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.754 1.754 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.720 1.750 acc_transpose_blocks_sync 82296 16.5 1.456 1.563 1.456 1.563 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.503 1.504 cp_fm_redistribute_end 50 14.0 0.752 1.479 0.756 1.480 cp_fm_diag_elpa_base 50 14.0 0.698 1.413 0.722 1.454 fft_wrap_pw1pw2_140 487 12.2 0.071 0.073 1.411 1.444 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.414 1.425 fft3d_ps 1201 13.6 0.573 0.635 1.381 1.409 grid_collocate_task_list 119 9.7 1.249 1.333 1.249 1.333 wfi_extrapolate 11 7.9 0.001 0.001 1.331 1.331 multiply_cannon_metrocomm3 27432 15.5 0.039 0.041 0.803 1.318 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.292 1.308 potential_pw2rs 119 12.3 0.008 0.009 1.240 1.242 acc_transpose_blocks_kernels 27432 16.5 0.269 0.276 1.125 1.152 qs_energies_init_hamiltonians 11 5.9 0.003 0.005 1.071 1.072 transfer_rs2pw 487 10.6 0.005 0.005 0.898 1.004 make_images_data 4572 15.5 0.047 0.051 0.853 0.983 hybrid_alltoall_any 4725 16.4 0.066 0.155 0.728 0.918 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.801 0.877 jit_kernel_transpose 5 15.6 0.856 0.877 0.856 0.877 mp_alltoall_d11v 2130 13.8 0.762 0.860 0.762 0.860 dbcsr_complete_redistribute 329 12.2 0.110 0.128 0.676 0.844 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.814 0.819 transfer_pw2rs 487 13.2 0.004 0.004 0.744 0.746 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.803000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.272727, yerr=2.339033 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 622.444544E+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 0.008 0.029 29.231 29.232 qs_mol_dyn_low 1 2.0 0.005 0.033 28.940 28.949 qs_forces 11 3.9 0.039 0.061 28.859 28.861 qs_energies 11 4.9 0.006 0.010 27.123 27.156 scf_env_do_scf 11 5.9 0.000 0.001 21.941 21.942 scf_env_do_scf_inner_loop 108 6.5 0.006 0.011 19.322 19.323 velocity_verlet 10 3.0 0.014 0.025 14.438 14.442 dbcsr_multiply_generic 2286 12.5 0.097 0.100 13.633 13.744 qs_scf_new_mos 108 7.5 0.001 0.001 11.820 11.848 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.820 11.847 multiply_cannon 2286 13.5 0.222 0.225 10.938 11.378 ot_scf_mini 108 9.5 0.002 0.002 11.105 11.134 multiply_cannon_loop 2286 14.5 0.641 0.659 10.011 10.197 ot_mini 108 10.5 0.001 0.001 6.402 6.435 multiply_cannon_multrec 9144 15.5 1.666 1.950 6.141 6.424 rebuild_ks_matrix 119 8.3 0.000 0.000 5.791 5.827 qs_ks_build_kohn_sham_matrix 119 9.3 0.078 0.099 5.790 5.826 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.158 5.191 qs_ot_get_derivative 108 11.5 0.001 0.001 5.114 5.143 dbcsr_mm_accdrv_process 12550 15.8 3.690 4.373 4.368 4.538 init_scf_run 11 5.9 0.000 0.001 3.720 3.720 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.720 3.720 sum_up_and_integrate 119 10.3 0.001 0.001 3.347 3.352 integrate_v_rspace 119 11.3 0.003 0.003 3.337 3.342 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.258 3.264 calculate_rho_elec 119 8.7 0.060 0.061 3.258 3.263 qs_ot_get_p 119 10.4 0.041 0.078 2.970 3.021 init_scf_loop 11 6.9 0.000 0.000 2.599 2.601 calculate_first_density_matrix 1 7.0 0.000 0.000 2.591 2.591 mp_waitall_1 115863 16.7 1.545 2.052 1.545 2.052 calculate_dm_sparse 119 9.5 0.000 0.000 2.008 2.028 make_m2s 4572 13.5 0.034 0.036 1.819 1.983 grid_integrate_task_list 119 12.3 1.872 1.950 1.872 1.950 make_images 4572 14.5 0.267 0.297 1.729 1.891 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.824 1.826 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.776 1.798 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.742 1.752 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.743 1.745 density_rs2pw 119 9.7 0.003 0.003 1.638 1.736 acc_transpose_blocks 9144 15.5 0.041 0.044 1.597 1.619 jit_kernel_multiply 8 15.6 0.639 1.610 0.639 1.610 prepare_preconditioner 11 7.9 0.000 0.000 1.557 1.587 make_preconditioner 11 8.9 0.000 0.000 1.557 1.587 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.585 1.586 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.572 1.583 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.452 1.504 fft_wrap_pw1pw2_140 487 12.2 0.094 0.096 1.438 1.448 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.391 1.397 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.360 1.374 grid_collocate_task_list 119 9.7 1.296 1.372 1.296 1.372 fft3d_ps 1201 13.6 0.632 0.647 1.355 1.366 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.289 1.298 cp_fm_diag_elpa_base 50 14.0 1.263 1.280 1.288 1.296 qs_energies_init_hamiltonians 11 5.9 0.004 0.012 1.249 1.280 ot_diis_step 108 11.5 0.013 0.013 1.275 1.275 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.221 1.253 apply_single 119 13.6 0.000 0.000 1.220 1.253 potential_pw2rs 119 12.3 0.010 0.010 1.155 1.157 wfi_extrapolate 11 7.9 0.001 0.001 1.079 1.079 hybrid_alltoall_any 4725 16.4 0.065 0.177 0.772 1.018 make_images_data 4572 15.5 0.042 0.046 0.790 0.985 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.876 0.929 transfer_rs2pw 487 10.6 0.004 0.005 0.818 0.908 mp_alltoall_d11v 2130 13.8 0.815 0.881 0.815 0.881 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.420 0.872 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.821 0.842 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.821 0.827 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.753 0.755 cp_fm_cholesky_invert 11 10.9 0.746 0.751 0.746 0.751 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.690 0.741 acc_transpose_blocks_sync 27432 16.5 0.719 0.737 0.719 0.737 jit_kernel_transpose 5 15.6 0.704 0.722 0.704 0.722 transfer_pw2rs 487 13.2 0.003 0.004 0.668 0.670 mp_allgather_i34 2286 14.5 0.244 0.669 0.244 0.669 mp_alltoall_z22v 1201 15.6 0.599 0.633 0.599 0.633 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.232000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=589.000000, yerr=5.045250 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 775.102464E+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 0.036 0.063 41.983 41.984 qs_mol_dyn_low 1 2.0 0.003 0.003 41.563 41.570 qs_forces 11 3.9 0.002 0.002 41.226 41.227 qs_energies 11 4.9 0.002 0.002 39.236 39.240 scf_env_do_scf 11 5.9 0.001 0.001 33.319 33.319 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.293 27.295 velocity_verlet 10 3.0 0.002 0.002 22.636 22.676 dbcsr_multiply_generic 2286 12.5 0.105 0.106 19.546 19.639 qs_scf_new_mos 108 7.5 0.001 0.001 17.936 18.033 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.936 18.032 ot_scf_mini 108 9.5 0.002 0.002 16.762 16.860 multiply_cannon 2286 13.5 0.295 0.299 15.439 16.299 multiply_cannon_loop 2286 14.5 0.859 0.881 14.164 15.017 ot_mini 108 10.5 0.001 0.001 10.281 10.399 multiply_cannon_multrec 9144 15.5 3.463 4.849 8.989 9.175 qs_ot_get_derivative 108 11.5 0.001 0.001 8.187 8.284 rebuild_ks_matrix 119 8.3 0.000 0.000 7.117 7.274 qs_ks_build_kohn_sham_matrix 119 9.3 0.048 0.062 7.116 7.274 dbcsr_mm_accdrv_process 12550 15.8 4.686 6.641 5.393 6.766 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.408 6.551 init_scf_loop 11 6.9 0.000 0.000 5.994 5.998 prepare_preconditioner 11 7.9 0.000 0.000 4.980 5.003 make_preconditioner 11 8.9 0.000 0.000 4.980 5.003 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.010 4.868 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.002 4.010 calculate_rho_elec 119 8.7 0.118 0.121 4.002 4.009 init_scf_run 11 5.9 0.000 0.001 3.804 3.804 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.803 3.803 sum_up_and_integrate 119 10.3 0.001 0.001 3.772 3.778 integrate_v_rspace 119 11.3 0.004 0.004 3.762 3.768 qs_ot_get_p 119 10.4 0.001 0.002 3.511 3.660 mp_waitall_1 94719 16.7 2.504 3.469 2.504 3.469 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.991 3.389 cp_fm_upper_to_full 72 14.2 2.069 2.947 2.069 2.947 make_m2s 4572 13.5 0.038 0.038 2.495 2.657 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.227 2.547 apply_single 119 13.6 0.000 0.000 2.227 2.547 make_images 4572 14.5 0.350 0.382 2.374 2.539 fft_wrap_pw1pw2 1201 11.6 0.014 0.014 2.424 2.429 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.430 2.356 calculate_first_density_matrix 1 7.0 0.000 0.000 2.335 2.335 calculate_dm_sparse 119 9.5 0.000 0.000 2.305 2.322 grid_integrate_task_list 119 12.3 2.126 2.171 2.126 2.171 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.078 2.129 dbcsr_complete_redistribute 329 12.2 0.259 0.264 1.572 2.116 ot_diis_step 108 11.5 0.014 0.015 2.063 2.063 qs_ot_p2m_diag 50 11.0 0.042 0.043 2.049 2.050 fft_wrap_pw1pw2_140 487 12.2 0.179 0.180 2.040 2.047 density_rs2pw 119 9.7 0.003 0.003 2.000 2.043 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.979 1.979 acc_transpose_blocks 9144 15.5 0.044 0.045 1.908 1.978 mp_sum_l 7287 12.8 1.066 1.888 1.066 1.888 fft3d_ps 1201 13.6 0.857 0.880 1.876 1.881 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.833 1.834 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.220 1.753 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.681 1.696 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.695 1.695 grid_collocate_task_list 119 9.7 1.554 1.586 1.554 1.586 mp_alltoall_i22 627 13.8 0.975 1.504 0.975 1.504 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 0.965 1.498 hybrid_alltoall_any 4725 16.4 0.091 0.152 1.170 1.422 wfi_extrapolate 11 7.9 0.001 0.001 1.405 1.405 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.377 1.377 cp_fm_diag_elpa_base 50 14.0 1.291 1.305 1.375 1.375 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.336 1.358 make_images_data 4572 15.5 0.046 0.048 1.122 1.336 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.290 1.299 mp_alltoall_d11v 2130 13.8 1.214 1.240 1.214 1.240 potential_pw2rs 119 12.3 0.013 0.014 1.230 1.231 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.162 1.179 cp_fm_cholesky_invert 11 10.9 1.150 1.157 1.150 1.157 acc_transpose_blocks_sync 27432 16.5 1.098 1.136 1.098 1.136 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.986 1.033 qs_create_task_list 11 7.9 0.001 0.001 0.931 0.943 generate_qs_task_list 11 8.9 0.367 0.388 0.930 0.943 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.919 0.933 jit_kernel_multiply 5 15.6 0.678 0.919 0.678 0.919 mp_alltoall_z22v 1201 15.6 0.884 0.911 0.884 0.911 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.984000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=728.000000, yerr=13.816986 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 497.176576E+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 1716596. 