=== 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: 7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318 ################# 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 1.16.2, LIBGRPP 20231215 # # 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: 25.01.2024 # # \ 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; \ ./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; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 #USE_DEEPMD := 2.2.7 #USE_SIRIUS := 7.5.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS #LIBVDWXC_VER := 0.4.0 #SPFFT_VER := 1.0.6 #SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_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 SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/01 job id: 51787265 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/02 job id: 51787267 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/03 job id: 51787268 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/04 job id: 51787269 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/05 job id: 51787270 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/06 job id: 51787271 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/07 job id: 51787272 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/08 job id: 51787273 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/09 job id: 51787274 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/10 job id: 51787275 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/11 job id: 51787276 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/12 job id: 51787278 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/13 job id: 51787279 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/14 job id: 51787280 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/15 job id: 51787281 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/16 job id: 51787282 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/17 job id: 51787284 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/18 job id: 51787286 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/19 job id: 51787287 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/20 job id: 51787288 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/21 job id: 51787289 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/22 job id: 51787290 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/23 job id: 51787291 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/24 job id: 51787292 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/25 job id: 51787293 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/26 job id: 51787294 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/27 job id: 51787295 --- 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/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 0.018 0.035 137.982 137.983 farming_run 1 2.0 137.519 137.520 137.952 137.955 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.542832E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.005 0.024 119.798 119.798 qs_energies 1 2.0 0.000 0.000 119.612 119.614 mp2_main 1 3.0 0.000 0.000 117.582 117.585 mp2_gpw_main 1 4.0 0.029 0.037 116.592 116.594 mp2_ri_gpw_compute_in 1 5.0 0.175 0.183 97.187 97.424 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.004 57.308 57.543 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.171 43.743 48.712 get_2c_integrals 1 6.0 0.008 0.009 38.964 39.703 integrate_v_rspace 273 8.0 0.437 0.453 25.555 30.092 fft_wrap_pw1pw2 5465 10.4 0.067 0.070 27.975 28.503 grid_integrate_task_list 273 9.0 20.806 26.297 20.806 26.297 fft_wrap_pw1pw2_100 2178 11.4 1.307 1.437 25.419 25.943 compute_2c_integrals 1 7.0 0.002 0.002 21.260 21.262 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 20.760 20.975 mp2_eri_2c_integrate_gpw 1 9.0 2.389 2.433 20.758 20.971 rpa_ri_compute_en 1 5.0 0.019 0.022 19.299 19.471 cp_fm_cholesky_decompose 12 8.2 17.630 18.364 17.630 18.364 fft3d_s 5443 12.4 16.846 17.378 16.867 17.399 cholesky_decomp 1 7.0 0.000 0.000 16.548 17.282 ao_to_mo_and_store_B_mult_1 272 7.0 10.773 15.317 10.773 15.317 calculate_wavefunction 272 8.0 5.383 5.523 12.844 13.772 calc_potential_gpw 544 9.5 0.004 0.004 10.919 11.597 rpa_num_int 1 6.0 0.000 0.001 11.182 11.183 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.685 11.163 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 9.769 10.456 potential_pw2rs 545 10.0 0.106 0.108 8.399 10.081 calc_mat_Q 8 8.0 0.000 0.000 9.550 9.960 contract_S_to_Q 8 9.0 0.000 0.000 8.969 9.378 parallel_gemm_fm 14 9.1 0.000 0.000 8.544 9.029 parallel_gemm_fm_cosma 14 10.1 8.544 9.029 8.544 9.029 collocate_single_gaussian 272 10.0 0.038 0.039 7.792 8.440 create_integ_mat 1 6.0 0.022 0.028 7.788 7.788 array2fm 1 7.0 0.000 0.000 6.709 7.196 pw_poisson_solve 545 10.5 0.010 0.010 4.665 4.851 pw_scatter_s 2720 12.7 4.496 4.833 4.496 4.833 pw_gather_s 2722 12.2 3.848 4.539 3.848 4.539 pw_poisson_set 548 11.5 0.018 0.019 3.366 3.538 array2fm_buffer_send 1 8.0 2.881 3.108 2.881 3.108 pw_copy 4911 11.6 2.617 2.726 2.617 2.726 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=116.594079, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2778.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 0.025 0.036 532.175 532.176 farming_run 1 2.0 531.427 531.431 532.127 532.128 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.269567E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 19685014. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.032 326.570 326.571 qs_energies 1 2.0 0.000 0.000 326.342 326.343 mp2_main 1 3.0 0.000 0.000 210.121 210.122 mp2_gpw_main 1 4.0 0.047 0.083 208.808 208.816 mp2_ri_gpw_compute_en 1 5.0 0.038 0.042 128.741 130.792 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.762 1.931 127.038 127.044 scf_env_do_scf 1 3.0 0.000 0.000 115.882 115.882 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.835 114.842 rebuild_ks_matrix 4 6.0 0.000 0.000 114.833 114.841 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.064 114.833 114.841 hfx_ks_matrix 4 8.0 0.001 0.001 114.430 114.433 integrate_four_center 4 9.0 0.152 0.462 114.429 114.433 mp2_ri_gpw_compute_en_expansio 172 7.0 0.462 0.495 111.019 111.988 local_gemm 172 8.0 110.557 111.514 110.557 111.514 integrate_four_center_main 4 10.0 0.113 0.391 102.282 105.752 integrate_four_center_bin 266 11.0 102.169 105.724 102.169 105.724 init_scf_loop 1 4.0 0.000 0.000 97.810 97.810 mp2_ri_gpw_compute_in 1 5.0 0.068 0.078 79.915 80.978 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 56.684 57.755 mp2_eri_3c_integrate_gpw 91 7.0 0.146 0.159 44.578 49.194 integrate_v_rspace 95 8.0 0.399 0.555 28.272 32.749 fft_wrap_pw1pw2 1868 10.4 0.029 0.034 30.502 31.012 ao_to_mo_and_store_B_mult_1 91 7.0 10.435 28.982 10.435 28.982 fft_wrap_pw1pw2_100 730 11.4 0.629 0.734 28.183 28.696 grid_integrate_task_list 95 9.0 23.515 28.171 23.515 28.171 get_2c_integrals 1 6.0 0.000 0.000 23.121 23.169 compute_2c_integrals 1 7.0 0.002 0.003 22.104 22.114 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 21.678 21.937 mp2_eri_2c_integrate_gpw 1 9.0 1.732 1.860 21.677 21.935 fft3d_s 1823 12.4 19.208 19.981 19.222 19.994 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 18.068 18.068 calc_potential_gpw 182 9.5 0.002 0.002 12.367 13.130 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 10.683 11.364 calculate_wavefunction 91 8.0 2.015 2.059 10.026 10.734 potential_pw2rs 186 10.0 0.033 0.034 9.074 10.420 mp2_ri_gpw_compute_en_comm 22 7.0 0.491 0.511 8.155 9.270 collocate_single_gaussian 91 10.0 0.017 0.022 8.275 8.846 mp_sendrecv_dm3 2068 8.0 6.213 7.335 6.213 7.335 mp_sync 37 10.5 4.120 6.911 4.120 6.911 integrate_four_center_load 4 10.0 0.000 0.000 6.760 6.764 hfx_load_balance 1 11.0 0.000 0.000 6.759 6.764 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=208.802480, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1524.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 445.485056E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.056 0.095 54.665 54.667 qs_mol_dyn_low 1 2.0 0.009 0.055 54.063 54.071 qs_forces 11 3.9 0.005 0.027 53.973 53.982 qs_energies 11 4.9 0.002 0.003 52.492 52.516 scf_env_do_scf 11 5.9 0.001 0.002 46.136 46.136 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.070 44.070 qs_scf_new_mos 108 7.5 0.000 0.001 34.232 34.531 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.231 34.530 dbcsr_multiply_generic 2286 12.5 0.094 0.099 34.047 34.475 ot_scf_mini 108 9.5 0.002 0.002 32.571 32.746 multiply_cannon 2286 13.5 0.186 0.195 26.431 27.888 multiply_cannon_loop 2286 14.5 1.791 1.905 25.772 27.270 velocity_verlet 10 3.0 0.002 0.005 25.885 25.885 ot_mini 108 10.5 0.001 0.001 19.668 19.913 qs_ot_get_derivative 108 11.5 0.001 0.001 16.658 16.850 mp_waitall_1 245248 16.5 8.383 14.824 8.383 14.824 multiply_cannon_metrocomm3 54864 15.5 0.072 0.077 6.041 12.963 multiply_cannon_multrec 54864 15.5 3.695 5.753 7.887 11.319 qs_ot_get_p 119 10.4 0.001 0.001 8.271 8.595 rebuild_ks_matrix 119 8.3 0.000 0.000 7.776 7.911 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.775 7.911 mp_sum_l 7287 12.8 5.657 7.332 5.657 7.332 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.842 6.962 multiply_cannon_sync_h2d 54864 15.5 5.221 6.705 5.221 6.705 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.618 6.106 dbcsr_mm_accdrv_process 76910 16.1 1.796 2.868 4.106 5.854 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.625 5.725 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.413 5.468 init_scf_run 11 5.9 0.000 0.001 4.986 4.986 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.986 4.986 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.583 4.583 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.370 4.371 sum_up_and_integrate 119 10.3 0.001 0.002 4.352 4.359 integrate_v_rspace 119 11.3 0.002 0.002 4.341 4.350 cp_fm_redistribute_end 50 14.0 2.228 4.343 2.234 4.346 cp_fm_diag_elpa_base 50 14.0 2.105 4.238 2.110 4.247 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.030 4.134 calculate_rho_elec 119 8.7 0.012 0.025 4.029 4.133 calculate_dm_sparse 119 9.5 0.000 0.000 3.041 3.184 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.942 3.154 apply_single 