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 0.216 0.450 86.493 86.568 qs_mol_dyn_low 1 2.0 0.008 0.040 85.311 85.365 qs_forces 11 3.9 0.012 0.081 84.897 84.952 qs_energies 11 4.9 0.002 0.005 81.948 81.993 scf_env_do_scf 11 5.9 0.000 0.001 72.248 72.293 scf_env_do_scf_inner_loop 99 6.5 0.005 0.023 66.505 66.547 dbcsr_multiply_generic 2055 12.4 0.107 0.110 53.413 53.645 qs_scf_new_mos 99 7.5 0.000 0.001 49.471 49.604 qs_scf_loop_do_ot 99 8.5 0.000 0.001 49.471 49.603 ot_scf_mini 99 9.5 0.002 0.002 47.017 47.137 multiply_cannon 2055 13.4 0.173 0.178 43.458 44.308 velocity_verlet 10 3.0 0.017 0.083 43.470 43.511 multiply_cannon_loop 2055 14.4 1.821 1.914 42.478 43.342 ot_mini 99 10.5 0.001 0.001 28.278 28.385 qs_ot_get_derivative 99 11.5 0.001 0.002 21.300 21.422 multiply_cannon_multrec 49320 15.4 11.119 12.021 17.476 18.152 rebuild_ks_matrix 110 8.3 0.000 0.001 14.672 14.865 qs_ks_build_kohn_sham_matrix 110 9.3 0.035 0.130 14.672 14.864 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.895 13.081 mp_waitall_1 220248 16.4 11.801 12.783 11.801 12.783 qs_ot_get_p 110 10.4 0.001 0.001 10.042 10.161 multiply_cannon_sync_h2d 49320 15.4 9.329 10.133 9.329 10.133 multiply_cannon_metrocomm3 49320 15.4 0.080 0.084 7.146 8.285 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.720 8.272 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.277 7.902 apply_single 110 13.6 0.000 0.000 7.277 7.902 init_scf_run 11 5.9 0.000 0.001 7.192 7.197 scf_env_initial_rho_setup 11 6.9 0.001 0.014 7.192 7.197 sum_up_and_integrate 110 10.3 0.004 0.024 6.926 6.939 integrate_v_rspace 110 11.3 0.003 0.003 6.898 6.920 mp_sum_l 6594 12.7 5.857 6.859 5.857 6.859 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.734 6.769 dbcsr_mm_accdrv_process 87628 16.1 3.195 3.447 6.231 6.669 ot_diis_step 99 11.5 0.005 0.006 6.639 6.643 qs_rho_update_rho_low 110 7.6 0.005 0.037 6.346 6.494 calculate_rho_elec 110 8.6 0.027 0.081 6.341 6.493 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.246 6.351 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.765 5.768 init_scf_loop 11 6.9 0.000 0.001 5.710 5.714 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.286 5.290 cp_fm_redistribute_end 48 14.0 2.683 5.260 2.692 5.266 cp_fm_diag_elpa_base 48 14.0 2.561 5.144 2.569 5.159 calculate_dm_sparse 110 9.5 0.001 0.001 3.919 4.018 make_m2s 4110 13.4 0.061 0.066 3.915 4.007 wfi_extrapolate 11 7.9 0.001 0.003 4.002 4.004 make_images 4110 14.4 0.177 0.190 3.819 3.916 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.783 3.787 multiply_cannon_metrocomm1 49320 15.4 0.063 0.066 2.771 3.773 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.395 3.479 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.419 3.457 grid_integrate_task_list 110 12.3 3.240 3.393 3.240 3.393 density_rs2pw 110 9.6 0.004 0.004 3.219 3.382 prepare_preconditioner 11 7.9 0.000 0.000 3.152 3.188 make_preconditioner 11 8.9 0.000 0.000 3.152 3.188 calculate_first_density_matrix 1 7.0 0.004 0.032 2.998 3.003 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.905 2.968 jit_kernel_multiply 13 15.9 2.753 2.816 2.753 2.816 acc_transpose_blocks 49320 15.4 0.210 0.225 2.640 2.795 mp_alltoall_d11v 2046 13.8 2.370 2.783 2.370 2.783 fft_wrap_pw1pw2 1111 11.6 0.014 0.017 2.672 2.738 potential_pw2rs 110 12.3 0.006 0.006 2.518 2.541 fft_wrap_pw1pw2_140 451 12.1 0.091 0.097 2.259 2.327 fft3d_ps 1111 13.6 0.726 0.812 2.213 2.271 grid_collocate_task_list 110 9.6 2.158 2.262 2.158 2.262 transfer_rs2pw 451 10.6 0.006 0.006 1.945 2.168 mp_sum_d 3891 11.9 1.616 2.163 1.616 2.163 mp_waitany 14300 13.8 1.826 2.052 1.826 2.052 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.917 1.932 make_images_data 4110 15.4 0.043 0.047 1.774 1.888 hybrid_alltoall_any 4261 16.3 0.084 0.486 1.551 1.835 transfer_pw2rs 451 13.1 0.005 0.007 1.778 1.795 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.752 1.787 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=86.568000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=472.545455, yerr=2.641062 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 590.123008E+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 3099482. 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 0.132 0.300 69.990 69.999 qs_mol_dyn_low 1 2.0 0.021 0.092 68.942 69.062 qs_forces 11 3.9 0.004 0.006 68.385 68.389 qs_energies 11 4.9 0.018 0.090 65.052 65.057 scf_env_do_scf 11 5.9 0.000 0.001 55.729 55.730 scf_env_do_scf_inner_loop 99 6.5 0.007 0.017 48.596 48.597 dbcsr_multiply_generic 2055 12.4 0.116 0.120 38.380 38.577 velocity_verlet 10 3.0 0.029 0.101 35.324 35.354 qs_scf_new_mos 99 7.5 0.001 0.001 33.033 33.207 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.032 33.206 multiply_cannon 2055 13.4 0.223 0.242 31.694 32.713 ot_scf_mini 99 9.5 0.003 0.003 31.367 31.551 multiply_cannon_loop 2055 14.4 1.160 1.185 30.429 31.126 ot_mini 99 10.5 0.001 0.001 18.319 18.506 multiply_cannon_multrec 24660 15.4 6.977 8.775 14.324 16.042 rebuild_ks_matrix 110 8.3 0.000 0.000 13.555 13.700 qs_ks_build_kohn_sham_matrix 110 9.3 0.035 0.091 13.554 13.700 qs_ot_get_derivative 99 11.5 0.001 0.002 12.564 12.750 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.967 12.111 mp_waitall_1 176588 16.5 7.507 10.189 7.507 10.189 multiply_cannon_metrocomm3 24660 15.4 0.072 0.074 5.160 8.019 multiply_cannon_sync_h2d 24660 15.4 6.387 7.594 6.387 7.594 dbcsr_mm_accdrv_process 52282 16.1 5.638 6.709 7.178 7.522 init_scf_loop 11 6.9 0.000 0.000 7.096 7.097 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.444 7.059 apply_single 110 13.6 0.000 0.001 6.443 7.059 qs_ot_get_p 110 10.4 0.023 0.073 6.521 6.711 init_scf_run 11 5.9 0.000 0.001 6.535 6.536 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.535 6.536 sum_up_and_integrate 110 10.3 0.007 0.031 6.185 6.210 integrate_v_rspace 110 11.3 0.002 0.003 6.153 6.166 ot_diis_step 99 11.5 0.010 0.010 5.708 5.708 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.695 5.706 calculate_rho_elec 110 8.6 0.051 0.144 5.695 5.705 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.738 5.461 prepare_preconditioner 11 7.9 0.000 0.000 4.806 4.841 make_preconditioner 11 8.9 0.000 0.000 4.806 4.841 make_m2s 4110 13.4 0.057 0.061 4.162 4.645 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.427 4.543 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.513 4.533 make_images 4110 14.4 0.403 0.457 4.050 4.529 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.057 4.058 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.536 3.536 cp_fm_redistribute_end 48 14.0 1.784 3.510 1.789 3.513 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.499 3.501 cp_fm_diag_elpa_base 48 14.0 1.663 3.371 1.718 3.450 wfi_extrapolate 11 7.9 0.002 0.010 3.427 3.427 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.264 3.356 grid_integrate_task_list 110 12.3 3.160 3.335 3.160 3.335 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.161 3.209 fft_wrap_pw1pw2 1111 11.6 0.014 0.017 3.024 3.161 calculate_dm_sparse 110 9.5 0.001 0.001 3.085 3.116 density_rs2pw 110 9.6 0.004 0.004 2.846 3.041 calculate_first_density_matrix 1 7.0 0.000 0.000 2.918 2.920 mp_sum_l 6594 12.7 2.088 2.802 2.088 2.802 make_images_data 4110 15.4 0.048 0.052 2.258 2.730 fft_wrap_pw1pw2_140 451 12.1 0.109 0.116 2.590 2.724 hybrid_alltoall_any 4261 16.3 0.105 0.450 1.972 2.706 fft3d_ps 1111 13.6 1.064 1.274 2.405 2.529 cp_fm_cholesky_invert 11 10.9 2.341 2.356 2.341 2.356 grid_collocate_task_list 110 9.6 2.169 2.329 2.169 2.329 qs_energies_init_hamiltonians 11 5.9 0.029 0.068 2.199 2.202 mp_alltoall_d11v 2046 13.8 1.859 2.120 1.859 2.120 jit_kernel_multiply 10 16.2 1.170 2.109 1.170 2.109 potential_pw2rs 110 12.3 0.008 0.008 2.069 2.075 acc_transpose_blocks 24660 15.4 0.114 0.118 1.991 2.034 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.950 1.965 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.798 1.806 multiply_cannon_metrocomm4 22605 15.4 0.078 0.082 0.778 1.722 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.655 1.668 transfer_rs2pw 451 10.6 0.007 0.007 1.448 1.633 mp_irecv_dv 57340 16.2 0.647 1.598 0.647 1.598 mp_allgather_i34 2055 14.4 0.576 1.584 0.576 1.584 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.388 1.492 mp_sum_d 3891 11.9 1.154 1.463 1.154 1.463 mp_waitany 10164 13.8 1.247 1.427 1.247 1.427 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.999000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=558.000000, yerr=6.438520 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 667.394048E+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 0.008 0.029 59.136 59.136 qs_mol_dyn_low 1 2.0 0.003 0.004 58.867 58.886 qs_forces 11 3.9 0.003 0.003 58.437 58.438 qs_energies 11 4.9 0.002 0.002 55.286 55.290 scf_env_do_scf 11 5.9 0.000 0.001 46.687 46.687 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 39.199 39.199 velocity_verlet 10 3.0 0.001 0.002 31.362 31.375 dbcsr_multiply_generic 2055 12.4 0.113 0.117 29.139 29.420 qs_scf_new_mos 99 7.5 0.001 0.001 24.939 25.049 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.939 25.048 multiply_cannon 2055 13.4 0.212 0.220 22.826 24.048 ot_scf_mini 99 9.5 0.002 0.002 23.698 23.833 multiply_cannon_loop 2055 14.4 0.815 0.839 21.669 22.701 ot_mini 99 10.5 0.001 0.001 13.618 13.756 rebuild_ks_matrix 110 8.3 0.000 0.000 12.065 12.227 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.064 12.227 multiply_cannon_multrec 16440 15.4 3.670 4.529 10.148 11.108 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.622 10.772 mp_waitall_1 139946 16.5 6.936 10.193 6.936 10.193 qs_ot_get_derivative 99 11.5 0.001 0.001 9.220 9.355 init_scf_loop 11 6.9 0.000 0.000 7.453 7.453 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.289 7.215 dbcsr_mm_accdrv_process 34862 16.1 5.663 6.123 6.323 6.514 init_scf_run 11 5.9 0.000 0.001 6.146 6.146 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.145 6.145 sum_up_and_integrate 110 10.3 0.001 0.002 6.099 6.114 integrate_v_rspace 110 11.3 0.003 0.003 6.073 6.088 prepare_preconditioner 11 7.9 0.000 0.000 5.595 5.617 make_preconditioner 11 8.9 0.000 0.000 5.595 5.617 qs_ot_get_p 110 10.4 0.010 0.041 5.329 5.483 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.402 5.409 calculate_rho_elec 110 8.6 0.058 0.058 5.401 5.409 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.914 5.343 apply_single 110 13.6 