119 13.6 0.000 0.000 2.942 3.154 jit_kernel_multiply 13 15.8 2.248 2.906 2.248 2.906 calculate_first_density_matrix 1 7.0 0.000 0.000 2.803 2.813 acc_transpose_blocks 54864 15.5 0.213 0.235 2.208 2.759 ot_diis_step 108 11.5 0.006 0.006 2.753 2.753 multiply_cannon_metrocomm1 54864 15.5 0.056 0.061 1.700 2.620 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.462 2.531 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.393 2.395 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.126 2.202 density_rs2pw 119 9.7 0.003 0.004 2.057 2.181 grid_integrate_task_list 119 12.3 2.019 2.116 2.019 2.116 wfi_extrapolate 11 7.9 0.001 0.001 2.109 2.109 init_scf_loop 11 6.9 0.000 0.000 2.048 2.048 mp_sum_d 4139 12.0 1.350 1.932 1.350 1.932 potential_pw2rs 119 12.3 0.004 0.004 1.735 1.745 make_m2s 4572 13.5 0.053 0.055 1.666 1.716 make_images 4572 14.5 0.132 0.138 1.584 1.634 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.544 1.565 fft_wrap_pw1pw2 1201 11.6 0.011 0.013 1.483 1.550 mp_alltoall_d11v 2130 13.8 1.303 1.466 1.303 1.466 transfer_rs2pw 487 10.6 0.006 0.007 1.370 1.466 acc_transpose_blocks_sync 164592 16.5 1.172 1.422 1.172 1.422 grid_collocate_task_list 119 9.7 1.341 1.399 1.341 1.399 mp_waitany 12084 13.8 1.252 1.386 1.252 1.386 transfer_pw2rs 487 13.2 0.005 0.006 1.312 1.321 fft3d_ps 1201 13.6 0.363 0.462 1.259 1.318 fft_wrap_pw1pw2_140 487 12.2 0.048 0.053 1.142 1.211 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.740 1.146 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.667000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=424.454545, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 481.808384E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1873097. 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 0.015 0.038 41.407 41.409 qs_mol_dyn_low 1 2.0 0.004 0.009 41.176 41.184 qs_forces 11 3.9 0.002 0.002 41.114 41.115 qs_energies 11 4.9 0.002 0.002 39.369 39.375 scf_env_do_scf 11 5.9 0.001 0.001 34.078 34.079 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 31.354 31.354 dbcsr_multiply_generic 2286 12.5 0.101 0.105 22.530 22.876 qs_scf_new_mos 108 7.5 0.001 0.001 21.638 21.872 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.638 21.871 ot_scf_mini 108 9.5 0.002 0.003 20.694 20.860 velocity_verlet 10 3.0 0.002 0.006 19.745 19.746 multiply_cannon 2286 13.5 0.209 0.219 16.947 18.376 multiply_cannon_loop 2286 14.5 1.182 1.235 15.699 17.162 ot_mini 108 10.5 0.001 0.001 12.593 12.827 mp_waitall_1 200699 16.5 6.139 11.347 6.139 11.347 qs_ot_get_derivative 108 11.5 0.001 0.001 10.199 10.371 multiply_cannon_metrocomm3 27432 15.5 0.071 0.073 4.004 9.512 multiply_cannon_multrec 27432 15.5 1.837 4.056 6.153 8.974 rebuild_ks_matrix 119 8.3 0.000 0.000 7.424 7.566 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 7.423 7.566 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.539 6.668 dbcsr_mm_accdrv_process 47894 16.0 3.540 5.610 4.235 6.304 qs_ot_get_p 119 10.4 0.001 0.002 5.115 5.331 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.816 4.739 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.259 4.292 calculate_rho_elec 119 8.7 0.021 0.024 4.259 4.292 sum_up_and_integrate 119 10.3 0.001 0.001 4.258 4.264 integrate_v_rspace 119 11.3 0.002 0.002 4.244 4.250 mp_sum_l 7287 12.8 2.161 4.173 2.161 4.173 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.093 4.073 apply_single 119 13.6 0.000 0.000 3.093 4.073 init_scf_run 11 5.9 0.000 0.001 3.982 3.982 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.982 3.982 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.326 3.344 make_m2s 4572 13.5 0.052 0.053 2.992 3.299 make_images 4572 14.5 0.207 0.245 2.902 3.212 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.908 2.909 init_scf_loop 11 6.9 0.000 0.000 2.698 2.698 density_rs2pw 119 9.7 0.003 0.004 2.550 2.620 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.463 2.551 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.501 2.502 cp_fm_redistribute_end 50 14.0 1.266 2.465 1.274 2.469 cp_fm_diag_elpa_base 50 14.0 1.156 2.342 1.188 2.389 ot_diis_step 108 11.5 0.011 0.011 2.344 2.345 calculate_first_density_matrix 1 7.0 0.000 0.000 2.279 2.281 multiply_cannon_sync_h2d 27432 15.5 1.709 2.234 1.709 2.234 calculate_dm_sparse 119 9.5 0.000 0.001 2.085 2.169 potential_pw2rs 119 12.3 0.006 0.006 1.988 2.016 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.906 1.960 acc_transpose_blocks 27432 15.5 0.113 0.120 1.602 1.943 grid_integrate_task_list 119 12.3 1.826 1.910 1.826 1.910 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.904 1.907 make_images_data 4572 15.5 0.047 0.053 1.335 1.780 transfer_rs2pw 487 10.6 0.005 0.006 1.668 1.758 prepare_preconditioner 11 7.9 0.000 0.000 1.729 1.754 make_preconditioner 11 8.9 0.000 0.000 1.729 1.754 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.667 1.709 fft3d_ps 1201 13.6 0.513 0.568 1.627 1.682 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.619 1.674 wfi_extrapolate 11 7.9 0.001 0.001 1.637 1.637 hybrid_alltoall_any 4725 16.4 0.053 0.116 1.179 1.553 mp_alltoall_d11v 2130 13.8 1.373 1.520 1.373 1.520 fft_wrap_pw1pw2_140 487 12.2 0.047 0.050 1.398 1.455 transfer_pw2rs 487 13.2 0.004 0.005 1.420 1.446 jit_kernel_multiply 7 16.4 0.636 1.399 0.636 1.399 grid_collocate_task_list 119 9.7 1.286 1.341 1.286 1.341 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.237 1.278 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.260 1.271 make_images_sizes 4572 15.5 0.005 0.005 0.790 1.101 mp_alltoall_i44 4572 16.5 0.785 1.096 0.785 1.096 mp_allgather_i34 2286 14.5 0.670 1.091 0.670 1.091 mp_sum_d 4139 12.0 0.565 1.051 0.565 1.051 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.015 1.017 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.941 0.955 acc_transpose_blocks_sync 82296 16.5 0.809 0.943 0.809 0.943 mp_alltoall_z22v 1201 15.6 0.838 0.931 0.838 0.931 mp_sendrecv_dv 16779 12.7 0.866 0.899 0.866 0.899 acc_transpose_blocks_kernels 27432 16.5 0.187 0.275 0.655 0.860 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.409000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=458.818182, yerr=1.192262 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 512.102400E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.118 35.705 35.707 qs_mol_dyn_low 1 2.0 0.012 0.054 35.325 35.333 qs_forces 11 3.9 0.002 0.003 35.192 35.199 qs_energies 11 4.9 0.002 0.005 33.567 33.574 scf_env_do_scf 11 5.9 0.000 0.001 28.300 28.301 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.383 25.384 dbcsr_multiply_generic 2286 12.5 0.095 0.097 17.542 17.614 velocity_verlet 10 3.0 0.011 0.040 16.787 16.794 qs_scf_new_mos 108 7.5 0.001 0.001 16.634 16.652 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.634 16.651 ot_scf_mini 108 9.5 0.002 0.003 15.849 15.865 multiply_cannon 2286 13.5 0.193 0.196 13.777 14.441 multiply_cannon_loop 2286 14.5 0.853 0.878 12.896 13.613 ot_mini 108 10.5 0.001 0.001 9.468 9.485 qs_ot_get_derivative 108 11.5 0.001 0.001 7.942 7.958 multiply_cannon_multrec 18288 15.5 1.869 2.843 6.946 7.273 rebuild_ks_matrix 119 8.3 0.000 0.000 6.610 6.628 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.610 6.628 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.844 5.860 dbcsr_mm_accdrv_process 38222 16.0 4.903 5.692 4.984 5.758 mp_waitall_1 158411 16.6 3.287 4.403 3.287 4.403 qs_ot_get_p 119 10.4 0.001 0.001 4.049 4.077 sum_up_and_integrate 119 10.3 0.001 0.001 4.065 4.072 integrate_v_rspace 119 11.3 0.002 0.003 4.053 4.061 init_scf_run 11 5.9 0.000 0.001 3.949 3.949 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.949 3.949 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.924 3.930 calculate_rho_elec 119 8.7 0.031 0.032 3.924 3.929 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.942 3.540 init_scf_loop 11 6.9 0.000 0.002 2.896 2.896 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.743 2.749 density_rs2pw 119 9.7 0.004 0.006 2.472 2.559 calculate_first_density_matrix 1 7.0 0.000 0.002 2.475 2.477 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.417 2.418 make_m2s 4572 13.5 0.044 0.044 2.238 2.400 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.029 2.374 apply_single 119 13.6 0.000 0.000 2.029 2.374 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.414 2.374 make_images 4572 14.5 0.193 0.205 2.153 2.314 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.090 2.099 cp_fm_diag_elpa_base 50 14.0 2.062 2.078 2.088 2.098 prepare_preconditioner 11 7.9 0.000 0.000 2.053 2.056 make_preconditioner 11 8.9 0.000 0.001 2.053 2.056 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.899 1.991 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.913 1.940 grid_integrate_task_list 119 12.3 1.799 1.907 1.799 1.907 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.895 1.903 calculate_dm_sparse 119 9.5 0.000 0.001 1.878 1.888 potential_pw2rs 119 12.3 0.007 0.007 1.835 1.854 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.769 1.771 transfer_rs2pw 487 10.6 0.005 0.005 1.616 1.713 acc_transpose_blocks 18288 15.5 0.078 0.080 1.630 1.687 fft3d_ps 1201 13.6 0.538 0.560 1.598 1.626 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.564 1.574 mp_sum_l 7287 12.8 1.192 1.533 1.192 1.533 ot_diis_step 108 11.5 0.011 0.012 1.507 1.508 fft_wrap_pw1pw2_140 487 12.2 0.061 0.066 1.413 1.448 wfi_extrapolate 11 7.9 0.001 0.001 1.401 1.401 grid_collocate_task_list 119 9.7 1.233 1.350 1.233 1.350 transfer_pw2rs 487 13.2 0.004 0.004 1.279 1.299 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.194 1.198 make_images_data 4572 15.5 0.047 0.051 0.952 1.184 multiply_cannon_sync_h2d 18288 15.5 1.039 1.150 1.039 1.150 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.812 1.093 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.036 1.042 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.999 1.018 mp_alltoall_z22v 1201 15.6 0.885 0.966 0.885 0.966 mp_alltoall_d11v 2130 13.8 0.815 0.936 0.815 0.936 cp_fm_cholesky_invert 11 10.9 0.896 0.900 0.896 0.900 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.838 0.840 acc_transpose_blocks_kernels 18288 16.5 0.216 0.221 0.802 0.822 mp_sendrecv_dv 11067 12.7 0.798 0.815 0.798 0.815 acc_transpose_blocks_sync 54864 16.5 0.731 0.771 0.731 0.771 build_core_hamiltonian_matrix_ 11 4.9 0.002 0.011 0.669 0.744 make_images_sizes 4572 15.5 0.005 0.005 0.530 0.730 mp_alltoall_i44 4572 16.5 0.525 0.725 0.525 0.725 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.489 0.715 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=35.707000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=488.545455, yerr=0.655555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 558.510080E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.055 38.911 38.911 qs_mol_dyn_low 1 2.0 0.003 0.003 38.611 38.631 qs_forces 11 3.9 0.002 0.002 38.552 38.553 qs_energies 11 4.9 0.002 0.002 36.723 36.731 scf_env_do_scf 11 5.9 0.001 0.001 31.502 31.503 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 27.773 27.773 dbcsr_multiply_generic 2286 12.5 0.102 0.105 20.509 20.633 velocity_verlet 10 3.0 0.001 0.002 19.642 19.644 qs_scf_new_mos 108 7.5 0.001 0.001 18.985 19.044 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.984 19.044 ot_scf_mini 108 9.5 0.002 0.003 17.927 17.971 multiply_cannon 2286 13.5 0.218 0.227 16.268 16.702 multiply_cannon_loop 2286 14.5 1.490 1.616 15.249 15.622 ot_mini 108 10.5 0.001 0.001 10.833 10.891 multiply_cannon_multrec 27432 15.5 2.452 3.114 9.096 9.376 qs_ot_get_derivative 108 11.5 0.001 0.001 8.934 8.981 dbcsr_mm_accdrv_process 47916 15.9 5.830 7.444 6.542 7.889 rebuild_ks_matrix 119 8.3 0.000 0.000 6.828 6.891 qs_ks_build_kohn_sham_matrix 119 9.3 0.022 0.034 6.828 6.890 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.998 6.055 qs_ot_get_p 119 10.4 0.001 0.001 4.186 4.253 sum_up_and_integrate 119 10.3 0.001 0.001 3.898 3.905 integrate_v_rspace 119 11.3 0.002 0.003 3.886 3.894 init_scf_run 11 5.9 0.000 0.001 3.820 3.821 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.820 3.820 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.752 3.786 calculate_rho_elec 119 8.7 0.040 0.046 3.752 3.785 init_scf_loop 11 6.9 0.000 0.000 3.702 3.702 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.266 3.654 mp_waitall_1 137007 16.6 2.386 3.020 2.386 3.020 prepare_preconditioner 11 7.9 0.000 0.000 2.815 2.830 make_preconditioner 11 8.9 0.000 0.000 2.815 2.830 make_m2s 4572 13.5 0.054 0.056 2.670 2.786 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.412 2.752 make_images 4572 14.5 0.275 0.335 2.561 2.676 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.561 2.575 acc_transpose_blocks 27432 15.5 0.115 0.120 2.363 2.556 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.489 2.515 calculate_first_density_matrix 1 7.0 0.000 0.000 2.305 2.307 calculate_dm_sparse 119 9.5 0.000 0.000 2.225 2.284 density_rs2pw 119 9.7 0.003 0.003 2.175 2.265 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.135 2.262 apply_single 119 13.6 0.000 0.000 2.135 2.262 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.146 2.146 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.893 1.931 grid_integrate_task_list 119 12.3 1.831 1.912 1.831 1.912 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.849 1.858 cp_fm_diag_elpa_base 50 14.0 1.808 1.826 1.846 1.856 ot_diis_step 108 11.5 0.012 0.012 1.855 1.855 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.852 1.853 potential_pw2rs 119 12.3 0.008 0.009 1.663 1.675 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.659 1.671 fft3d_ps 1201 13.6 0.579 0.641 1.556 1.593 acc_transpose_blocks_sync 82296 16.5 1.384 1.571 1.384 1.571 jit_kernel_multiply 8 15.9 0.646 1.551 0.646 1.551 fft_wrap_pw1pw2_140 487 12.2 0.071 0.074 1.497 1.534 wfi_extrapolate 11 7.9 0.001 0.001 1.456 1.456 mp_sum_l 7287 12.8 1.159 1.455 1.159 1.455 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.793 1.452 transfer_rs2pw 487 10.6 0.005 0.005 1.282 1.381 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.320 1.339 grid_collocate_task_list 119 9.7 1.250 1.331 1.250 1.331 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.305 1.316 cp_fm_upper_to_full 72 14.2 0.834 1.189 0.834 1.189 make_images_data 4572 15.5 0.047 0.051 1.030 1.184 transfer_pw2rs 487 13.2 0.004 0.004 1.119 1.133 dbcsr_complete_redistribute 329 12.2 0.119 0.150 0.841 1.128 qs_energies_init_hamiltonians 11 5.9 0.009 0.019 1.098 1.099 hybrid_alltoall_any 4725 16.4 0.065 0.155 0.850 1.038 mp_alltoall_d11v 2130 13.8 0.868 0.985 0.868 0.985 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.919 0.923 cp_fm_cholesky_invert 11 10.9 0.917 0.920 0.917 0.920 mp_alltoall_z22v 1201 15.6 0.860 0.892 0.860 0.892 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.801 0.879 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.594 0.878 acc_transpose_blocks_kernels 27432 16.5 0.270 0.277 0.836 0.854 multiply_cannon_metrocomm1 27432 15.5 0.034 0.036 0.227 0.780 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=38.911000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.272727, yerr=4.201535 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 629.112832E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.034 29.025 29.025 qs_mol_dyn_low 1 2.0 0.003 0.003 28.812 28.858 qs_forces 11 3.9 0.002 0.002 28.751 28.755 qs_energies 11 4.9 0.024 0.025 27.039 27.044 scf_env_do_scf 11 5.9 0.000 0.001 22.335 22.336 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.877 19.878 velocity_verlet 10 3.0 0.002 0.002 14.714 14.717 dbcsr_multiply_generic 2286 12.5 0.094 0.097 13.258 13.379 qs_scf_new_mos 108 7.5 0.001 0.001 12.266 12.290 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.266 12.289 ot_scf_mini 108 9.5 0.002 0.002 11.544 11.566 multiply_cannon 2286 13.5 0.223 0.226 10.546 11.095 multiply_cannon_loop 2286 14.5 0.643 0.665 9.617 9.837 ot_mini 108 10.5 0.001 0.001 6.586 6.613 multiply_cannon_multrec 9144 15.5 1.700 1.951 5.976 6.167 rebuild_ks_matrix 119 8.3 0.000 0.000 5.777 5.800 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.776 5.799 qs_ot_get_derivative 108 11.5 0.001 0.001 5.293 5.315 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.137 5.157 dbcsr_mm_accdrv_process 12550 15.8 3.685 4.200 4.166 4.243 sum_up_and_integrate 119 10.3 0.001 0.001 3.409 3.418 integrate_v_rspace 119 11.3 0.003 0.003 3.399 3.408 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.295 3.301 calculate_rho_elec 119 8.7 0.060 0.061 3.295 3.301 init_scf_run 11 5.9 0.000 0.001 3.225 3.225 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.225 3.225 qs_ot_get_p 119 10.4 0.001 0.001 3.050 3.082 init_scf_loop 11 6.9 0.000 0.000 2.434 2.437 calculate_first_density_matrix 1 7.0 0.000 0.000 2.083 2.084 mp_waitall_1 115863 16.7 1.518 2.066 1.518 2.066 make_m2s 4572 13.5 0.033 0.034 1.870 2.019 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.001 2.004 make_images 4572 14.5 0.270 0.299 1.781 1.928 grid_integrate_task_list 119 12.3 1.868 1.903 1.868 1.903 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.784 1.790 calculate_dm_sparse 119 9.5 0.000 0.000 1.753 1.770 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.744 1.762 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.749 1.749 density_rs2pw 119 9.7 0.003 0.003 1.667 1.715 prepare_preconditioner 11 7.9 0.000 0.000 1.711 1.715 make_preconditioner 11 8.9 0.000 0.000 1.711 1.715 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.603 1.629 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.536 1.538 fft_wrap_pw1pw2_140 487 12.2 0.093 0.095 1.461 1.469 acc_transpose_blocks 9144 15.5 0.041 0.042 1.427 1.456 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.440 1.448 cp_fm_diag_elpa_base 50 14.0 1.411 1.430 1.438 1.446 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.397 1.408 fft3d_ps 1201 13.6 0.638 0.653 1.395 1.405 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.371 1.380 grid_collocate_task_list 119 9.7 1.298 1.369 1.298 1.369 ot_diis_step 108 11.5 0.013 0.013 1.279 1.279 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.217 1.255 apply_single 119 13.6 0.000 0.000 1.216 1.255 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.231 1.235 potential_pw2rs 119 12.3 0.010 0.010 1.202 1.206 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.131 1.137 wfi_extrapolate 11 7.9 0.001 0.001 1.093 1.093 jit_kernel_multiply 6 15.4 0.441 1.017 0.441 1.017 hybrid_alltoall_any 4725 16.4 0.066 0.180 0.779 0.990 make_images_data 4572 15.5 0.042 0.046 0.799 0.956 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.877 0.927 mp_alltoall_d11v 2130 13.8 0.819 0.867 0.819 0.867 transfer_rs2pw 487 10.6 0.004 0.005 0.824 0.867 cp_fm_cholesky_invert 11 10.9 0.854 0.857 0.854 0.857 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.840 0.848 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.762 0.764 acc_transpose_blocks_sync 27432 16.5 0.716 0.742 0.716 0.742 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.677 0.732 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.399 0.724 transfer_pw2rs 487 13.2 0.003 0.004 0.694 0.696 mp_alltoall_z22v 1201 15.6 0.632 0.669 0.632 0.669 mp_allgather_i34 2286 14.5 0.243 0.664 0.243 0.664 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.654 0.657 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.314 0.623 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.025000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=593.818182, yerr=8.144359 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 780.161024E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.034 46.216 46.216 qs_mol_dyn_low 1 2.0 0.003 0.003 45.993 46.000 qs_forces 11 3.9 0.002 0.002 45.935 45.936 qs_energies 11 4.9 0.001 0.002 43.875 43.881 scf_env_do_scf 11 5.9 0.000 0.001 38.046 38.046 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 29.376 29.378 velocity_verlet 10 3.0 0.002 0.002 26.133 26.139 dbcsr_multiply_generic 2286 12.5 0.104 0.105 20.246 20.588 qs_scf_new_mos 108 7.5 0.001 0.001 19.247 19.344 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.247 19.343 ot_scf_mini 108 9.5 0.002 0.002 18.009 18.106 multiply_cannon 2286 13.5 0.300 0.309 15.350 16.490 multiply_cannon_loop 2286 14.5 0.860 0.878 13.984 15.110 ot_mini 108 10.5 0.001 0.001 10.710 10.833 multiply_cannon_multrec 9144 15.5 3.442 4.793 8.781 8.840 init_scf_loop 11 6.9 0.000 0.000 8.634 8.637 qs_ot_get_derivative 108 11.5 0.001 0.001 