0.000 0.000 4.914 5.343 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.084 5.312 make_m2s 4110 13.4 0.049 0.050 4.205 4.543 make_images 4110 14.4 0.397 0.524 4.089 4.425 ot_diis_step 99 11.5 0.011 0.011 4.368 4.368 multiply_cannon_sync_h2d 16440 15.4 3.264 3.885 3.264 3.885 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.130 3.766 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.628 3.631 grid_integrate_task_list 110 12.3 3.202 3.436 3.202 3.436 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.266 3.267 calculate_first_density_matrix 1 7.0 0.000 0.000 3.196 3.197 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.038 3.039 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 2.987 2.992 density_rs2pw 110 9.6 0.004 0.004 2.666 2.964 wfi_extrapolate 11 7.9 0.001 0.001 2.860 2.860 make_images_data 4110 15.4 0.045 0.050 2.385 2.805 hybrid_alltoall_any 4261 16.3 0.109 0.380 2.093 2.780 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.728 2.740 cp_fm_diag_elpa_base 48 14.0 2.688 2.707 2.727 2.738 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.686 2.738 calculate_dm_sparse 110 9.5 0.001 0.001 2.709 2.737 fft_wrap_pw1pw2_140 451 12.1 0.126 0.131 2.576 2.584 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.400 2.479 multiply_cannon_metrocomm4 14385 15.4 0.047 0.053 0.883 2.372 grid_collocate_task_list 110 9.6 2.220 2.366 2.220 2.366 cp_fm_cholesky_invert 11 10.9 2.335 2.347 2.335 2.347 fft3d_ps 1111 13.6 1.080 1.088 2.301 2.316 mp_irecv_dv 48980 15.7 0.808 2.240 0.808 2.240 mp_sum_l 6594 12.7 1.476 2.136 1.476 2.136 qs_energies_init_hamiltonians 11 5.9 0.055 0.091 2.048 2.049 mp_alltoall_d11v 2046 13.8 1.719 2.044 1.719 2.044 potential_pw2rs 110 12.3 0.010 0.010 1.965 1.970 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.654 1.667 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.640 1.658 dbcsr_complete_redistribute 325 12.2 0.339 0.363 1.349 1.645 acc_transpose_blocks 16440 15.4 0.077 0.080 1.588 1.636 mp_allgather_i34 2055 14.4 0.448 1.549 0.448 1.549 transfer_rs2pw 451 10.6 0.005 0.006 1.219 1.516 mp_waitany 17072 13.8 1.127 1.492 1.127 1.492 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.365 1.455 cp_fm_cholesky_decompose 22 10.9 1.309 1.331 1.309 1.331 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.259 1.264 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.136000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.272727, yerr=9.845081 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 745.725952E+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 0.098 0.190 64.147 64.151 qs_mol_dyn_low 1 2.0 0.012 0.077 63.470 63.479 qs_forces 11 3.9 0.003 0.003 63.269 63.270 qs_energies 11 4.9 0.007 0.049 59.876 59.891 scf_env_do_scf 11 5.9 0.003 0.023 51.033 51.038 scf_env_do_scf_inner_loop 99 6.5 0.020 0.035 41.543 41.547 velocity_verlet 10 3.0 0.010 0.073 34.772 34.781 dbcsr_multiply_generic 2055 12.4 0.121 0.123 31.167 31.413 qs_scf_new_mos 99 7.5 0.001 0.001 27.239 27.341 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.238 27.340 ot_scf_mini 99 9.5 0.003 0.003 25.557 25.650 multiply_cannon 2055 13.4 0.237 0.255 24.071 25.083 multiply_cannon_loop 2055 14.4 1.415 1.491 22.660 23.324 ot_mini 99 10.5 0.001 0.001 14.844 14.957 multiply_cannon_multrec 24660 15.4 4.109 6.840 13.392 14.552 rebuild_ks_matrix 110 8.3 0.000 0.000 11.947 12.055 qs_ks_build_kohn_sham_matrix 110 9.3 0.032 0.065 11.946 12.054 qs_ot_get_derivative 99 11.5 0.001 0.001 10.688 10.780 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.561 10.651 dbcsr_mm_accdrv_process 52304 16.0 7.943 9.315 9.127 10.106 init_scf_loop 11 6.9 0.006 0.045 9.447 9.447 prepare_preconditioner 11 7.9 0.000 0.000 7.478 7.509 make_preconditioner 11 8.9 0.002 0.015 7.478 7.509 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.341 7.170 sum_up_and_integrate 110 10.3 0.002 0.003 6.077 6.094 integrate_v_rspace 110 11.3 0.003 0.003 6.050 6.067 mp_waitall_1 121746 16.5 4.203 6.064 4.203 6.064 init_scf_run 11 5.9 0.000 0.001 6.025 6.025 scf_env_initial_rho_setup 11 6.9 0.002 0.011 6.024 6.025 qs_ot_get_p 110 10.4 0.022 0.057 5.644 5.797 make_m2s 4110 13.4 0.060 0.062 5.407 5.791 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.648 5.662 calculate_rho_elec 110 8.6 0.077 0.081 5.647 5.662 make_images 4110 14.4 0.576 0.696 5.265 5.646 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.036 4.130 apply_single 110 13.6 0.000 0.000 4.036 4.130 ot_diis_step 99 11.5 0.011 0.012 4.117 4.117 qs_ot_p2m_diag 48 11.0 0.059 0.102 3.714 3.762 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.597 3.654 grid_integrate_task_list 110 12.3 3.276 3.409 3.276 3.409 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.318 3.369 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.203 3.239 calculate_dm_sparse 110 9.5 0.001 0.001 3.142 3.171 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.154 3.155 hybrid_alltoall_any 4261 16.3 0.122 0.458 2.290 3.076 make_images_data 4110 15.4 0.048 0.052 2.715 3.067 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.029 3.031 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.373 3.009 density_rs2pw 110 9.6 0.010 0.054 2.754 2.924 wfi_extrapolate 11 7.9 0.003 0.020 2.922 2.922 cp_fm_upper_to_full 70 14.2 2.062 2.908 2.062 2.908 calculate_first_density_matrix 1 7.0 0.000 0.004 2.897 2.903 dbcsr_complete_redistribute 325 12.2 0.385 0.444 2.101 2.776 fft_wrap_pw1pw2_140 451 12.1 0.146 0.147 2.717 2.751 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.650 2.685 acc_transpose_blocks 24660 15.4 0.109 0.112 2.572 2.674 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.621 2.631 cp_fm_diag_elpa_base 48 14.0 2.557 2.578 2.617 2.629 fft3d_ps 1111 13.6 1.126 1.165 2.471 2.501 multiply_cannon_sync_h2d 24660 15.4 2.353 2.477 2.353 2.477 cp_fm_cholesky_invert 11 10.9 2.384 2.394 2.384 2.394 grid_collocate_task_list 110 9.6 2.265 2.375 2.265 2.375 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 2.284 2.285 copy_fm_to_dbcsr 174 11.2 0.001 0.004 1.551 2.224 mp_alltoall_d11v 2046 13.8 1.845 2.143 1.845 2.143 potential_pw2rs 110 12.3 0.012 0.013 1.883 1.892 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.823 1.864 mp_alltoall_i22 605 13.7 1.066 1.800 1.066 1.800 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.125 1.775 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.707 1.725 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.612 1.713 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.631 1.644 jit_kernel_multiply 8 15.7 0.835 1.628 0.835 1.628 mp_sum_l 6594 12.7 1.085 1.611 1.085 1.611 acc_transpose_blocks_sync 73980 16.4 1.422 1.544 1.422 1.544 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.854 1.514 mp_irecv_dv 62702 16.1 0.750 1.429 0.750 1.429 mp_allgather_i34 2055 14.4 0.489 1.405 0.489 1.405 transfer_rs2pw 451 10.6 0.005 0.006 1.213 1.399 cp_fm_cholesky_decompose 22 10.9 1.346 1.392 1.346 1.392 mp_waitany 13376 13.8 1.089 1.317 1.089 1.317 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.218 1.306 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=64.151000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=706.000000, yerr=10.929526 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 854.450176E+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 0.036 0.062 54.566 54.567 qs_mol_dyn_low 1 2.0 0.003 0.003 54.228 54.286 qs_forces 11 3.9 0.020 0.047 54.129 54.130 qs_energies 11 4.9 0.001 0.002 50.484 50.504 scf_env_do_scf 11 5.9 0.000 0.001 41.949 41.949 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.719 34.721 velocity_verlet 10 3.0 0.017 0.019 29.987 29.992 dbcsr_multiply_generic 2055 12.4 0.109 0.111 23.854 23.985 qs_scf_new_mos 99 7.5 0.001 0.001 20.630 20.698 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.629 20.697 multiply_cannon 2055 13.4 0.236 0.244 18.348 19.583 ot_scf_mini 99 9.5 0.002 0.002 19.382 19.437 multiply_cannon_loop 2055 14.4 0.601 0.619 17.091 17.301 rebuild_ks_matrix 110 8.3 0.000 0.000 11.625 11.677 qs_ks_build_kohn_sham_matrix 110 9.3 0.028 0.038 11.625 11.677 ot_mini 99 10.5 0.001 0.001 10.856 10.911 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.365 10.414 multiply_cannon_multrec 8220 15.4 3.229 4.493 7.877 8.845 mp_waitall_1 103326 16.6 5.959 7.669 5.959 7.669 qs_ot_get_derivative 99 11.5 0.001 0.001 7.132 7.187 init_scf_loop 11 6.9 0.000 0.000 7.184 7.185 sum_up_and_integrate 110 10.3 0.001 0.002 6.014 6.028 integrate_v_rspace 110 11.3 0.003 0.003 5.987 6.002 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.610 5.623 calculate_rho_elec 110 8.6 0.115 0.116 5.610 5.623 dbcsr_mm_accdrv_process 17442 15.9 3.199 4.347 4.507 5.480 init_scf_run 11 5.9 0.000 0.001 5.442 5.442 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.442 5.442 prepare_preconditioner 11 7.9 0.000 0.000 5.361 5.373 make_preconditioner 11 8.9 0.000 0.000 5.361 5.373 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.966 5.041 qs_ot_get_p 110 10.4 0.001 0.001 4.574 4.629 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 3.129 4.460 make_m2s 4110 13.4 0.038 0.039 4.152 4.404 make_images 4110 14.4 0.644 0.707 4.022 4.276 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.808 3.909 apply_single 110 13.6 0.000 0.000 3.808 3.909 ot_diis_step 99 11.5 0.012 0.012 3.703 3.703 grid_integrate_task_list 110 12.3 3.384 3.471 3.384 3.471 fft_wrap_pw1pw2 1111 11.6 0.014 0.015 3.293 3.306 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.228 3.231 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.892 2.893 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.873 2.875 fft_wrap_pw1pw2_140 451 12.1 0.192 0.194 2.844 2.861 calculate_first_density_matrix 1 7.0 0.000 0.000 2.831 2.832 density_rs2pw 110 9.6 0.004 0.004 2.640 2.812 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.678 2.697 calculate_dm_sparse 110 9.5 0.001 0.001 2.593 2.634 hybrid_alltoall_any 4261 16.3 0.200 0.849 2.180 2.602 make_images_data 4110 15.4 0.040 0.046 2.213 2.545 cp_fm_cholesky_invert 11 10.9 2.527 2.535 2.527 2.535 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.506 2.519 wfi_extrapolate 11 7.9 0.001 0.001 2.517 2.517 multiply_cannon_sync_h2d 8220 15.4 2.361 2.482 2.361 2.482 fft3d_ps 1111 13.6 1.268 1.296 2.453 2.464 grid_collocate_task_list 110 9.6 2.365 2.448 2.365 2.448 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.375 2.382 cp_fm_diag_elpa_base 48 14.0 2.320 2.345 2.373 2.381 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.046 