8.530 8.621 rebuild_ks_matrix 119 8.3 0.000 0.000 7.589 7.746 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 7.588 7.746 prepare_preconditioner 11 7.9 0.000 0.000 7.603 7.617 make_preconditioner 11 8.9 0.000 0.000 7.603 7.617 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.134 7.478 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.833 6.976 dbcsr_mm_accdrv_process 12550 15.8 4.433 5.809 5.202 6.530 cp_fm_upper_to_full 72 14.2 3.253 4.653 3.253 4.653 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.399 4.419 calculate_rho_elec 119 8.7 0.118 0.122 4.399 4.418 mp_waitall_1 94719 16.7 3.043 4.387 3.043 4.387 qs_ot_get_p 119 10.4 0.001 0.001 4.145 4.288 sum_up_and_integrate 119 10.3 0.001 0.001 4.127 4.134 integrate_v_rspace 119 11.3 0.003 0.003 4.116 4.124 init_scf_run 11 5.9 0.000 0.001 3.738 3.738 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.738 3.738 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.050 3.477 make_m2s 4572 13.5 0.037 0.037 2.954 3.168 make_images 4572 14.5 0.352 0.387 2.834 3.049 dbcsr_complete_redistribute 329 12.2 0.282 0.284 2.129 3.007 fft_wrap_pw1pw2 1201 11.6 0.014 0.014 2.732 2.740 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.555 2.653 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.746 2.603 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.293 2.564 apply_single 119 13.6 0.000 0.000 2.292 2.563 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.508 2.510 density_rs2pw 119 9.7 0.003 0.003 2.344 2.391 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.314 2.363 calculate_dm_sparse 119 9.5 0.000 0.000 2.332 2.350 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.464 2.320 mp_alltoall_i22 627 13.8 1.458 2.298 1.458 2.298 fft_wrap_pw1pw2_140 487 12.2 0.179 0.180 2.237 2.245 fft3d_ps 1201 13.6 0.860 0.882 2.183 2.192 mp_sum_l 7287 12.8 1.404 2.175 1.404 2.175 grid_integrate_task_list 119 12.3 2.134 2.171 2.134 2.171 ot_diis_step 108 11.5 0.014 0.014 2.122 2.123 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.117 2.118 calculate_first_density_matrix 1 7.0 0.000 0.000 2.094 2.099 cp_fm_cholesky_invert 11 10.9 1.964 1.970 1.964 1.970 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.871 1.872 acc_transpose_blocks 9144 15.5 0.044 0.045 1.803 1.835 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.779 1.780 hybrid_alltoall_any 4725 16.4 0.090 0.149 1.340 1.755 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.752 1.752 cp_fm_diag_elpa_base 50 14.0 1.601 1.659 1.749 1.749 make_images_data 4572 15.5 0.045 0.048 1.341 1.719 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.576 1.615 grid_collocate_task_list 119 9.7 1.559 1.597 1.559 1.597 wfi_extrapolate 11 7.9 0.001 0.001 1.576 1.576 potential_pw2rs 119 12.3 0.013 0.014 1.536 1.540 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.445 1.480 mp_alltoall_d11v 2130 13.8 1.369 1.450 1.369 1.450 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.264 1.276 mp_alltoall_z22v 1201 15.6 1.188 1.218 1.188 1.218 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.108 1.126 acc_transpose_blocks_sync 27432 16.5 1.093 1.123 1.093 1.123 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.991 1.031 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.969 0.984 transfer_rs2pw 487 10.6 0.005 0.005 0.930 0.969 qs_create_task_list 11 7.9 0.000 0.000 0.941 0.952 generate_qs_task_list 11 8.9 0.368 0.388 0.941 0.952 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=46.216000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=730.545455, yerr=14.411887 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 496.455680E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 2449583. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.032 87.289 87.292 qs_mol_dyn_low 1 2.0 0.008 0.029 86.930 86.938 qs_forces 11 3.9 0.004 0.009 86.747 86.750 qs_energies 11 4.9 0.004 0.016 83.771 83.781 scf_env_do_scf 11 5.9 0.001 0.001 74.428 74.430 scf_env_do_scf_inner_loop 99 6.5 0.003 0.012 68.188 68.189 dbcsr_multiply_generic 2055 12.4 0.105 0.107 52.313 52.499 qs_scf_new_mos 99 7.5 0.000 0.001 50.024 50.126 qs_scf_loop_do_ot 99 8.5 0.001 0.001 50.023 50.125 ot_scf_mini 99 9.5 0.002 0.002 47.585 47.679 velocity_verlet 10 3.0 0.025 0.068 46.072 46.076 multiply_cannon 2055 13.4 0.184 0.188 43.014 43.735 multiply_cannon_loop 2055 14.4 1.773 1.807 41.896 42.626 ot_mini 99 10.5 0.001 0.001 27.731 27.835 qs_ot_get_derivative 99 11.5 0.001 0.001 20.894 21.011 multiply_cannon_multrec 49320 15.4 11.320 12.140 17.356 18.056 rebuild_ks_matrix 110 8.3 0.000 0.000 15.252 15.367 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.015 15.251 15.367 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.406 13.514 mp_waitall_1 220248 16.4 11.825 12.810 11.825 12.810 qs_ot_get_p 110 10.4 0.001 0.002 11.248 11.386 multiply_cannon_sync_h2d 49320 15.4 9.529 10.056 9.529 10.056 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.286 8.852 multiply_cannon_metrocomm3 49320 15.4 0.083 0.087 6.723 7.837 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.241 7.748 apply_single 110 13.6 0.000 0.000 7.240 7.748 qs_ot_p2m_diag 48 11.0 0.013 0.019 7.623 7.641 sum_up_and_integrate 110 10.3 0.002 0.005 7.539 7.550 integrate_v_rspace 110 11.3 0.003 0.003 7.513 7.532 init_scf_run 11 5.9 0.000 0.001 7.160 7.160 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.159 7.160 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.936 7.059 calculate_rho_elec 110 8.6 0.020 0.025 6.935 7.059 cp_dbcsr_syevd 48 12.0 0.002 0.003 6.707 6.708 ot_diis_step 99 11.5 0.005 0.006 6.622 6.622 init_scf_loop 11 6.9 0.000 0.000 6.200 6.201 dbcsr_mm_accdrv_process 87628 16.1 3.053 3.137 5.906 6.192 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.987 6.013 cp_fm_diag_elpa_base 48 14.0 5.968 5.993 5.984 6.011 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.613 5.666 mp_sum_l 6594 12.7 4.377 5.077 4.377 5.077 make_m2s 4110 13.4 0.061 0.065 4.478 4.620 make_images 4110 14.4 0.178 0.190 4.380 4.526 wfi_extrapolate 11 7.9 0.003 0.016 4.292 4.293 density_rs2pw 110 9.6 0.003 0.004 3.953 4.130 prepare_preconditioner 11 7.9 0.000 0.000 3.917 3.939 make_preconditioner 11 8.9 0.000 0.000 3.917 3.938 calculate_dm_sparse 110 9.5 0.001 0.001 3.834 3.931 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.681 3.729 multiply_cannon_metrocomm1 49320 15.4 0.063 0.066 2.650 3.633 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 3.613 3.617 grid_integrate_task_list 110 12.3 3.242 3.424 3.242 3.424 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.303 3.376 fft_wrap_pw1pw2 1111 11.6 0.014 0.020 3.243 3.359 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.211 3.270 potential_pw2rs 110 12.3 0.006 0.006 3.083 3.122 mp_alltoall_d11v 2046 13.8 2.301 2.945 2.301 2.945 fft3d_ps 1111 13.6 0.731 0.822 2.786 2.896 calculate_first_density_matrix 1 7.0 0.000 0.001 2.760 2.765 fft_wrap_pw1pw2_140 451 12.1 0.092 0.098 2.600 2.718 acc_transpose_blocks 49320 15.4 0.204 0.211 2.555 2.616 jit_kernel_multiply 13 15.9 2.573 2.607 2.573 2.607 transfer_rs2pw 451 10.6 0.006 0.006 2.372 2.536 grid_collocate_task_list 110 9.6 2.156 2.265 2.156 2.265 transfer_pw2rs 451 13.1 0.006 0.007 2.192 2.233 mp_waitany 14300 13.8 1.852 2.135 1.852 2.135 make_images_data 4110 15.4 0.043 0.046 1.937 2.120 cp_fm_cholesky_invert 11 10.9 2.107 2.113 2.107 2.113 mp_sum_d 3893 11.9 1.446 2.048 1.446 2.048 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.988 2.004 hybrid_alltoall_any 4261 16.3 0.084 0.483 1.684 1.956 mp_alltoall_z22v 1111 15.6 1.690 1.885 1.690 1.885 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=87.292000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=471.181818, yerr=2.979267 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 589.594624E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2731430. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.047 72.971 72.973 qs_mol_dyn_low 1 2.0 0.010 0.033 72.663 72.691 qs_forces 11 3.9 0.003 0.003 72.508 72.514 qs_energies 11 4.9 0.002 0.002 69.113 69.124 scf_env_do_scf 11 5.9 0.001 0.001 60.368 60.371 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 51.842 51.843 dbcsr_multiply_generic 2055 12.4 0.116 0.120 39.106 39.463 velocity_verlet 10 3.0 0.020 0.065 38.138 38.144 qs_scf_new_mos 99 7.5 0.001 0.001 35.326 35.451 qs_scf_loop_do_ot 99 8.5 0.001 0.001 35.325 35.450 ot_scf_mini 99 9.5 0.003 0.003 33.614 33.727 multiply_cannon 2055 13.4 0.225 0.244 31.754 33.205 multiply_cannon_loop 2055 14.4 1.158 1.182 30.249 31.454 ot_mini 99 10.5 0.001 0.001 19.243 19.371 multiply_cannon_multrec 24660 15.4 6.954 8.389 14.026 15.293 rebuild_ks_matrix 110 8.3 0.000 0.000 14.035 14.111 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 14.034 14.111 qs_ot_get_derivative 99 11.5 0.001 0.001 13.410 13.528 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.361 12.433 mp_waitall_1 176588 16.5 8.154 11.434 8.154 11.434 multiply_cannon_metrocomm3 24660 15.4 0.071 0.074 5.303 8.651 init_scf_loop 11 6.9 0.000 0.000 8.482 8.482 qs_ot_get_p 110 10.4 0.001 0.001 7.495 7.633 multiply_cannon_sync_h2d 24660 15.4 6.394 7.574 6.394 7.574 dbcsr_mm_accdrv_process 52282 16.1 5.574 6.425 6.903 7.228 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.470 7.022 apply_single 110 13.6 0.000 0.001 6.470 7.022 sum_up_and_integrate 110 10.3 0.001 0.002 6.642 6.652 integrate_v_rspace 110 11.3 0.002 0.003 6.615 6.627 prepare_preconditioner 11 7.9 0.000 0.000 6.405 6.425 make_preconditioner 11 8.9 0.000 0.000 6.405 6.425 init_scf_run 11 5.9 0.000 0.001 6.310 6.310 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.309 6.310 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.167 6.182 calculate_rho_elec 110 8.6 0.039 0.047 6.166 6.181 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.967 6.128 ot_diis_step 99 11.5 0.010 0.010 5.752 5.752 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.967 5.665 qs_ot_p2m_diag 48 11.0 0.029 0.045 5.346 5.373 make_m2s 4110 13.4 0.056 0.059 4.602 5.110 make_images 4110 14.4 0.408 0.465 4.490 4.993 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.838 4.838 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.162 4.178 cp_fm_diag_elpa_base 48 14.0 4.109 