2.102 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.785 2.007 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.885 1.916 potential_pw2rs 110 12.3 0.015 0.015 1.812 1.815 mp_alltoall_d11v 2046 13.8 1.646 1.806 1.646 1.806 acc_transpose_blocks 8220 15.4 0.039 0.040 1.626 1.674 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.503 1.635 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.588 1.592 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.551 1.563 dbcsr_complete_redistribute 325 12.2 0.516 0.528 1.418 1.497 mp_allgather_i34 2055 14.4 0.429 1.485 0.429 1.485 cp_fm_cholesky_decompose 22 10.9 1.412 1.440 1.412 1.440 qs_create_task_list 11 7.9 0.001 0.001 1.211 1.313 generate_qs_task_list 11 8.9 0.376 0.444 1.210 1.312 transfer_rs2pw 451 10.6 0.005 0.006 1.110 1.288 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.476 1.271 mp_waitany 9240 13.8 1.062 1.261 1.062 1.261 mp_irecv_dv 24056 15.7 0.449 1.226 0.449 1.226 jit_kernel_multiply 6 15.6 0.986 1.220 0.986 1.220 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.136 1.172 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.809 1.130 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.567000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=804.909091, yerr=11.579406 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 1.347023E+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 0.055 0.060 78.928 78.929 qs_mol_dyn_low 1 2.0 0.003 0.003 78.504 78.521 qs_forces 11 3.9 0.003 0.003 78.406 78.407 qs_energies 11 4.9 0.003 0.003 74.258 74.260 scf_env_do_scf 11 5.9 0.001 0.001 63.821 63.822 velocity_verlet 10 3.0 0.008 0.034 48.054 48.060 scf_env_do_scf_inner_loop 99 6.5 0.031 0.034 43.455 43.456 dbcsr_multiply_generic 2055 12.4 0.124 0.126 30.149 30.333 qs_scf_new_mos 99 7.5 0.001 0.001 26.577 26.673 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.576 26.672 ot_scf_mini 99 9.5 0.002 0.002 24.794 24.873 multiply_cannon 2055 13.4 0.336 0.358 22.974 23.578 multiply_cannon_loop 2055 14.4 0.825 0.838 21.167 21.580 init_scf_loop 11 6.9 0.000 0.000 20.295 20.296 prepare_preconditioner 11 7.9 0.000 0.000 18.281 18.293 make_preconditioner 11 8.9 0.000 0.000 18.281 18.293 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.469 17.798 ot_mini 99 10.5 0.001 0.001 14.327 14.406 rebuild_ks_matrix 110 8.3 0.000 0.001 13.423 13.519 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.423 13.519 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.180 12.266 cp_fm_upper_to_full 70 14.2 8.267 11.711 8.267 11.711 multiply_cannon_multrec 8220 15.4 4.136 4.329 10.038 10.178 qs_ot_get_derivative 99 11.5 0.001 0.001 9.553 9.634 mp_waitall_1 84994 16.7 7.590 8.584 7.590 8.584 dbcsr_complete_redistribute 325 12.2 0.937 0.954 5.338 7.353 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.792 6.827 calculate_rho_elec 110 8.6 0.227 0.227 6.792 6.826 sum_up_and_integrate 110 10.3 0.002 0.002 6.538 6.552 integrate_v_rspace 110 11.3 0.004 0.004 6.509 6.524 copy_fm_to_dbcsr 174 11.2 0.001 0.001 4.331 6.334 init_scf_run 11 5.9 0.000 0.001 6.085 6.085 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.085 6.085 dbcsr_mm_accdrv_process 11614 15.7 3.883 4.159 5.753 6.004 make_m2s 4110 13.4 0.043 0.043 5.287 5.808 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 3.796 5.787 make_images 4110 14.4 0.879 0.930 5.098 5.618 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.986 5.534 apply_single 110 13.6 0.000 0.000 4.986 5.534 mp_alltoall_i22 605 13.7 3.459 5.505 3.459 5.505 qs_ot_get_p 110 10.4 0.001 0.001 5.194 5.293 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.945 5.275 ot_diis_step 99 11.5 0.015 0.016 4.760 4.760 fft_wrap_pw1pw2 1111 11.6 0.016 0.016 4.435 4.447 fft_wrap_pw1pw2_140 451 12.1 0.363 0.364 3.886 3.903 qs_energies_init_hamiltonians 11 5.9 0.005 0.010 3.775 3.776 cp_fm_cholesky_invert 11 10.9 3.743 3.752 3.743 3.752 grid_integrate_task_list 110 12.3 3.673 3.734 3.673 3.734 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.056 3.627 qs_ot_p2m_diag 48 11.0 0.151 0.155 3.582 3.589 hybrid_alltoall_any 4261 16.3 0.266 0.584 2.787 3.507 make_images_data 4110 15.4 0.044 0.047 2.803 3.419 calculate_dm_sparse 110 9.5 0.001 0.001 3.320 3.349 fft3d_ps 1111 13.6 1.826 1.851 3.292 3.302 density_rs2pw 110 9.6 0.004 0.004 3.272 3.286 wfi_extrapolate 11 7.9 0.001 0.001 3.207 3.207 multiply_cannon_sync_h2d 8220 15.4 3.127 3.153 3.127 3.153 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.083 3.091 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.068 3.068 calculate_first_density_matrix 1 7.0 0.000 0.000 2.769 2.770 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.633 2.676 grid_collocate_task_list 110 9.6 2.662 2.675 2.662 2.675 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.568 2.608 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.492 2.492 cp_fm_diag_elpa_base 48 14.0 2.197 2.272 2.491 2.491 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.285 2.348 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.126 2.228 potential_pw2rs 110 12.3 0.021 0.021 2.088 2.096 mp_alltoall_d11v 2046 13.8 1.986 2.045 1.986 2.045 qs_create_task_list 11 7.9 0.036 0.100 1.976 2.022 generate_qs_task_list 11 8.9 0.730 0.783 1.940 1.985 jit_kernel_multiply 10 15.2 1.672 1.932 1.672 1.932 acc_transpose_blocks 8220 15.4 0.041 0.041 1.844 1.874 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.843 1.852 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.836 1.852 cp_fm_cholesky_decompose 22 10.9 1.723 1.770 1.723 1.770 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.721 1.768 qs_ot_get_orbitals 99 10.5 0.000 0.000 1.596 1.608 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=78.929000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1220.909091, yerr=48.023582 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_performance_tests/15/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 1.094965E+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 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 633.868288E+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 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3999 67369. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2253977. 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 0.063 0.073 205.454 205.455 qs_mol_dyn_low 1 2.0 0.006 0.023 204.949 204.963 qs_forces 11 3.9 0.004 0.005 204.624 204.626 qs_energies 11 4.9 0.002 0.003 199.006 199.019 scf_env_do_scf 11 5.9 0.001 0.001 182.222 182.239 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 163.255 163.257 dbcsr_multiply_generic 2507 12.6 0.179 0.182 126.531 127.335 qs_scf_new_mos 117 7.6 0.001 0.001 124.169 124.467 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.169 124.466 velocity_verlet 10 3.0 0.009 0.059 122.199 122.205 ot_scf_mini 117 9.6 0.003 0.003 117.450 117.698 multiply_cannon 2507 13.6 0.236 0.247 102.237 104.490 multiply_cannon_loop 2507 14.6 2.398 2.457 100.085 102.393 ot_mini 117 10.6 0.001 0.001 66.504 66.743 multiply_cannon_multrec 60168 15.6 31.841 33.916 41.802 43.871 qs_ot_get_derivative 117 11.6 0.001 0.001 41.653 41.898 rebuild_ks_matrix 128 8.3 0.001 0.001 33.891 34.614 qs_ks_build_kohn_sham_matrix 128 9.3 0.014 0.017 33.890 34.613 mp_waitall_1 267128 16.5 29.284 31.761 29.284 31.761 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.489 31.125 qs_ot_get_p 128 10.4 0.004 0.029 28.647 28.979 multiply_cannon_sync_h2d 60168 15.6 26.456 28.935 26.456 28.935 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.363 24.913 apply_single 128 13.6 0.001 0.001 24.363 24.913 ot_diis_step 117 11.6 0.007 0.008 24.579 24.581 qs_ot_p2m_diag 83 11.4 0.079 0.091 21.876 21.974 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.660 19.886 cp_dbcsr_syevd 83 12.4 0.005 0.005 18.978 18.979 init_scf_loop 11 6.9 0.000 0.001 18.887 18.888 multiply_cannon_metrocomm3 60168 15.6 0.118 0.123 16.366 18.007 cp_fm_diag_elpa 83 13.4 0.000 0.000 15.935 15.965 cp_fm_diag_elpa_base 83 14.4 15.884 15.919 15.928 15.957 make_m2s 5014 13.6 0.106 0.114 13.956 14.440 make_images 5014 14.6 0.399 0.416 13.770 14.264 prepare_preconditioner 11 7.9 0.000 0.000 14.037 14.099 make_preconditioner 11 8.9 0.000 0.000 14.037 14.099 sum_up_and_integrate 128 10.3 0.002 0.004 13.708 13.724 integrate_v_rspace 128 11.3 0.003 0.004 13.649 13.669 make_full_inverse_cholesky 11 9.9 0.000 0.000 13.281 13.453 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.624 12.805 calculate_rho_elec 128 8.7 0.045 0.064 12.624 12.804 init_scf_run 11 5.9 0.000 0.001 12.577 12.577 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.576 12.577 mp_sum_l 7950 12.9 9.084 10.376 9.084 10.376 dbcsr_mm_accdrv_process 124484 16.2 4.709 4.858 9.524 10.048 wfi_extrapolate 11 7.9 0.001 0.001 8.890 8.890 calculate_dm_sparse 128 9.5 0.001 0.001 8.649 8.745 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.540 8.665 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.181 8.289 multiply_cannon_metrocomm1 60168 15.6 0.094 0.098 6.371 8.111 cp_fm_cholesky_invert 11 10.9 7.910 7.928 7.910 7.928 make_images_data 5014 15.6 0.066 0.072 6.875 7.813 hybrid_alltoall_any 5200 16.5 0.294 2.257 6.030 7.546 grid_integrate_task_list 128 12.3 7.068 7.483 7.068 7.483 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.829 6.841 density_rs2pw 128 9.7 0.005 0.006 6.246 6.811 fft_wrap_pw1pw2 1291 11.7 0.018 0.026 6.487 6.757 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.903 6.001 fft_wrap_pw1pw2_140 523 12.2 0.233 0.243 5.592 5.760 fft3d_ps 1291 13.7 2.090 2.601 5.309 5.517 mp_alltoall_d11v 2415 14.1 4.473 5.439 4.473 5.439 grid_collocate_task_list 128 9.7 4.842 5.198 4.842 5.198 mp_sum_d 4467 12.1 3.680 4.580 3.680 4.580 potential_pw2rs 128 12.3 0.009 0.010 4.207 4.222 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.455000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.363636, yerr=7.737426 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 850.112512E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4008 67670. MP_Allreduce 11081 960. MP_Sync 86 MP_Alltoall 1955 6640307. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.043 182.594 182.596 qs_mol_dyn_low 1 2.0 0.016 0.080 182.067 182.083 qs_forces 11 3.9 0.011 0.054 181.858 181.860 qs_energies 11 4.9 0.012 0.082 175.045 175.056 scf_env_do_scf 11 5.9 0.001 0.002 157.689 157.705 scf_env_do_scf_inner_loop 116 6.6 0.004 0.014 131.192 131.195 velocity_verlet 10 3.0 0.015 0.059 111.674 111.681 dbcsr_multiply_generic 2485 12.5 0.186 0.193 97.180 98.360 qs_scf_new_mos 116 7.6 0.001 0.001 93.305 93.811 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.304 93.810 ot_scf_mini 116 9.6 0.004 0.010 88.524 89.128 multiply_cannon 2485 13.5 0.498 0.560 77.035 81.978 multiply_cannon_loop 2485 14.5 1.554 1.618 73.659 76.392 ot_mini 116 10.6 0.001 0.001 50.022 50.604 mp_waitall_1 212858 16.6 24.011 39.933 24.011 39.933 multiply_cannon_multrec 29820 15.5 20.610 25.756 31.386 37.012 rebuild_ks_matrix 127 8.3 0.001 0.001 32.207 32.961 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.022 32.206 32.960 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.966 29.664 multiply_cannon_metrocomm3 29820 15.5 0.096 0.102 15.718 29.564 qs_ot_get_derivative 116 11.6 0.001 0.002 28.210 28.805 init_scf_loop 11 6.9 0.005 0.041 26.398 26.400 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.937 23.015 apply_single 127 13.6 0.001 0.001 21.936 23.015 prepare_preconditioner 11 7.9 0.000 0.000 21.871 21.945 make_preconditioner 11 8.9 0.000 0.001 21.871 21.945 ot_diis_step 116 11.6 0.014 0.015 21.633 21.635 qs_ot_get_p 127 10.4 0.001 0.002 20.691 21.430 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.717 21.153 multiply_cannon_sync_h2d 29820 15.5 17.801 20.452 17.801 20.452 qs_ot_p2m_diag 82 11.4 0.186 0.214 15.767 15.800 make_m2s 4970 13.5 0.087 0.092 14.085 15.779 make_images 4970 14.5 1.150 1.343 13.870 15.560 cp_dbcsr_syevd 82 12.4 0.005 0.006 14.584 14.585 sum_up_and_integrate 127 10.3 0.010 0.045 13.609 13.670 integrate_v_rspace 127 11.3 0.003 0.004 13.541 13.568 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.805 12.841 calculate_rho_elec 127 8.7 0.087 0.105 12.805 12.840 init_scf_run 11 5.9 0.000 0.001 12.133 12.135 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.133 12.135 cp_fm_cholesky_invert 11 10.9 12.004 12.040 12.004 12.040 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.116 11.535 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.243 11.274 cp_fm_diag_elpa_base 82 14.4 11.067 11.101 11.232 11.260 multiply_cannon_metrocomm4 27335 15.5 0.103 0.119 3.903 11.042 dbcsr_mm_accdrv_process 61726 16.2 5.434 6.140 10.224 10.825 mp_irecv_dv 68888 16.3 3.696 10.633 3.696 10.633 make_images_data 4970 15.5 0.065 0.075 8.292 10.416 hybrid_alltoall_any 5155 16.4 0.344 1.493 7.100 9.789 wfi_extrapolate 11 7.9 0.002 0.005 8.201 8.201 grid_integrate_task_list 127 12.3 7.111 7.557 7.111 7.557 fft_wrap_pw1pw2 1281 11.7 0.019 0.024 7.329 7.405 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.336 7.189 density_rs2pw 127 9.7 0.006 0.006 6.465 6.915 calculate_dm_sparse 127 9.5 0.001 0.002 6.589 6.738 fft_wrap_pw1pw2_140 519 12.2 0.248 0.257 6.445 6.525 mp_sum_l 7884 12.9 4.385 6.401 4.385 6.401 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.248 6.260 fft3d_ps 1281 13.7 2.758 2.961 5.782 5.847 cp_fm_cholesky_decompose 22 10.9 5.643 5.716 5.643 5.716 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.309 5.499 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.337 5.412 grid_collocate_task_list 127 9.7 4.964 5.317 4.964 5.317 mp_alltoall_d11v 2401 14.1 4.576 5.195 4.576 5.195 mp_allgather_i34 2485 14.5 1.910 4.937 1.910 4.937 potential_pw2rs 127 12.3 0.014 0.015 4.141 4.154 mp_sum_d 4453 12.1 2.686 4.036 2.686 4.036 dbcsr_complete_redistribute 393 12.7 0.717 0.880 3.327 4.019 calculate_first_density_matrix 1 7.0 0.001 0.004 3.740 3.744 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.652 3.714 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.668 3.671 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=182.596000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=810.363636, yerr=2.012359 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 2.951022E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 4017216 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.3 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.031799E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1052064 MPI messages size (bytes): total size 2.737208E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.601750E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 284544 37295751168 131072 < size <= 4194304 665856 1004519030784 4194304 < size <= 16777216 66080 937890532432 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4050 66975. MP_Allreduce 11184 996. MP_Sync 88 MP_Alltoall 1724 9394234. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.080 0.140 171.641 171.643 qs_mol_dyn_low 1 2.0 0.011 0.043 170.692 170.730 qs_forces 11 3.9 0.004 0.006 170.555 170.558 qs_energies 11 4.9 0.009 0.058 164.022 164.030 scf_env_do_scf 11 5.9 0.001 0.006 147.711 147.711 scf_env_do_scf_inner_loop 118 6.6 0.004 0.008 118.707 118.708 velocity_verlet 10 3.0 0.002 0.006 107.579 107.582 dbcsr_multiply_generic 2529 12.6 0.190 0.194 82.746 83.743 qs_scf_new_mos 118 7.6 0.001 0.001 81.809 82.174 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.809 82.173 ot_scf_mini 118 9.6 0.003 0.005 77.594 77.993 multiply_cannon 2529 13.6 0.508 0.529 62.726 67.101 multiply_cannon_loop 2529 14.6 1.142 1.199 60.057 62.762 ot_mini 118 10.6 0.001 0.001 42.900 43.317 mp_waitall_1 172006 16.6 25.354 35.015 25.354 35.015 rebuild_ks_matrix 129 8.3 0.001 0.001 30.536 31.077 qs_ks_build_kohn_sham_matrix 129 9.3 0.031 0.075 30.535 31.076 init_scf_loop 11 6.9 0.001 0.003 28.899 28.900 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.572 28.066 multiply_cannon_multrec 20232 15.6 13.091 16.229 22.780 26.123 multiply_cannon_metrocomm3 20232 15.6 0.060 0.063 15.677 25.065 prepare_preconditioner 11 7.9 0.000 0.000 24.804 24.864 make_preconditioner 11 8.9 0.000 0.001 24.804 24.864 make_full_inverse_cholesky 11 9.9 0.000 0.000 22.822 23.762 qs_ot_get_derivative 118 11.6 0.001 0.002 23.079 23.493 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.808 20.887 apply_single 129 13.6 0.001 0.001 19.808 20.886 qs_ot_get_p 129 10.4 0.005 0.030 20.060 20.548 ot_diis_step 118 11.6 0.018 0.018 19.709 19.710 make_m2s 5058 13.6 0.079 0.082 14.979 15.762 qs_ot_p2m_diag 84 11.4 0.269 0.275 15.661 15.671 make_images 5058 14.6 1.146 1.242 14.742 15.522 multiply_cannon_sync_h2d 20232 15.6 13.590 15.428 13.590 15.428 cp_dbcsr_syevd 84 12.4 0.005 0.006 14.572 14.572 sum_up_and_integrate 129 10.3 0.002 0.003 13.805 13.835 integrate_v_rspace 129 11.3 0.003 0.004 13.745 13.777 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.064 13.105 calculate_rho_elec 129 8.7 0.133 0.148 13.063 13.104 cp_fm_cholesky_invert 11 10.9 11.805 11.825 11.805 11.825 cp_fm_diag_elpa 84 13.4 0.000 0.000 11.138 11.157 cp_fm_diag_elpa_base 84 14.4 10.887 10.964 11.134 11.152 init_scf_run 11 5.9 0.000 0.001 10.999 10.999 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.998 10.999 make_images_data 5058 15.6 0.063 0.073 9.042 10.193 hybrid_alltoall_any 5245 16.5 0.455 2.090 7.798 9.567 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.178 9.490 multiply_cannon_metrocomm4 17703 15.6 0.067 0.077 3.540 9.479 dbcsr_mm_accdrv_process 41846 16.2 5.965 6.266 9.136 9.334 mp_irecv_dv 50659 16.2 3.407 9.219 3.407 9.219 grid_integrate_task_list 129 12.3 7.357 7.808 7.357 7.808 fft_wrap_pw1pw2 1301 11.7 0.019 0.025 7.596 7.708 wfi_extrapolate 11 7.9 0.004 0.029 7.403 7.403 fft_wrap_pw1pw2_140 527 12.2 0.289 0.297 6.661 6.787 density_rs2pw 129 9.7 0.005 0.006 6.480 6.769 cp_fm_cholesky_decompose 22 10.9 6.007 6.052 6.007 6.052 calculate_dm_sparse 129 9.5 0.001 0.001 5.966 6.036 fft3d_ps 1301 13.7 2.795 3.021 5.883 5.971 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.757 5.769 dbcsr_complete_redistribute 397 12.7 1.036 1.097 4.465 5.630 grid_collocate_task_list 129 9.7 5.199 5.585 5.199 5.585 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.701 5.299 mp_alltoall_d11v 2429 14.1 4.614 5.144 4.614 5.144 mp_sum_l 8016 12.9 3.367 4.981 3.367 4.981 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.772 4.925 cp_fm_upper_to_full 106 14.8 3.925 4.922 3.925 4.922 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.226 4.388 mp_allgather_i34 2529 14.6 1.159 4.327 1.159 4.327 potential_pw2rs 129 12.3 0.020 0.022 4.173 4.186 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.023 4.045 qs_energies_init_hamiltonians 11 5.9 0.008 0.024 3.978 3.980 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.481 3.533 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=171.643000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=979.272727, yerr=5.785419 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 1.091842E+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 0.039 177.375 177.376 qs_mol_dyn_low 1 2.0 0.010 0.057 177.001 177.014 qs_forces 11 3.9 0.004 0.004 176.775 176.787 qs_energies 11 4.9 0.013 0.091 169.649 169.694 scf_env_do_scf 11 5.9 0.008 0.061 152.390 152.411 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 115.132 115.134 velocity_verlet 10 3.0 0.005 0.033 114.971 114.982 dbcsr_multiply_generic 2507 12.6 0.197 0.221 81.325 82.229 qs_scf_new_mos 117 7.6 0.001 0.001 79.526 79.854 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.525 79.853 ot_scf_mini 117 9.6 0.003 0.004 74.968 75.290 multiply_cannon 2507 13.6 0.552 0.581 56.521 59.310 multiply_cannon_loop 2507 14.6 1.839 1.934 52.839 54.816 ot_mini 117 10.6 0.001 0.001 43.105 43.409 init_scf_loop 11 6.9 0.011 0.086 37.127 37.140 prepare_preconditioner 11 7.9 0.000 0.000 32.799 32.869 make_preconditioner 11 8.9 0.006 0.049 32.799 32.869 make_full_inverse_cholesky 11 9.9 0.017 0.024 27.989 31.637 multiply_cannon_multrec 30084 15.6 13.681 18.683 26.844 31.481 rebuild_ks_matrix 128 8.3 0.001 0.001 29.319 29.611 qs_ks_build_kohn_sham_matrix 128 9.3 0.056 0.155 29.318 29.611 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.471 26.731 mp_waitall_1 147882 16.7 17.073 26.141 17.073 26.141 qs_ot_get_derivative 117 11.6 0.001 0.002 23.316 23.644 make_m2s 5014 13.6 0.094 0.099 20.302 21.216 make_images 5014 14.6 1.987 2.231 19.992 20.912 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.270 19.836 apply_single 128 13.6 0.001 0.001 19.270 19.835 ot_diis_step 117 11.6 0.017 0.019 19.659 19.661 qs_ot_get_p 128 10.4 0.001 0.002 17.999 18.342 multiply_cannon_metrocomm3 30084 15.6 0.048 0.051 6.327 14.898 qs_ot_p2m_diag 83 11.4 0.343 0.390 13.810 13.866 sum_up_and_integrate 128 10.3 0.002 0.003 13.585 13.614 integrate_v_rspace 128 11.3 0.003 0.004 13.525 13.557 dbcsr_mm_accdrv_process 62264 16.2 8.467 9.359 12.733 13.243 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.100 13.143 calculate_rho_elec 128 8.7 0.174 0.190 13.100 13.143 cp_fm_upper_to_full 105 14.8 8.751 12.487 8.751 12.487 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.414 12.416 make_images_data 5014 15.6 0.065 0.070 10.795 12.232 cp_fm_cholesky_invert 11 10.9 11.918 11.940 11.918 11.940 multiply_cannon_sync_h2d 30084 15.6 10.457 11.330 10.457 11.330 hybrid_alltoall_any 5200 16.5 0.530 2.215 9.809 11.316 init_scf_run 11 5.9 0.000 0.001 11.199 11.200 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.198 11.200 dbcsr_complete_redistribute 395 12.7 1.409 1.502 7.556 10.453 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.795 10.022 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.221 9.242 cp_fm_diag_elpa_base 83 14.4 8.739 8.870 9.215 9.229 copy_fm_to_dbcsr 209 11.7 0.002 0.002 6.210 9.048 grid_integrate_task_list 128 12.3 7.533 7.960 7.533 7.960 fft_wrap_pw1pw2 1291 11.7 0.018 0.023 7.691 7.750 transfer_fm_to_dbcsr 11 9.9 0.003 0.008 4.781 7.565 wfi_extrapolate 11 7.9 0.001 0.001 7.377 7.377 mp_alltoall_i22 716 14.1 4.321 7.251 4.321 7.251 fft_wrap_pw1pw2_140 523 12.2 0.324 0.334 6.825 6.913 multiply_cannon_metrocomm4 25070 15.6 0.079 0.088 2.785 6.798 calculate_dm_sparse 128 9.5 0.001 0.001 6.506 6.604 mp_irecv_dv 76098 16.2 2.630 6.528 2.630 6.528 density_rs2pw 128 9.7 0.006 0.013 6.163 6.406 cp_fm_cholesky_decompose 22 10.9 5.970 6.024 5.970 6.024 fft3d_ps 1291 13.7 2.933 2.992 5.887 5.942 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.778 5.811 grid_collocate_task_list 128 9.7 5.291 5.517 5.291 5.517 mp_alltoall_d11v 2415 14.1 4.862 5.199 4.862 5.199 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.678 4.767 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.497 4.627 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 4.572 4.573 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.210 4.274 potential_pw2rs 128 12.3 0.022 0.022 3.858 3.869 dbcsr_special_finalize 12535 15.6 0.066 0.068 3.502 3.758 dbcsr_merge_single_wm 12535 16.6 0.322 0.340 3.411 3.667 mp_sum_l 7950 12.9 2.578 3.637 2.578 3.637 calculate_first_density_matrix 1 7.0 0.000 0.002 3.613 3.619 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=177.376000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1034.909091, yerr=12.936590 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.409130E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108580288 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 759871. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4014 67559. MP_Allreduce 11095 1166. MP_Sync 86 MP_Alltoall 1700 18828153. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.049 160.855 160.855 qs_mol_dyn_low 1 2.0 0.003 0.004 160.195 160.208 qs_forces 11 3.9 0.004 0.004 160.096 160.099 qs_energies 11 4.9 0.002 0.002 152.688 152.695 scf_env_do_scf 11 5.9 0.001 0.001 135.863 135.873 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 105.276 105.277 velocity_verlet 10 3.0 0.019 0.019 104.129 104.132 dbcsr_multiply_generic 2485 12.5 0.187 0.191 70.982 71.343 qs_scf_new_mos 116 7.6 0.001 0.001 69.617 69.720 qs_scf_loop_do_ot 116 8.6 0.001 0.001 69.616 69.719 ot_scf_mini 116 9.6 0.003 0.003 65.249 65.338 multiply_cannon 2485 13.5 0.554 0.575 52.773 56.166 multiply_cannon_loop 2485 14.5 0.799 0.844 49.883 50.627 ot_mini 116 10.6 0.001 0.001 36.810 36.887 init_scf_loop 11 6.9 0.000 0.000 30.434 30.436 mp_waitall_1 124680 16.7 24.166 30.237 24.166 30.237 rebuild_ks_matrix 127 8.3 0.001 0.001 28.672 28.772 qs_ks_build_kohn_sham_matrix 127 9.3 0.026 0.038 28.672 28.771 prepare_preconditioner 11 7.9 0.000 0.000 26.560 26.579 make_preconditioner 11 8.9 0.000 0.000 26.560 26.579 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.145 26.242 make_full_inverse_cholesky 11 9.9 0.022 0.028 24.675 24.932 multiply_cannon_multrec 9940 15.5 10.295 14.486 17.947 21.207 qs_ot_get_derivative 116 11.6 0.001 0.002 20.041 20.125 multiply_cannon_metrocomm3 9940 15.5 0.024 0.026 12.283 19.374 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 16.844 16.968 apply_single 127 13.6 0.001 0.001 16.844 16.968 ot_diis_step 116 11.6 0.019 0.020 16.701 16.701 qs_ot_get_p 127 10.4 0.029 0.040 15.953 16.035 make_m2s 4970 13.5 0.065 0.068 14.449 15.253 make_images 4970 14.5 2.141 2.548 14.142 14.938 cp_fm_cholesky_invert 11 10.9 14.585 14.595 14.585 14.595 sum_up_and_integrate 127 10.3 0.002 0.002 13.582 13.635 integrate_v_rspace 127 11.3 0.004 0.004 13.523 13.577 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.198 13.226 calculate_rho_elec 127 8.7 0.255 0.266 13.197 13.226 qs_ot_p2m_diag 82 11.4 0.490 0.495 12.268 12.285 cp_dbcsr_syevd 82 12.4 0.005 0.006 11.127 11.127 multiply_cannon_sync_h2d 9940 15.5 10.793 11.075 10.793 11.075 init_scf_run 11 5.9 0.000 0.001 10.096 10.096 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.095 10.096 hybrid_alltoall_any 5155 16.4 0.837 3.759 8.192 9.606 make_images_data 4970 15.5 0.054 0.066 8.325 9.573 cp_fm_diag_elpa 82 13.4 0.000 0.000 8.320 8.331 cp_fm_diag_elpa_base 82 14.4 8.089 8.170 8.316 8.328 dbcsr_mm_accdrv_process 20590 16.1 3.102 4.205 7.305 7.998 fft_wrap_pw1pw2 1281 11.7 0.017 0.019 7.874 7.903 grid_integrate_task_list 127 12.3 7.697 7.894 7.697 7.894 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.787 7.849 multiply_cannon_metrocomm1 9940 15.5 0.030 0.030 4.352 7.260 wfi_extrapolate 11 7.9 0.001 0.001 7.055 7.055 fft_wrap_pw1pw2_140 519 12.2 0.431 0.436 6.959 6.994 cp_fm_cholesky_decompose 22 10.9 6.460 6.532 6.460 6.532 density_rs2pw 127 9.7 0.005 0.005 5.997 6.232 calculate_dm_sparse 127 9.5 0.001 0.001 6.080 6.154 fft3d_ps 1281 13.7 3.077 3.179 5.869 5.893 grid_collocate_task_list 127 9.7 5.519 5.726 5.519 5.726 dbcsr_complete_redistribute 393 12.7 1.986 2.016 4.951 5.290 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.241 5.242 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.143 5.164 mp_alltoall_d11v 2401 14.1 4.452 4.790 4.452 4.790 mp_allgather_i34 2485 14.5 1.095 4.498 1.095 4.498 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.063 4.112 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.598 3.942 potential_pw2rs 127 12.3 0.026 0.027 3.833 3.838 multiply_cannon_metrocomm4 7455 15.5 0.025 0.027 1.728 3.816 mp_irecv_dv 28618 15.9 1.689 3.740 1.689 3.740 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.574 3.635 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.508 3.540 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.379 3.467 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.157 3.459 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.426 3.437 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=160.855000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1328.000000, yerr=19.729996 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 2.969711E+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 0.021 0.040 248.058 248.059 qs_mol_dyn_low 1 2.0 0.003 0.004 247.063 247.502 qs_forces 11 3.9 0.005 0.005 246.905 246.908 qs_energies 11 4.9 0.058 0.072 238.111 238.116 scf_env_do_scf 11 5.9 0.001 0.001 215.143 215.162 velocity_verlet 10 3.0 0.011 0.015 172.028 172.036 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 131.827 131.829 qs_scf_new_mos 118 7.6 0.001 0.001 88.711 88.785 qs_scf_loop_do_ot 118 8.6 0.001 0.001 88.710 88.784 dbcsr_multiply_generic 2535 12.6 0.245 0.256 83.260 83.780 ot_scf_mini 118 9.6 0.003 0.004 83.695 83.726 init_scf_loop 11 6.9 0.000 0.000 82.998 83.002 prepare_preconditioner 11 7.9 0.000 0.000 78.163 78.187 make_preconditioner 11 8.9 0.000 0.000 78.163 78.187 make_full_inverse_cholesky 11 9.9 0.038 0.039 61.947 75.384 multiply_cannon 2535 13.6 0.684 0.730 59.569 60.960 multiply_cannon_loop 2535 14.6 1.055 1.082 55.493 56.859 cp_fm_upper_to_full 106 14.8 34.429 49.019 34.429 49.019 ot_mini 118 10.6 0.001 0.001 44.763 44.804 rebuild_ks_matrix 129 8.3 0.001 0.001 34.110 34.172 qs_ks_build_kohn_sham_matrix 129 9.3 0.057 0.099 34.110 34.172 mp_waitall_1 104820 16.8 28.416 32.669 28.416 32.669 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.752 31.817 dbcsr_complete_redistribute 397 12.7 3.839 3.982 22.321 30.815 copy_fm_to_dbcsr 210 11.7 0.001 0.001 18.733 27.246 qs_ot_get_derivative 118 11.6 0.002 0.002 24.638 24.669 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 16.174 24.577 qs_ot_get_p 129 10.4 0.002 0.002 23.965 24.040 mp_alltoall_i22 720 14.1 13.943 22.646 13.943 22.646 make_m2s 5070 13.6 0.075 0.077 18.995 20.493 multiply_cannon_metrocomm3 10140 15.6 0.024 0.025 18.894 20.271 ot_diis_step 118 11.6 0.022 0.022 20.091 20.092 make_images 5070 14.6 3.113 3.327 18.511 20.013 qs_ot_p2m_diag 84 11.4 0.890 0.895 19.858 19.887 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.432 19.560 apply_single 129 13.6 0.001 0.001 19.431 19.559 multiply_cannon_multrec 10140 15.6 10.375 12.238 18.432 18.562 cp_fm_cholesky_invert 11 10.9 18.371 18.383 18.371 18.383 cp_dbcsr_syevd 84 12.4 0.006 0.006 18.036 18.039 sum_up_and_integrate 129 10.3 0.002 0.002 15.916 16.003 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.939 15.988 calculate_rho_elec 129 8.7 0.486 0.486 15.938 15.987 integrate_v_rspace 129 11.3 0.004 0.005 15.852 15.940 cp_fm_diag_elpa 84 13.4 0.000 0.000 14.803 14.803 cp_fm_diag_elpa_base 84 14.4 11.932 12.981 14.798 14.798 multiply_cannon_sync_h2d 10140 15.6 14.421 14.493 14.421 14.493 hybrid_alltoall_any 5257 16.5 1.333 3.098 10.989 13.223 make_images_data 5070 15.6 0.063 0.068 10.807 13.072 init_scf_run 11 5.9 0.000 0.001 12.489 12.490 scf_env_initial_rho_setup 11 6.9 0.002 0.003 12.489 12.490 fft_wrap_pw1pw2 1301 11.7 0.020 0.020 10.072 10.093 qs_ot_get_derivative_diag 78 12.4 0.003 0.003 9.853 9.873 dbcsr_mm_accdrv_process 20958 16.1 4.381 6.082 7.804 9.660 fft_wrap_pw1pw2_140 527 12.2 0.831 0.833 8.916 8.928 wfi_extrapolate 11 7.9 0.001 0.001 8.804 8.804 grid_integrate_task_list 129 12.3 8.618 8.802 8.618 8.802 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 8.164 8.165 cp_fm_cholesky_decompose 22 10.9 7.399 7.456 7.399 7.456 mp_alltoall_d11v 2429 14.1 7.315 7.433 7.315 7.433 fft3d_ps 1301 13.7 4.014 4.039 7.342 7.381 calculate_dm_sparse 129 9.5 0.001 0.001 7.167 7.202 density_rs2pw 129 9.7 0.005 0.005 6.945 6.995 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.561 6.619 grid_collocate_task_list 129 9.7 6.455 6.527 6.455 6.527 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.276 6.467 qs_env_update_s_mstruct 11 6.9 0.000 0.000 4.987 5.192 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.644 5.071 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=248.059000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2635.181818, yerr=161.656099 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 1.270743E+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 45497430. 