4.126 4.159 4.174 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.608 3.670 wfi_extrapolate 11 7.9 0.001 0.001 3.662 3.662 density_rs2pw 110 9.6 0.004 0.004 3.401 3.606 fft_wrap_pw1pw2 1111 11.6 0.015 0.016 3.341 3.495 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.346 3.348 grid_integrate_task_list 110 12.3 3.171 3.346 3.171 3.346 cp_fm_cholesky_invert 11 10.9 3.171 3.181 3.171 3.181 calculate_dm_sparse 110 9.5 0.001 0.001 3.021 3.049 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.981 3.010 make_images_data 4110 15.4 0.048 0.053 2.467 3.006 hybrid_alltoall_any 4261 16.3 0.105 0.451 2.121 2.983 fft3d_ps 1111 13.6 1.070 1.276 2.722 2.870 fft_wrap_pw1pw2_140 451 12.1 0.110 0.118 2.670 2.821 mp_sum_l 6594 12.7 2.133 2.765 2.133 2.765 calculate_first_density_matrix 1 7.0 0.000 0.000 2.545 2.547 potential_pw2rs 110 12.3 0.008 0.009 2.511 2.544 grid_collocate_task_list 110 9.6 2.168 2.299 2.168 2.299 mp_alltoall_d11v 2046 13.8 1.840 2.137 1.840 2.137 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.088 2.112 acc_transpose_blocks 24660 15.4 0.114 0.117 1.993 2.039 transfer_rs2pw 451 10.6 0.006 0.008 1.775 1.978 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 1.895 1.900 mp_allgather_i34 2055 14.4 0.810 1.893 0.810 1.893 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.836 1.843 jit_kernel_multiply 9 16.3 0.958 1.770 0.958 1.770 multiply_cannon_metrocomm4 22605 15.4 0.078 0.082 0.771 1.720 cp_fm_cholesky_decompose 22 10.9 1.587 1.593 1.587 1.593 transfer_pw2rs 451 13.1 0.005 0.007 1.545 1.569 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.549 1.557 dbcsr_complete_redistribute 325 12.2 0.240 0.301 1.254 1.526 build_core_hamiltonian_matrix_ 11 4.9 0.002 0.007 1.389 1.497 mp_irecv_dv 57340 16.2 0.641 1.497 0.641 1.497 mp_alltoall_z22v 1111 15.6 1.370 1.475 1.370 1.475 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=72.973000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.363636, yerr=7.201928 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 663.367680E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.054 59.849 59.851 qs_mol_dyn_low 1 2.0 0.005 0.019 59.501 59.512 qs_forces 11 3.9 0.003 0.003 59.370 59.371 qs_energies 11 4.9 0.006 0.035 56.215 56.221 scf_env_do_scf 11 5.9 0.001 0.007 47.686 47.687 scf_env_do_scf_inner_loop 99 6.5 0.008 0.023 39.168 39.169 velocity_verlet 10 3.0 0.008 0.053 32.161 32.165 dbcsr_multiply_generic 2055 12.4 0.108 0.111 28.561 28.820 qs_scf_new_mos 99 7.5 0.001 0.001 24.866 24.972 qs_scf_loop_do_ot 99 8.5 0.003 0.016 24.866 24.971 ot_scf_mini 99 9.5 0.003 0.009 23.620 23.751 multiply_cannon 2055 13.4 0.213 0.220 22.428 23.473 multiply_cannon_loop 2055 14.4 0.812 0.845 21.248 22.350 ot_mini 99 10.5 0.001 0.001 13.575 13.705 rebuild_ks_matrix 110 8.3 0.000 0.000 12.035 12.159 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.034 12.158 multiply_cannon_multrec 16440 15.4 3.664 4.849 9.864 10.942 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.596 10.711 mp_waitall_1 139946 16.5 6.683 9.946 6.683 9.946 qs_ot_get_derivative 99 11.5 0.001 0.001 9.171 9.304 init_scf_loop 11 6.9 0.001 0.003 8.484 8.485 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.233 7.268 prepare_preconditioner 11 7.9 0.000 0.000 6.689 6.708 make_preconditioner 11 8.9 0.000 0.001 6.689 6.708 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.945 6.305 dbcsr_mm_accdrv_process 34862 16.1 5.339 5.720 6.044 6.195 sum_up_and_integrate 110 10.3 0.001 0.002 6.098 6.113 init_scf_run 11 5.9 0.000 0.001 6.098 6.098 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.098 6.098 integrate_v_rspace 110 11.3 0.003 0.003 6.071 6.088 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.443 5.451 calculate_rho_elec 110 8.6 0.058 0.058 5.443 5.450 qs_ot_get_p 110 10.4 0.022 0.040 5.273 5.413 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.908 5.347 apply_single 110 13.6 0.000 0.000 4.908 5.347 make_m2s 4110 13.4 0.049 0.050 4.045 4.446 ot_diis_step 99 11.5 0.011 0.011 4.375 4.375 make_images 4110 14.4 0.402 0.529 3.929 4.329 multiply_cannon_sync_h2d 16440 15.4 3.264 3.796 3.264 3.796 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.143 3.794 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.610 3.614 grid_integrate_task_list 110 12.3 3.202 3.417 3.202 3.417 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.246 3.247 calculate_first_density_matrix 1 7.0 0.000 0.003 3.103 3.105 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.025 3.036 wfi_extrapolate 11 7.9 0.001 0.001 2.907 2.907 density_rs2pw 110 9.6 0.004 0.004 2.689 2.845 hybrid_alltoall_any 4261 16.3 0.109 0.379 2.052 2.763 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.754 2.756 make_images_data 4110 15.4 0.045 0.050 2.256 2.745 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.722 2.733 cp_fm_diag_elpa_base 48 14.0 2.657 2.691 2.721 2.732 fft_wrap_pw1pw2_140 451 12.1 0.125 0.128 2.609 2.626 calculate_dm_sparse 110 9.5 0.001 0.001 2.528 2.563 cp_fm_cholesky_invert 11 10.9 2.545 2.551 2.545 2.551 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.442 2.485 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.395 2.459 multiply_cannon_metrocomm4 14385 15.4 0.048 0.053 0.883 2.387 grid_collocate_task_list 110 9.6 2.212 2.380 2.212 2.380 fft3d_ps 1111 13.6 1.094 1.112 2.339 2.350 mp_irecv_dv 48980 15.7 0.806 2.250 0.806 2.250 mp_alltoall_d11v 2046 13.8 1.843 2.181 1.843 2.181 mp_sum_l 6594 12.7 1.478 2.033 1.478 2.033 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.963 1.964 potential_pw2rs 110 12.3 0.010 0.010 1.943 1.948 dbcsr_complete_redistribute 325 12.2 0.337 0.380 1.480 1.935 cp_fm_upper_to_full 70 14.2 1.393 1.758 1.393 1.758 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.635 1.651 acc_transpose_blocks 16440 15.4 0.076 0.079 1.578 1.643 cp_fm_cholesky_decompose 22 10.9 1.534 1.557 1.534 1.557 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.039 1.484 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.467 1.480 mp_allgather_i34 2055 14.4 0.483 1.467 0.483 1.467 parallel_gemm_fm 81 9.0 0.000 0.000 1.461 1.463 parallel_gemm_fm_cosma 81 10.0 1.461 1.463 1.461 1.463 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.008 1.361 1.463 make_basis_sm 11 9.8 0.000 0.000 1.451 1.454 transfer_rs2pw 451 10.6 0.005 0.006 1.248 1.416 mp_waitany 17072 13.8 1.124 1.332 1.124 1.332 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.268 1.284 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.851000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=626.909091, yerr=8.490150 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 738.025472E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.071 0.101 66.825 66.828 qs_mol_dyn_low 1 2.0 0.004 0.009 66.241 66.311 qs_forces 11 3.9 0.003 0.003 66.172 66.174 qs_energies 11 4.9 0.026 0.038 62.729 62.733 scf_env_do_scf 11 5.9 0.001 0.002 53.044 53.047 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.435 41.436 velocity_verlet 10 3.0 0.002 0.006 36.845 36.847 dbcsr_multiply_generic 2055 12.4 0.115 0.119 30.492 30.705 qs_scf_new_mos 99 7.5 0.001 0.001 27.166 27.277 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.165 27.277 ot_scf_mini 99 9.5 0.003 0.003 25.482 25.589 multiply_cannon 2055 13.4 0.238 0.262 23.628 24.990 multiply_cannon_loop 2055 14.4 1.398 1.483 22.113 22.663 ot_mini 99 10.5 0.001 0.001 14.635 14.759 multiply_cannon_multrec 24660 15.4 4.103 6.701 13.053 14.216 rebuild_ks_matrix 110 8.3 0.000 0.000 11.959 12.084 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.023 11.959 12.084 init_scf_loop 11 6.9 0.001 0.003 11.571 11.571 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.513 10.621 qs_ot_get_derivative 99 11.5 0.001 0.001 10.475 10.579 prepare_preconditioner 11 7.9 0.000 0.000 9.772 9.787 make_preconditioner 11 8.9 0.000 0.001 9.772 9.787 dbcsr_mm_accdrv_process 52304 16.0 7.810 9.094 8.792 9.734 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.992 9.468 init_scf_run 11 5.9 0.000 0.001 6.383 6.383 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.382 6.383 sum_up_and_integrate 110 10.3 0.001 0.002 6.116 6.135 integrate_v_rspace 110 11.3 0.003 0.004 6.090 6.108 qs_ot_get_p 110 10.4 0.022 0.024 5.786 5.933 mp_waitall_1 121746 16.5 3.968 5.911 3.968 5.911 make_m2s 4110 13.4 0.059 0.061 5.257 5.728 make_images 4110 14.4 0.578 0.700 5.115 5.583 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.545 5.556 calculate_rho_elec 110 8.6 0.077 0.081 5.545 5.556 cp_fm_upper_to_full 70 14.2 3.368 4.886 3.368 4.886 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.028 4.151 apply_single 110 13.6 0.000 0.000 4.027 4.150 ot_diis_step 99 11.5 0.011 0.011 4.117 4.117 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.902 3.919 dbcsr_complete_redistribute 325 12.2 0.418 0.460 2.675 3.820 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.456 3.505 grid_integrate_task_list 110 12.3 3.289 3.440 3.289 3.440 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.318 3.375 cp_dbcsr_syevd 48 12.0 0.003 0.005 3.340 3.340 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.164 3.305 calculate_first_density_matrix 1 7.0 0.001 0.007 3.300 3.303 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.208 3.239 hybrid_alltoall_any 4261 16.3 0.122 0.467 2.211 3.072 calculate_dm_sparse 110 9.5 0.001 0.001 3.005 3.046 make_images_data 4110 15.4 0.048 0.052 2.546 3.015 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.318 2.997 wfi_extrapolate 11 7.9 0.001 0.001 2.988 2.988 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.771 2.899 mp_alltoall_i22 605 13.7 1.697 2.882 1.697 2.882 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.859 2.865 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.815 2.824 density_rs2pw 110 9.6 0.004 0.005 2.674 2.822 cp_fm_diag_elpa_base 48 14.0 2.668 2.726 2.812 2.822 fft_wrap_pw1pw2_140 451 12.1 0.145 0.150 2.762 2.799 cp_fm_cholesky_invert 11 10.9 2.595 2.603 2.595 2.603 qs_energies_init_hamiltonians 11 5.9 0.174 0.235 2.577 2.579 acc_transpose_blocks 24660 15.4 0.107 0.112 2.419 2.563 multiply_cannon_sync_h2d 24660 15.4 2.384 2.551 2.384 2.551 fft3d_ps 1111 13.6 1.138 1.183 2.476 2.504 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.457 