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 0.123 0.346 86.184 86.188 qs_energies 1 2.0 0.002 0.018 84.758 84.777 ls_scf 1 3.0 0.008 0.062 83.633 83.644 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.598 72.729 multiply_cannon 111 7.7 0.017 0.020 55.803 57.057 multiply_cannon_loop 111 8.7 0.227 0.245 52.418 53.822 ls_scf_main 1 4.0 0.000 0.003 52.161 52.162 density_matrix_trs4 2 5.0 0.002 0.003 46.664 46.732 ls_scf_init_scf 1 4.0 0.000 0.001 28.390 28.394 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.161 27.217 mp_waitall_1 11031 10.9 22.235 25.105 22.235 25.105 matrix_sqrt_Newton_Schulz 2 6.5 0.003 0.014 25.051 25.071 multiply_cannon_multrec 2664 9.7 8.139 8.912 15.632 17.436 multiply_cannon_sync_h2d 2664 9.7 13.483 14.944 13.483 14.944 make_m2s 222 7.7 0.009 0.011 13.053 13.565 make_images 222 8.7 0.099 0.110 13.030 13.546 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.547 11.657 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.465 8.398 dbcsr_mm_accdrv_process 4760 10.4 0.591 0.700 7.112 8.189 make_images_data 222 9.7 0.004 0.005 7.570 8.080 hybrid_alltoall_any 227 10.6 0.216 1.831 6.532 8.010 dbcsr_mm_accdrv_process_sort 4760 11.4 6.321 7.297 6.321 7.297 calculate_norms 4752 9.8 5.507 6.035 5.507 6.035 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.041 5.138 mp_sum_l 887 5.1 3.216 4.827 3.216 4.827 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.059 4.025 mp_irecv_dv 6231 10.9 2.042 3.995 2.042 3.995 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.413 3.602 make_images_sizes 222 9.7 0.000 0.000 0.739 3.493 mp_alltoall_i44 222 10.7 0.739 3.493 0.739 3.493 arnoldi_extremal 4 6.8 0.000 0.000 3.187 3.217 arnoldi_normal_ev 4 7.8 0.001 0.004 3.187 3.217 build_subspace 16 8.4 0.009 0.012 3.093 3.096 ls_scf_post 1 4.0 0.001 0.008 3.074 3.090 ls_scf_store_result 1 5.0 0.000 0.000 2.870 2.924 dbcsr_special_finalize 555 9.7 0.005 0.006 2.376 2.830 dbcsr_merge_single_wm 555 10.7 0.453 0.608 2.368 2.821 make_images_pack 222 9.7 2.213 2.629 2.215 2.631 dbcsr_sort_data 658 11.4 2.171 2.555 2.171 2.555 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.299 2.541 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.456 2.066 2.457 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.268 2.329 buffer_matrices_ensure_size 222 8.7 1.751 2.098 1.751 2.098 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.798 1.800 rebuild_ks_matrix 3 7.3 0.000 0.000 1.788 1.791 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.004 1.788 1.791 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.188000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 2.176176E+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 0.049 0.185 90.432 90.433 qs_energies 1 2.0 0.000 0.000 89.461 89.467 ls_scf 1 3.0 0.000 0.000 87.851 87.857 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.157 74.477 multiply_cannon 111 7.7 0.027 0.039 52.881 56.803 ls_scf_main 1 4.0 0.000 0.000 54.304 54.311 multiply_cannon_loop 111 8.7 0.136 0.148 50.054 52.844 density_matrix_trs4 2 5.0 0.002 0.003 48.703 48.863 ls_scf_init_scf 1 4.0 0.000 0.000 29.876 29.877 mp_waitall_1 9105 10.9 20.739 29.657 20.739 29.657 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.707 28.774 multiply_cannon_multrec 1332 9.7 13.200 17.394 22.562 27.948 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.358 26.367 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.619 20.348 make_m2s 222 7.7 0.006 0.007 14.926 15.513 make_images 222 8.7 1.374 1.697 14.896 15.483 dbcsr_mm_accdrv_process 4041 10.4 0.329 0.524 8.957 10.443 dbcsr_mm_accdrv_process_sort 4041 11.4 8.472 9.921 8.472 9.921 make_images_data 222 9.7 0.004 0.005 8.583 9.462 hybrid_alltoall_any 227 10.6 0.543 2.550 7.944 9.360 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.172 7.836 mp_irecv_dv 3311 11.0 3.151 7.779 3.151 7.779 mp_sum_l 887 5.1 4.889 7.727 4.889 7.727 multiply_cannon_sync_h2d 1332 9.7 4.945 6.765 4.945 6.765 calculate_norms 2376 9.8 6.071 6.757 6.071 6.757 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.704 6.061 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.855 5.011 arnoldi_extremal 4 6.8 0.000 0.000 4.574 4.588 arnoldi_normal_ev 4 7.8 0.001 0.005 4.574 4.588 build_subspace 16 8.4 0.014 0.021 4.328 4.331 ls_scf_post 1 4.0 0.000 0.000 3.671 3.678 ls_scf_store_result 1 5.0 0.000 0.000 3.389 3.496 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.114 3.345 dbcsr_matrix_vector_mult_local 304 10.0 2.742 3.218 2.744 3.220 mp_allgather_i34 111 8.7 0.799 3.003 0.799 3.003 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.189 2.947 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.573 2.654 dbcsr_data_new 4174 10.1 2.130 2.426 2.130 2.426 dbcsr_sort_data 436 11.2 1.855 2.128 1.855 2.128 make_images_pack 222 9.7 1.820 2.124 1.823 2.126 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.893 1.895 rebuild_ks_matrix 3 7.3 0.000 0.000 1.881 1.883 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.881 1.882 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.433000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1774.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 2.896957E+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 0.096 0.255 92.888 92.896 qs_energies 1 2.0 0.013 0.072 91.550 91.631 ls_scf 1 3.0 0.000 0.002 89.686 89.744 dbcsr_multiply_generic 111 6.7 0.016 0.018 74.179 74.495 multiply_cannon 111 7.7 0.038 0.079 51.272 55.842 ls_scf_main 1 4.0 0.006 0.066 55.747 55.758 multiply_cannon_loop 111 8.7 0.117 0.129 48.538 52.166 density_matrix_trs4 2 5.0 0.003 0.016 49.816 49.947 mp_waitall_1 7281 11.0 22.638 32.102 22.638 32.102 ls_scf_init_scf 1 4.0 0.000 0.002 30.199 30.215 ls_scf_init_matrix_S 1 5.0 0.006 0.028 28.844 28.921 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.007 26.499 26.519 multiply_cannon_multrec 888 9.7 12.765 15.054 21.486 24.788 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.187 21.619 make_m2s 222 7.7 0.006 0.007 16.425 17.108 make_images 222 8.7 1.587 1.893 16.387 17.071 make_images_data 222 9.7 0.004 0.004 9.587 10.515 hybrid_alltoall_any 227 10.6 0.638 2.925 9.007 10.228 dbcsr_mm_accdrv_process 3754 10.4 0.315 0.512 8.216 9.505 dbcsr_mm_accdrv_process_sort 3754 11.4 7.777 8.992 7.777 8.992 mp_sum_l 887 5.1 4.891 7.803 4.891 7.803 multiply_cannon_sync_h2d 888 9.7 6.026 7.624 6.026 7.624 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.506 7.327 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.480 7.167 mp_irecv_dv 2335 11.1 2.463 7.110 2.463 7.110 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.623 5.798 arnoldi_extremal 4 6.8 0.000 0.000 5.057 5.077 arnoldi_normal_ev 4 7.8 0.001 0.005 5.057 5.077 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.768 5.018 calculate_norms 1584 9.8 4.392 4.857 4.392 4.857 build_subspace 16 8.4 0.014 0.020 4.765 4.770 mp_allgather_i34 111 8.7 0.850 3.838 0.850 3.838 ls_scf_post 1 4.0 0.012 0.063 3.740 3.798 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.427 3.770 dbcsr_matrix_vector_mult_local 304 10.0 3.021 3.600 3.023 3.602 ls_scf_store_result 1 5.0 0.000 0.000 3.438 3.519 ls_scf_dm_to_ks 2 5.0 0.003 0.049 2.831 2.933 dbcsr_data_new 4116 9.9 2.107 2.469 2.107 2.469 make_images_sizes 222 9.7 0.000 0.000 0.976 2.143 mp_alltoall_i44 222 10.7 0.976 2.143 0.976 2.143 dbcsr_sort_data 325 11.1 1.899 2.134 1.899 2.134 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.991 2.007 rebuild_ks_matrix 3 7.3 0.000 0.000 1.973 1.989 qs_ks_build_kohn_sham_matrix 3 8.3 0.007 0.036 1.973 1.989 make_images_pack 222 9.7 1.623 1.877 1.626 1.880 dbcsr_finalize 304 7.8 0.026 0.032 1.614 1.871 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.896000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2309.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 3.388510E+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 0.034 0.045 97.236 97.237 qs_energies 1 2.0 0.000 0.000 96.515 96.521 ls_scf 1 3.0 0.000 0.000 94.792 94.798 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.279 78.606 ls_scf_main 1 4.0 0.000 0.000 58.689 58.702 multiply_cannon 111 7.7 0.046 0.102 51.493 56.517 density_matrix_trs4 2 5.0 0.002 0.003 52.539 52.665 multiply_cannon_loop 111 8.7 0.155 0.175 46.402 49.612 ls_scf_init_scf 1 4.0 0.000 0.000 32.867 32.886 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.516 31.605 mp_waitall_1 6369 11.0 22.359 29.126 22.359 29.126 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.005 28.857 28.870 multiply_cannon_multrec 1332 9.7 14.207 17.172 22.235 25.018 make_m2s 222 7.7 0.007 0.008 21.132 22.613 make_images 222 8.7 3.141 3.599 21.081 22.564 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 8.794 17.354 make_images_data 222 9.7 0.004 0.004 11.788 13.620 hybrid_alltoall_any 227 10.6 0.797 3.786 11.146 12.805 dbcsr_mm_accdrv_process 3641 10.4 0.295 0.511 7.675 9.200 dbcsr_mm_accdrv_process_sort 3641 11.4 7.222 8.706 7.222 8.706 mp_sum_l 887 5.1 4.069 7.514 4.069 7.514 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.607 6.301 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.097 6.090 mp_irecv_dv 3229 10.9 2.072 6.008 2.072 6.008 multiply_cannon_sync_h2d 1332 9.7 5.384 5.917 5.384 5.917 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.108 5.762 arnoldi_extremal 4 6.8 0.000 0.000 5.168 5.180 arnoldi_normal_ev 4 7.8 0.001 0.005 5.168 5.180 build_subspace 16 8.4 0.014 0.021 4.833 4.842 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.539 4.708 mp_allgather_i34 111 8.7 2.208 4.657 2.208 4.657 calculate_norms 2376 9.8 4.158 4.536 4.158 4.536 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.561 3.859 dbcsr_matrix_vector_mult_local 304 10.0 3.187 3.685 3.190 3.687 dbcsr_sort_data 658 11.4 3.012 3.500 3.012 3.500 ls_scf_post 1 4.0 0.000 0.000 3.237 3.243 dbcsr_special_finalize 555 9.7 0.006 0.007 2.759 3.230 dbcsr_merge_single_wm 555 10.7 0.533 0.669 2.751 3.222 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.004 3.076 ls_scf_store_result 1 5.0 0.000 0.000 2.980 3.040 dbcsr_data_release 10477 10.7 1.573 2.432 1.573 2.432 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.081 2.100 rebuild_ks_matrix 3 7.3 0.000 0.000 2.058 2.076 qs_ks_build_kohn_sham_matrix 3 8.3 0.009 0.026 2.058 2.076 dbcsr_finalize 304 7.8 0.049 0.061 1.802 2.020 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.237000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2750.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 4.837335E+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 0.069 0.174 92.407 92.408 qs_energies 1 2.0 0.000 0.000 91.269 91.277 ls_scf 1 3.0 0.000 0.000 89.306 89.315 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.911 71.094 ls_scf_main 1 4.0 0.000 0.000 56.622 56.636 multiply_cannon 111 7.7 0.105 0.185 52.891 56.527 multiply_cannon_loop 111 8.7 0.093 0.158 50.297 51.589 density_matrix_trs4 2 5.0 0.002 0.003 49.681 49.746 ls_scf_init_scf 1 4.0 0.000 0.000 29.267 29.280 mp_waitall_1 5436 11.0 24.395 28.536 24.395 28.536 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.034 28.067 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.913 25.927 multiply_cannon_multrec 444 9.7 13.751 16.765 20.967 22.431 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.542 15.611 make_m2s 222 7.7 0.005 0.005 13.438 14.357 make_images 222 8.7 2.042 2.478 13.371 14.289 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.510 13.643 make_images_data 222 9.7 0.003 0.004 8.164 9.678 hybrid_alltoall_any 227 10.6 0.804 3.821 8.085 9.611 dbcsr_mm_accdrv_process 3003 10.4 0.372 0.480 6.903 8.158 multiply_cannon_sync_h2d 444 9.7 6.756 7.703 6.756 7.703 dbcsr_mm_accdrv_process_sort 3003 11.4 6.517 7.678 6.517 7.678 arnoldi_extremal 4 6.8 0.000 0.000 5.782 5.798 arnoldi_normal_ev 4 7.8 0.001 0.005 5.782 5.797 build_subspace 16 8.4 0.015 0.020 5.387 5.397 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.439 4.703 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.164 4.371 mp_sum_l 887 5.1 2.777 4.246 2.777 4.246 dbcsr_matrix_vector_mult_local 304 10.0 3.702 4.186 3.704 4.188 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.551 3.902 mp_irecv_dv 1241 11.2 1.535 3.855 1.535 3.855 mp_allgather_i34 111 8.7 1.176 3.733 1.176 3.733 calculate_norms 792 9.8 3.625 3.719 3.625 3.719 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.496 3.544 ls_scf_post 1 4.0 0.000 0.000 3.418 3.425 ls_scf_store_result 1 5.0 0.000 0.000 3.189 3.225 make_images_sizes 222 9.7 0.000 0.000 0.871 3.139 mp_alltoall_i44 222 10.7 0.870 3.138 0.870 3.138 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.766 2.988 dbcsr_finalize 304 7.8 0.062 0.077 2.206 2.295 dbcsr_data_new 4608 9.7 1.772 2.240 1.772 2.240 dbcsr_merge_all 275 8.9 0.481 0.528 2.061 2.125 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.046 2.054 rebuild_ks_matrix 3 7.3 0.000 0.000 2.013 2.021 qs_ks_build_kohn_sham_matrix 3 8.3 0.015 0.023 2.013 2.021 qs_energies_init_hamiltonians 1 3.0 0.001 0.004 1.947 1.947 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.408000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3794.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_performance_tests/26/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 30.358840E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 339931 0.0% 0.0% 100.0% average stack size 0.0 0.0 28341.7 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 8.983904E+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 0.142 0.200 105.216 105.217 qs_energies 1 2.0 0.000 0.000 103.464 103.473 ls_scf 1 3.0 0.000 0.000 100.485 100.494 dbcsr_multiply_generic 111 6.7 0.024 0.026 74.287 74.435 ls_scf_main 1 4.0 0.000 0.000 63.135 63.135 density_matrix_trs4 2 5.0 0.002 0.003 54.331 54.383 multiply_cannon 111 7.7 0.170 0.261 48.307 50.412 multiply_cannon_loop 111 8.7 0.110 0.183 45.277 45.742 ls_scf_init_scf 1 4.0 0.000 0.000 33.575 33.575 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.165 32.188 matrix_sqrt_Newton_Schulz 2 6.5 0.005 0.030 29.469 29.475 mp_waitall_1 4527 11.1 21.255 24.805 21.255 24.805 make_m2s 222 7.7 0.005 0.005 22.318 23.401 multiply_cannon_multrec 444 9.7 17.801 18.541 22.538 23.335 make_images 222 8.7 3.574 3.886 22.210 23.292 hybrid_alltoall_any 227 10.6 1.659 3.626 12.509 15.294 make_images_data 222 9.7 0.003 0.004 12.700 14.824 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.973 10.267 multiply_cannon_sync_h2d 444 9.7 8.788 8.828 8.788 8.828 arnoldi_extremal 4 6.8 0.000 0.000 7.402 7.422 arnoldi_normal_ev 4 7.8 0.003 0.009 7.402 7.422 build_subspace 16 8.4 0.026 0.036 6.808 6.821 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.446 5.619 dbcsr_matrix_vector_mult_local 304 10.0 5.034 5.348 5.037 5.351 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.873 5.124 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.922 5.013 dbcsr_mm_accdrv_process 1814 10.4 0.305 0.378 4.546 4.678 dbcsr_mm_accdrv_process_sort 1814 11.4 4.194 4.333 4.194 4.333 ls_scf_post 1 4.0 0.000 0.000 3.775 3.785 make_images_sizes 222 9.7 0.000 0.000 1.478 3.637 mp_alltoall_i44 222 10.7 1.477 3.637 1.477 3.637 ls_scf_store_result 1 5.0 0.000 0.000 3.490 3.528 mp_allgather_i34 111 8.7 1.106 3.482 1.106 3.482 calculate_norms 792 9.8 3.230 3.270 3.230 3.270 dbcsr_finalize 304 7.8 0.082 0.089 3.080 3.187 dbcsr_merge_all 275 8.9 0.884 0.924 2.864 2.964 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.949 2.949 dbcsr_complete_redistribute 5 7.6 1.438 1.493 2.745 2.859 matrix_ls_to_qs 2 6.0 0.000 0.000 2.407 2.540 dbcsr_sort_data 325 11.1 2.441 2.513 2.441 2.513 dbcsr_new_transposed 4 7.5 0.274 0.330 2.383 2.400 dbcsr_data_new 6591 9.6 1.806 2.315 1.806 2.315 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.275 2.276 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.255 dbcsr_add_anytype 103 7.2 0.860 0.893 2.132 2.255 dbcsr_frobenius_norm 74 6.6 2.061 2.138 2.201 2.233 rebuild_ks_matrix 3 7.3 0.000 0.000 2.209 2.211 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.209 2.211 dbcsr_data_release 12724 10.6 1.973 2.164 1.973 2.164 dbcsr_redistribute 4 8.5 1.383 1.442 2.082 2.114 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.217000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7207.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/be1f5882581616708fdc25ee40722957c884746f_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 586.764288E+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 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 0.024 0.072 231.020 231.021 qs_mol_dyn_low 1 2.0 0.007 0.069 229.307 229.372 qs_forces 5 3.8 0.004 0.012 229.031 229.035 qs_energies 5 4.8 0.012 0.064 226.001 226.031 scf_env_do_scf 5 5.8 0.002 0.134 205.408 205.408 scf_env_do_scf_inner_loop 105 6.6 0.008 0.057 179.903 179.904 qs_scf_new_mos 105 7.6 0.000 0.001 141.504 141.703 qs_scf_loop_do_ot 105 8.6 0.001 0.003 141.504 141.702 dbcsr_multiply_generic 1445 12.2 0.129 0.138 139.798 140.205 ot_scf_mini 105 9.6 0.003 0.005 131.600 131.773 multiply_cannon 1445 13.2 0.274 0.359 120.497 122.763 multiply_cannon_loop 1445 14.2 2.854 2.991 118.844 119.989 velocity_verlet 4 3.0 0.020 0.139 102.157 102.160 ot_mini 105 10.6 0.001 0.001 60.451 60.559 multiply_cannon_multrec 69360 15.2 29.692 34.843 39.971 45.206 mp_waitall_1 488190 16.1 35.348 43.973 35.348 43.973 qs_ot_get_p 112 10.4 0.001 0.001 41.091 41.403 qs_ot_get_derivative 55 11.6 0.001 0.014 38.777 38.899 multiply_cannon_metrocomm3 69360 15.2 0.202 0.216 26.095 35.136 multiply_cannon_sync_h2d 69360 15.2 29.063 33.474 29.063 33.474 qs_ot_p2m_diag 40 11.0 0.020 0.030 30.187 30.279 rebuild_ks_matrix 110 8.4 0.000 0.000 28.214 28.392 qs_ks_build_kohn_sham_matrix 110 9.4 0.015 0.088 28.214 28.392 cp_dbcsr_syevd 40 12.0 0.002 0.002 26.937 26.937 qs_ks_update_qs_env 112 7.6 0.001 0.001 25.879 26.048 init_scf_loop 7 6.6 0.000 0.015 25.470 25.473 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.956 23.217 apply_single 62 13.6 0.000 0.000 22.956 23.217 cp_fm_syevd 40 13.0 0.000 0.000 21.201 21.405 ot_new_cg_direction 55 11.6 0.001 0.001 20.972 20.972 prepare_preconditioner 7 7.6 0.000 0.000 20.439 20.480 make_preconditioner 7 8.6 0.000 0.015 20.439 20.480 init_scf_run 5 5.8 0.000 0.001 18.024 18.025 scf_env_initial_rho_setup 5 6.8 0.002 0.003 18.024 18.025 cp_fm_redistribute_end 40 14.0 8.295 16.530 8.309 16.536 qs_rho_update_rho_low 110 7.6 0.000 0.001 16.132 16.524 calculate_rho_elec 110 8.6 0.028 0.031 16.131 16.524 cp_fm_syevd_base 40 14.0 8.218 16.455 8.218 16.455 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.908 15.124 mp_sum_l 4764 12.2 12.923 13.718 12.923 13.718 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.601 13.715 calculate_first_density_matrix 1 7.0 0.002 0.036 13.693 13.705 cp_dbcsr_sm_fm_multiply 15 9.3 0.002 0.046 13.498 13.519 acc_transpose_blocks 69360 15.2 0.346 0.366 12.742 13.506 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 12.852 12.925 make_full_inverse_cholesky 7 9.6 0.000 0.000 12.523 12.590 calculate_dm_sparse 110 9.5 0.000 0.001 11.597 11.818 dbcsr_mm_accdrv_process 154766 15.8 6.284 6.479 10.146 10.999 density_rs2pw 110 9.6 0.005 0.109 10.421 10.812 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.665 10.753 acc_transpose_blocks_kernels 69360 16.2 0.846 0.885 9.982 10.693 fft_wrap_pw1pw2 1425 12.5 0.019 0.025 10.362 10.563 check_diag 80 13.5 8.659 9.013 9.733 9.899 jit_kernel_transpose 5 15.0 9.135 9.834 9.135 9.834 qs_vxc_create 110 10.4 0.007 0.069 9.758 9.773 fft_wrap_pw1pw2_240 915 14.0 0.201 0.247 9.084 9.262 multiply_cannon_metrocomm1 69360 15.2 0.096 0.102 5.012 8.758 cp_fm_cholesky_invert 7 10.6 8.404 8.421 8.404 8.421 sum_up_and_integrate 60 10.3 0.001 0.029 8.191 8.224 integrate_v_rspace 60 11.3 0.001 0.002 8.174 8.201 fft3d_pb 915 15.0 2.254 2.418 8.035 8.191 transfer_rs2pw 445 10.6 0.008 0.009 7.538 7.929 make_full_single_inverse 7 9.6 0.001 0.034 7.654 7.695 xc_rho_set_and_dset_create 110 12.4 0.076 0.131 6.910 7.174 make_m2s 2890 13.2 0.079 0.089 6.233 6.793 make_images 2890 14.2 0.240 0.259 6.125 6.686 xc_vxc_pw_create 60 11.3 0.036 0.038 6.525 6.549 xc_pw_derive 510 13.4 0.005 0.006 5.609 5.673 mp_alltoall_z22v 2340 16.7 5.182 5.423 5.182 5.423 mp_waitany 7680 13.5 4.562 5.151 4.562 5.151 multiply_cannon_metrocomm4 67915 15.2 0.182 0.203 2.025 4.959 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=231.021000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=558.000000, yerr=1.788854 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: be1f5882581616708fdc25ee40722957c884746f Summary: empty Status: OK