2.499 grid_collocate_task_list 110 9.6 2.264 2.426 2.264 2.426 mp_alltoall_d11v 2046 13.8 1.746 2.077 1.746 2.077 potential_pw2rs 110 12.3 0.012 0.013 1.922 1.932 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.775 1.810 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.613 1.714 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.688 1.702 cp_fm_cholesky_decompose 22 10.9 1.592 1.643 1.592 1.643 jit_kernel_multiply 9 15.8 0.635 1.598 0.635 1.598 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.562 1.579 multiply_cannon_metrocomm4 20550 15.4 0.059 0.063 0.852 1.547 mp_allgather_i34 2055 14.4 0.595 1.537 0.595 1.537 acc_transpose_blocks_sync 73980 16.4 1.374 1.519 1.374 1.519 mp_sum_l 6594 12.7 0.939 1.494 0.939 1.494 mp_irecv_dv 62702 16.1 0.748 1.465 0.748 1.465 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.293 1.427 make_basis_sm 11 9.8 0.000 0.001 1.402 1.408 transfer_rs2pw 451 10.6 0.005 0.006 1.198 1.378 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.828000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=697.090909, yerr=10.166384 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 859.951104E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.054 54.895 54.896 qs_mol_dyn_low 1 2.0 0.003 0.003 54.577 54.604 qs_forces 11 3.9 0.003 0.003 54.513 54.514 qs_energies 11 4.9 0.002 0.002 50.900 50.904 scf_env_do_scf 11 5.9 0.000 0.001 42.056 42.056 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.381 34.381 velocity_verlet 10 3.0 0.002 0.002 30.601 30.604 dbcsr_multiply_generic 2055 12.4 0.106 0.108 23.187 23.321 qs_scf_new_mos 99 7.5 0.001 0.001 20.298 20.347 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.298 20.346 multiply_cannon 2055 13.4 0.239 0.251 17.720 19.124 ot_scf_mini 99 9.5 0.002 0.002 19.044 19.074 multiply_cannon_loop 2055 14.4 0.605 0.624 16.444 16.779 rebuild_ks_matrix 110 8.3 0.000 0.000 11.460 11.497 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.459 11.496 ot_mini 99 10.5 0.001 0.001 10.606 10.627 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.196 10.226 multiply_cannon_multrec 8220 15.4 3.184 4.392 7.851 8.768 init_scf_loop 11 6.9 0.000 0.000 7.629 7.630 mp_waitall_1 103326 16.6 5.562 7.220 5.562 7.220 qs_ot_get_derivative 99 11.5 0.001 0.001 6.915 6.945 sum_up_and_integrate 110 10.3 0.001 0.002 6.002 6.015 prepare_preconditioner 11 7.9 0.000 0.000 5.988 5.993 make_preconditioner 11 8.9 0.000 0.000 5.988 5.993 integrate_v_rspace 110 11.3 0.003 0.003 5.975 5.988 init_scf_run 11 5.9 0.000 0.001 5.717 5.717 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.717 5.717 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.569 5.644 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.619 5.630 calculate_rho_elec 110 8.6 0.115 0.116 5.619 5.630 dbcsr_mm_accdrv_process 17442 15.9 3.290 4.487 4.526 5.543 qs_ot_get_p 110 10.4 0.001 0.001 4.507 4.521 make_m2s 4110 13.4 0.037 0.039 4.158 4.453 make_images 4110 14.4 0.647 0.707 4.029 4.322 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 2.869 4.283 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.729 3.790 apply_single 110 13.6 0.000 0.000 3.729 3.789 ot_diis_step 99 11.5 0.012 0.012 3.665 3.665 grid_integrate_task_list 110 12.3 3.385 3.516 3.385 3.516 fft_wrap_pw1pw2 1111 11.6 0.013 0.015 3.293 3.297 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.174 3.178 calculate_first_density_matrix 1 7.0 0.000 0.000 2.966 2.967 fft_wrap_pw1pw2_140 451 12.1 0.191 0.193 2.849 2.858 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.839 2.839 cp_fm_cholesky_invert 11 10.9 2.827 2.831 2.827 2.831 density_rs2pw 110 9.6 0.003 0.004 2.640 2.752 hybrid_alltoall_any 4261 16.3 0.201 0.853 2.189 2.710 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.674 2.675 wfi_extrapolate 11 7.9 0.001 0.001 2.659 2.659 calculate_dm_sparse 110 9.5 0.001 0.001 2.615 2.652 make_images_data 4110 15.4 0.041 0.047 2.231 2.617 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.475 2.475 multiply_cannon_sync_h2d 8220 15.4 2.386 2.467 2.386 2.467 grid_collocate_task_list 110 9.6 2.366 2.467 2.366 2.467 fft3d_ps 1111 13.6 1.280 1.290 2.453 2.463 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.335 2.342 cp_fm_diag_elpa_base 48 14.0 2.279 2.299 2.333 2.341 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.164 2.202 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.121 2.137 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.783 1.994 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.891 1.904 potential_pw2rs 110 12.3 0.015 0.015 1.811 1.813 mp_alltoall_d11v 2046 13.8 1.569 1.704 1.569 1.704 cp_fm_cholesky_decompose 22 10.9 1.663 1.680 1.663 1.680 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.495 1.611 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.592 1.598 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.563 1.574 dbcsr_complete_redistribute 325 12.2 0.568 0.599 1.479 1.568 mp_allgather_i34 2055 14.4 0.444 1.523 0.444 1.523 acc_transpose_blocks 8220 15.4 0.039 0.041 1.394 1.426 jit_kernel_multiply 7 15.4 0.916 1.387 0.916 1.387 qs_create_task_list 11 7.9 0.000 0.000 1.213 1.314 generate_qs_task_list 11 8.9 0.374 0.442 1.212 1.314 make_basis_sm 11 9.8 0.000 0.000 1.262 1.264 parallel_gemm_fm 81 9.0 0.000 0.000 1.238 1.240 parallel_gemm_fm_cosma 81 10.0 1.238 1.240 1.238 1.240 transfer_rs2pw 451 10.6 0.005 0.006 1.098 1.215 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.175 1.190 mp_waitany 9240 13.8 1.053 1.180 1.053 1.180 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.896000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=806.272727, yerr=14.591971 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 1.421206E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.059 87.237 87.238 qs_mol_dyn_low 1 2.0 0.003 0.003 86.854 86.884 qs_forces 11 3.9 0.003 0.003 86.786 86.787 qs_energies 11 4.9 0.002 0.002 82.674 82.675 scf_env_do_scf 11 5.9 0.000 0.001 71.749 71.750 velocity_verlet 10 3.0 0.002 0.002 55.226 55.232 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.446 43.447 dbcsr_multiply_generic 2055 12.4 0.123 0.127 29.703 29.813 init_scf_loop 11 6.9 0.000 0.000 28.233 28.235 qs_scf_new_mos 99 7.5 0.001 0.001 26.581 26.620 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.580 26.619 prepare_preconditioner 11 7.9 0.000 0.000 26.244 26.252 make_preconditioner 11 8.9 0.000 0.000 26.244 26.252 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.396 25.722 ot_scf_mini 99 9.5 0.002 0.002 24.772 24.799 multiply_cannon 2055 13.4 0.336 0.358 22.719 23.434 multiply_cannon_loop 2055 14.4 0.824 0.842 20.894 21.218 cp_fm_upper_to_full 70 14.2 12.884 18.452 12.884 18.452 ot_mini 99 10.5 0.001 0.001 13.995 14.022 rebuild_ks_matrix 110 8.3 0.001 0.001 13.359 13.395 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.359 13.394 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.129 12.161 dbcsr_complete_redistribute 325 12.2 1.021 1.050 7.371 10.670 multiply_cannon_multrec 8220 15.4 4.079 4.225 9.977 10.185 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.384 9.681 qs_ot_get_derivative 99 11.5 0.001 0.001 9.430 9.457 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.832 9.102 mp_alltoall_i22 605 13.7 5.445 8.720 5.445 8.720 mp_waitall_1 84994 16.7 7.331 8.194 7.331 8.194 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.828 6.867 calculate_rho_elec 110 8.6 0.227 0.227 6.828 6.867 init_scf_run 11 5.9 0.000 0.001 6.612 6.612 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.612 6.612 sum_up_and_integrate 110 10.3 0.002 0.002 6.535 6.550 integrate_v_rspace 110 11.3 0.004 0.004 6.507 6.522 dbcsr_mm_accdrv_process 11614 15.7 3.855 4.167 5.749 5.993 make_m2s 4110 13.4 0.043 0.043 5.259 5.736 make_images 4110 14.4 0.886 0.940 5.070 5.547 qs_ot_get_p 110 10.4 0.001 0.001 5.438 5.468 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.691 5.155 apply_single 110 13.6 0.000 0.000 4.691 5.155 cp_fm_cholesky_invert 11 10.9 5.087 5.091 5.087 5.091 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.786 5.065 ot_diis_step 99 11.5 0.015 0.016 4.546 4.546 fft_wrap_pw1pw2 1111 11.6 0.016 0.016 4.457 4.472 fft_wrap_pw1pw2_140 451 12.1 0.363 0.365 3.910 3.926 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.895 3.902 grid_integrate_task_list 110 12.3 3.675 3.746 3.675 3.746 qs_energies_init_hamiltonians 11 5.9 0.013 0.016 3.708 3.709 hybrid_alltoall_any 4261 16.3 0.267 0.581 2.770 3.486 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.954 3.397 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.380 3.380 make_images_data 4110 15.4 0.044 0.047 2.740 3.334 fft3d_ps 1111 13.6 1.851 1.874 3.317 3.332 calculate_dm_sparse 110 9.5 0.001 0.001 3.304 3.330 density_rs2pw 110 9.6 0.004 0.004 3.290 3.309 wfi_extrapolate 11 7.9 0.001 0.001 3.262 3.262 calculate_first_density_matrix 1 7.0 0.000 0.000 3.240 3.241 multiply_cannon_sync_h2d 8220 15.4 3.125 3.138 3.125 3.138 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.065 3.067 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.801 2.802 cp_fm_diag_elpa_base 48 14.0 2.270 2.458 2.800 2.800 grid_collocate_task_list 110 9.6 2.666 2.674 2.666 2.674 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.618 2.636 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.574 2.587 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.210 2.250 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.102 2.233 potential_pw2rs 110 12.3 0.021 0.021 2.073 2.076 mp_alltoall_d11v 2046 13.8 1.993 2.056 1.993 2.056 cp_fm_cholesky_decompose 22 10.9 1.922 1.945 1.922 1.945 jit_kernel_multiply 10 15.2 1.696 1.941 1.696 1.941 qs_create_task_list 11 7.9 0.000 0.000 1.889 1.937 generate_qs_task_list 11 8.9 0.731 0.785 1.889 1.937 acc_transpose_blocks 8220 15.4 0.041 0.042 1.822 1.859 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.833 1.836 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.794 1.810 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.238000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1275.272727, yerr=64.595371 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 626.495488E+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 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2019702. 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 0.013 0.035 206.890 206.892 qs_mol_dyn_low 1 2.0 0.009 0.053 206.055 206.068 qs_forces 11 3.9 0.077 0.089 205.867 205.869 qs_energies 11 4.9 0.003 0.016 200.215 200.302 scf_env_do_scf 11 5.9 0.001 0.001 183.579 183.583 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 163.327 163.329 dbcsr_multiply_generic 2507 12.6 0.178 0.184 126.235 126.900 qs_scf_new_mos 117 7.6 0.001 0.001 124.424 124.720 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.423 124.720 velocity_verlet 10 3.0 0.006 0.041 122.755 122.757 ot_scf_mini 117 9.6 0.003 0.003 117.661 117.917 multiply_cannon 2507 13.6 0.237 0.245 101.839 103.731 multiply_cannon_loop 2507 14.6 2.378 2.432 99.732 101.653 ot_mini 117 10.6 0.001 0.001 66.917 67.174 multiply_cannon_multrec 60168 15.6 31.846 34.064 41.621 43.490 qs_ot_get_derivative 117 11.6 0.001 0.001 41.949 42.223 rebuild_ks_matrix 128 8.3 0.001 0.001 33.575 34.134 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.021 33.575 34.134 mp_waitall_1 267128 16.5 29.012 32.228 29.012 32.228 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.160 30.669 qs_ot_get_p 128 10.4 0.001 0.002 28.542 28.947 multiply_cannon_sync_h2d 60168 15.6 26.352 28.930 26.352 28.930 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.378 25.259 apply_single 128 13.6 0.001 0.001 24.377 25.259 ot_diis_step 117 11.6 0.007 0.008 24.687 24.688 qs_ot_p2m_diag 83 11.4 0.079 0.092 21.980 22.042 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.353 20.527 init_scf_loop 11 6.9 0.000 0.000 20.178 20.179 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.209 19.210 multiply_cannon_metrocomm3 60168 15.6 0.117 0.122 16.272 18.336 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.274 16.303 cp_fm_diag_elpa_base 83 14.4 16.208 16.240 16.270 16.300 prepare_preconditioner 11 7.9 0.000 0.000 15.601 15.642 make_preconditioner 11 8.9 0.000 0.000 15.601 15.642 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.861 15.011 make_m2s 5014 13.6 0.104 0.112 13.750 14.185 make_images 5014 14.6 0.397 0.416 13.565 14.011 sum_up_and_integrate 128 10.3 0.002 0.005 13.626 13.641 integrate_v_rspace 128 11.3 0.003 0.004 13.567 13.584 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.573 12.762 calculate_rho_elec 128 8.7 0.045 0.064 12.572 12.761 init_scf_run 11 5.9 0.000 0.001 12.446 12.447 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.446 12.447 mp_sum_l 7950 12.9 9.409 10.618 9.409 10.618 dbcsr_mm_accdrv_process 124484 16.2 4.709 4.890 9.340 9.922 multiply_cannon_metrocomm1 60168 15.6 0.091 0.095 6.374 9.191 wfi_extrapolate 11 7.9 0.001 0.003 8.987 8.987 cp_fm_cholesky_invert 11 10.9 8.864 8.871 8.864 8.871 calculate_dm_sparse 128 9.5 0.001 0.001 8.586 8.701 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.102 8.266 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.035 8.131 make_images_data 5014 15.6 0.065 0.072 6.725 7.656 grid_integrate_task_list 128 12.3 7.053 7.576 7.053 7.576 hybrid_alltoall_any 5200 16.5 0.293 2.254 5.919 7.085 density_rs2pw 128 9.7 0.005 0.006 6.159 6.825 cp_dbcsr_sm_fm_multiply 37 9.5 0.006 0.039 6.744 6.758 fft_wrap_pw1pw2 1291 11.7 0.018 0.025 6.402 6.669 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.794 5.946 fft_wrap_pw1pw2_140 523 12.2 0.234 0.244 5.577 5.750 mp_alltoall_d11v 2415 14.1 4.293 5.441 4.293 5.441 fft3d_ps 1291 13.7 2.102 2.625 5.222 5.432 grid_collocate_task_list 128 9.7 4.842 5.275 4.842 5.275 cp_fm_cholesky_decompose 22 10.9 4.585 4.601 4.585 4.601 mp_sum_d 4469 12.1 3.522 4.256 3.522 4.256 potential_pw2rs 128 12.3 0.009 0.010 4.145 4.160 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.892000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=593.000000, yerr=5.768252 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 840.245248E+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 4007 58142. MP_Allreduce 11096 960. MP_Sync 86 MP_Alltoall 1955 5191135. 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 0.046 0.090 188.587 188.589 qs_mol_dyn_low 1 2.0 0.008 0.026 188.018 188.030 qs_forces 11 3.9 0.004 0.004 187.463 187.468 qs_energies 11 4.9 0.003 0.014 180.710 180.724 scf_env_do_scf 11 5.9 0.001 0.001 163.763 163.773 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 131.172 131.174 velocity_verlet 10 3.0 0.013 0.029 117.661 117.674 dbcsr_multiply_generic 2485 12.5 0.184 0.191 96.939 98.101 qs_scf_new_mos 116 7.6 0.001 0.001 93.382 94.006 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.381 94.005 ot_scf_mini 116 9.6 0.004 0.004 88.620 89.354 multiply_cannon 2485 13.5 0.500 0.557 76.909 81.410 multiply_cannon_loop 2485 14.5 1.547 1.613 73.526 76.253 ot_mini 116 10.6 0.001 0.001 50.122 50.824 mp_waitall_1 212858 16.6 23.766 38.556 23.766 38.556 multiply_cannon_multrec 29820 15.5 20.756 25.254 31.368 36.184 rebuild_ks_matrix 127 8.3 0.001 0.001 32.093 32.796 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.022 32.093 32.796 init_scf_loop 11 6.9 0.000 0.000 32.503 32.504 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.846 29.502 qs_ot_get_derivative 116 11.6 0.001 0.002 28.342 29.073 multiply_cannon_metrocomm3 29820 15.5 0.094 0.100 15.633 28.460 prepare_preconditioner 11 7.9 0.000 0.000 28.174 28.265 make_preconditioner 11 8.9 0.000 0.000 28.174 28.265 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.884 27.424 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 22.062 23.197 apply_single 127 13.6 0.001 0.001 22.062 23.197 ot_diis_step 116 11.6 0.014 0.015 21.605 21.607 qs_ot_get_p 127 10.4 0.002 0.004 20.680 21.500 multiply_cannon_sync_h2d 29820 15.5 17.925 20.754 17.925 20.754 cp_fm_cholesky_invert 11 10.9 16.411 16.423 16.411 16.423 qs_ot_p2m_diag 82 11.4 0.208 0.387 15.749 15.784 make_m2s 4970 13.5 0.086 0.093 13.901 15.633 make_images 4970 14.5 1.164 1.368 13.686 15.420 cp_dbcsr_syevd 82 12.4 0.007 0.021 14.549 14.572 sum_up_and_integrate 127 10.3 0.002 0.005 13.531 13.563 integrate_v_rspace 127 11.3 0.003 0.004 13.470 13.505 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.752 12.787 calculate_rho_elec 127 8.7 0.087 0.105 12.752 12.786 init_scf_run 11 5.9 0.000 0.001 11.846 11.847 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.846 11.847 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.101 11.632 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.210 11.253 cp_fm_diag_elpa_base 82 14.4 10.948 11.026 11.191 11.228 dbcsr_mm_accdrv_process 61726 16.2 5.494 6.302 10.057 10.733 multiply_cannon_metrocomm4 27335 15.5 0.104 0.118 3.846 10.642 mp_irecv_dv 68888 16.3 3.638 10.235 3.638 10.235 make_images_data 4970 15.5 0.065 0.075 8.176 10.233 hybrid_alltoall_any 5155 16.4 0.346 1.505 7.168 9.709 wfi_extrapolate 11 7.9 0.002 0.010 8.323 8.323 grid_integrate_task_list 127 12.3 7.153 7.606 7.153 7.606 fft_wrap_pw1pw2 1281 11.7 0.019 0.025 7.286 7.362 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.328 7.318 cp_fm_cholesky_decompose 22 10.9 6.856 6.954 6.856 6.954 density_rs2pw 127 9.7 0.005 0.006 6.398 6.812 mp_sum_l 7884 12.9 4.399 6.705 4.399 6.705 calculate_dm_sparse 127 9.5 0.001 0.001 6.416 6.572 fft_wrap_pw1pw2_140 519 12.2 0.250 0.259 6.415 6.470 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.140 6.149 fft3d_ps 1281 13.7 2.765 2.933 5.735 5.814 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.238 5.435 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.321 5.390 grid_collocate_task_list 127 9.7 4.969 5.331 4.969 5.331 mp_allgather_i34 2485 14.5 1.902 4.868 1.902 4.868 mp_alltoall_d11v 2401 14.1 4.406 4.860 4.406 4.860 potential_pw2rs 127 12.3 0.014 0.016 4.094 4.111 mp_sum_d 4458 12.1 2.676 4.052 2.676 4.052 dbcsr_complete_redistribute 393 12.7 0.764 0.865 3.059 3.909 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=188.589000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=801.727273, yerr=0.862439 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239167967232 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239167967232 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 94.233007E+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 6806383584 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 963.682304E+06 # 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 937890820720 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57629. MP_Allreduce 11184 996. MP_Sync 88 MP_Alltoall 1724 9394234. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.044 175.856 175.856 qs_mol_dyn_low 1 2.0 0.003 0.003 175.432 175.451 qs_forces 11 3.9 0.004 0.005 175.337 175.340 qs_energies 11 4.9 0.005 0.006 168.771 168.781 scf_env_do_scf 11 5.9 0.001 0.001 152.999 153.001 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 117.928 117.928 velocity_verlet 10 3.0 0.001 0.002 112.923 112.925 dbcsr_multiply_generic 2529 12.6 0.185 0.189 82.116 83.091 qs_scf_new_mos 118 7.6 0.001 0.001 80.925 81.298 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.924 81.297 ot_scf_mini 118 9.6 0.003 0.004 76.658 77.114 multiply_cannon 2529 13.6 0.506 0.527 62.216 66.532 multiply_cannon_loop 2529 14.6 1.137 1.198 59.536 62.348 ot_mini 118 10.6 0.001 0.001 42.703 43.154 mp_waitall_1 172006 16.6 24.955 35.075 24.955 35.075 init_scf_loop 11 6.9 0.000 0.000 34.965 34.966 rebuild_ks_matrix 129 8.3 0.001 0.001 30.532 31.029 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.017 30.531 31.029 prepare_preconditioner 11 7.9 0.000 0.000 30.950 31.006 make_preconditioner 11 8.9 0.000 0.000 30.950 31.006 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.477 29.889 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.538 27.995 multiply_cannon_multrec 20232 15.6 13.145 16.686 22.579 26.003 multiply_cannon_metrocomm3 20232 15.6 0.060 0.063 15.493 25.170 qs_ot_get_derivative 118 11.6 0.001 0.002 22.916 23.359 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.760 20.868 apply_single 129 13.6 0.001 0.001 19.760 20.867 qs_ot_get_p 129 10.4 0.001 0.002 19.250 19.737 ot_diis_step 118 11.6 0.018 0.018 19.672 19.673 make_m2s 5058 13.6 0.078 0.084 14.903 15.716 multiply_cannon_sync_h2d 20232 15.6 13.598 15.548 13.598 15.548 make_images 5058 14.6 1.161 1.271 14.666 15.469 qs_ot_p2m_diag 84 11.4 0.268 0.275 14.941 14.950 cp_fm_cholesky_invert 11 10.9 14.328 14.337 14.328 14.337 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.869 13.870 sum_up_and_integrate 129 10.3 0.002 0.003 13.807 13.835 integrate_v_rspace 129 11.3 0.003 0.003 13.746 13.776 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.151 13.188 calculate_rho_elec 129 8.7 0.133 0.147 13.150 13.187 init_scf_run 11 5.9 0.000 0.001 10.759 10.760 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.759 10.759 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.638 10.655 cp_fm_diag_elpa_base 84 14.4 10.224 10.386 10.635 10.652 make_images_data 5058 15.6 0.062 0.073 8.953 10.161 hybrid_alltoall_any 5245 16.5 0.452 2.058 7.676 9.552 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.111 9.468 multiply_cannon_metrocomm4 17703 15.6 0.066 0.076 3.531 9.421 mp_irecv_dv 50659 16.2 3.400 9.158 3.400 9.158 dbcsr_mm_accdrv_process 41846 16.2 5.647 5.968 8.889 9.038 fft_wrap_pw1pw2 1301 11.7 0.019 0.024 7.615 7.730 grid_integrate_task_list 129 12.3 7.354 7.700 7.354 7.700 wfi_extrapolate 11 7.9 0.001 0.001 7.373 7.373 cp_fm_cholesky_decompose 22 10.9 7.329 7.366 7.329 7.366 cp_fm_upper_to_full 106 14.8 5.748 7.294 5.748 7.294 density_rs2pw 129 9.7 0.005 0.006 6.467 6.990 fft_wrap_pw1pw2_140 527 12.2 0.286 0.295 6.694 6.816 dbcsr_complete_redistribute 397 12.7 1.324 1.439 4.929 6.701 fft3d_ps 1301 13.7 2.791 3.031 5.898 5.979 calculate_dm_sparse 129 9.5 0.001 0.001 5.875 5.953 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.678 5.683 grid_collocate_task_list 129 9.7 5.196 5.612 5.196 5.612 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.763 5.532 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.630 5.228 mp_alltoall_d11v 2429 14.1 4.422 5.074 4.422 5.074 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.546 4.667 mp_sum_l 8016 12.9 3.162 4.624 3.162 4.624 mp_allgather_i34 2529 14.6 1.166 4.356 1.166 4.356 potential_pw2rs 129 12.3 0.020 0.022 4.169 4.180 transfer_fm_to_dbcsr 11 9.9 0.018 0.025 2.454 4.171 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.991 4.024 mp_alltoall_i22 720 14.1 1.955 3.835 1.955 3.835 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.717 3.720 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.508 3.549 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.856000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=916.545455, yerr=8.478461 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/18/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 4.353788E+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 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 1.177612E+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 1.770319E+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 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58348. MP_Allreduce 11057 1083. MP_Sync 87 MP_Alltoall 1712 12503084. 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 0.032 0.073 189.543 189.545 qs_mol_dyn_low 1 2.0 0.049 0.371 188.917 188.930 qs_forces 11 3.9 0.005 0.012 188.288 188.298 qs_energies 11 4.9 0.003 0.014 181.168 181.177 scf_env_do_scf 11 5.9 0.001 0.002 163.368 163.378 velocity_verlet 10 3.0 0.041 0.071 124.379 124.399 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.939 116.940 qs_scf_new_mos 117 7.6 0.001 0.001 81.199 81.546 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.199 81.545 dbcsr_multiply_generic 2507 12.6 0.190 0.194 80.434 81.347 ot_scf_mini 117 9.6 0.003 0.004 76.652 76.992 multiply_cannon 2507 13.6 0.553 0.585 55.909 58.193 multiply_cannon_loop 2507 14.6 1.843 1.919 52.154 54.324 init_scf_loop 11 6.9 0.001 0.004 46.306 46.308 ot_mini 117 10.6 0.001 0.001 42.987 43.330 prepare_preconditioner 11 7.9 0.000 0.000 42.207 42.235 make_preconditioner 11 8.9 0.000 0.001 42.207 42.235 make_full_inverse_cholesky 11 9.9 0.011 0.023 35.883 40.854 multiply_cannon_multrec 30084 15.6 13.519 18.651 26.466 31.286 rebuild_ks_matrix 128 8.3 0.001 0.001 29.329 29.692 qs_ks_build_kohn_sham_matrix 128 9.3 0.021 0.052 29.328 29.691 mp_waitall_1 147882 16.7 17.107 28.010 17.107 28.010 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.482 26.802 qs_ot_get_derivative 117 11.6 0.001 0.002 23.243 23.576 make_m2s 5014 13.6 0.093 0.097 20.152 21.251 make_images 5014 14.6 1.948 2.243 19.842 20.943 qs_ot_get_p 128 10.4 0.001 0.002 19.795 20.159 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.162 19.830 apply_single 128 13.6 0.001 0.001 19.161 19.829 ot_diis_step 117 11.6 0.018 0.019 19.623 19.624 cp_fm_upper_to_full 105 14.8 11.482 16.863 11.482 16.863 cp_fm_cholesky_invert 11 10.9 16.147 16.156 16.147 16.156 qs_ot_p2m_diag 83 11.4 0.343 0.390 15.611 15.662 multiply_cannon_metrocomm3 30084 15.6 0.050 0.054 6.468 15.557 cp_dbcsr_syevd 83 12.4 0.006 0.008 14.201 14.203 sum_up_and_integrate 128 10.3 0.002 0.003 13.675 13.702 integrate_v_rspace 128 11.3 0.003 0.004 13.614 13.643 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.141 13.185 calculate_rho_elec 128 8.7 0.174 0.190 13.140 13.185 dbcsr_mm_accdrv_process 62264 16.2 8.479 9.304 12.511 12.982 dbcsr_complete_redistribute 395 12.7 1.493 1.638 9.078 12.819 make_images_data 5014 15.6 0.064 0.070 10.759 12.436 multiply_cannon_sync_h2d 30084 15.6 10.324 11.736 10.324 11.736 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.713 11.469 init_scf_run 11 5.9 0.000 0.001 11.375 11.377 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.375 11.377 hybrid_alltoall_any 5200 16.5 0.528 2.209 9.719 11.350 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.941 10.959 cp_fm_diag_elpa_base 83 14.4 9.960 10.286 10.933 10.949 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.795 10.030 transfer_fm_to_dbcsr 11 9.9 0.002 0.009 6.302 9.959 mp_alltoall_i22 716 14.1 5.578 9.336 5.578 9.336 grid_integrate_task_list 128 12.3 7.472 7.871 7.472 7.871 fft_wrap_pw1pw2 1291 11.7 0.019 0.024 7.781 7.837 multiply_cannon_metrocomm4 25070 15.6 0.079 0.090 2.864 7.738 cp_fm_cholesky_decompose 22 10.9 7.530 7.641 7.530 7.641 wfi_extrapolate 11 7.9 0.001 0.001 7.574 7.575 mp_irecv_dv 76098 16.2 2.708 7.453 2.708 7.453 fft_wrap_pw1pw2_140 523 12.2 0.328 0.337 6.871 6.955 density_rs2pw 128 9.7 0.006 0.013 6.220 6.606 calculate_dm_sparse 128 9.5 0.001 0.001 6.230 6.323 fft3d_ps 1291 13.7 2.945 3.001 5.960 6.007 grid_collocate_task_list 128 9.7 5.283 5.590 5.283 5.590 mp_alltoall_d11v 2415 14.1 4.946 5.573 4.946 5.573 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.503 5.545 qs_energies_init_hamiltonians 11 5.9 0.065 0.096 4.746 4.749 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.464 4.565 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.402 4.501 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.249 4.307 potential_pw2rs 128 12.3 0.022 0.023 3.952 3.961 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.545000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1102.272727, yerr=26.917689 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410020794368 0.0% 0.0% 100.0% flops 32 x 32 x 32 1947767668736 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957875093504 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544039424 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714619674624 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444697722880 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019193401344 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019193401344 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624869085184 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.820223E+12 0.0% 0.0% 100.0% flops max/rank 5.821533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705860672 0.0% 0.0% 100.0% number of processed stacks 1948544 0.0% 0.0% 100.0% average stack size 0.0 0.0 3441.5 marketing flops 143.531365E+12 ------------------------------------------------------------------------------- # multiplications 2496 max memory usage/rank 1.569341E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 239616 MPI messages size (bytes): total size 1.321658E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.515734E+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 113856 59693334528 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108831376 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8885 51. MP_Alltoall 9606 802511. MP_ISend 39892 2096308. MP_IRecv 39892 2095413. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3951 58940. MP_Allreduce 10955 1170. MP_Sync 86 MP_Alltoall 1700 18828273. 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 0.020 0.049 170.354 170.355 qs_mol_dyn_low 1 2.0 0.003 0.003 169.946 169.968 qs_forces 11 3.9 0.004 0.004 169.818 169.827 qs_energies 11 4.9 0.140 0.141 162.438 162.444 scf_env_do_scf 11 5.9 0.001 0.001 144.957 144.969 velocity_verlet 10 3.0 0.014 0.015 110.911 110.916 scf_env_do_scf_inner_loop 116 6.6 0.004 0.012 109.000 109.001 qs_scf_new_mos 116 7.6 0.001 0.001 72.856 72.954 qs_scf_loop_do_ot 116 8.6 0.001 0.001 72.856 72.953 dbcsr_multiply_generic 2496 12.6 0.185 0.192 72.084 72.592 ot_scf_mini 116 9.6 0.003 0.003 68.463 68.553 multiply_cannon 2496 13.6 0.556 0.580 53.480 56.674 multiply_cannon_loop 2496 14.6 0.809 0.835 50.394 51.453 ot_mini 116 10.6 0.001 0.001 37.777 37.865 init_scf_loop 11 6.9 0.000 0.000 35.802 35.804 prepare_preconditioner 11 7.9 0.000 0.000 31.963 31.993 make_preconditioner 11 8.9 0.000 0.000 31.963 31.993 mp_waitall_1 125208 16.7 24.923 31.444 24.923 31.444 make_full_inverse_cholesky 11 9.9 0.017 0.029 29.765 30.046 rebuild_ks_matrix 127 8.3 0.001 0.001 28.929 29.068 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 28.929 29.067 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.368 26.494 multiply_cannon_multrec 9984 15.6 10.248 15.142 17.819 22.375 qs_ot_get_derivative 116 11.6 0.001 0.002 20.583 20.673 multiply_cannon_metrocomm3 9984 15.6 0.024 0.026 12.975 20.412 cp_fm_cholesky_invert 11 10.9 18.268 18.275 18.268 18.275 qs_ot_get_p 127 10.4 0.001 0.001 18.014 18.149 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.259 17.451 apply_single 127 13.6 0.001 0.001 17.258 17.451 ot_diis_step 116 11.6 0.020 0.021 17.120 17.120 make_m2s 4992 13.6 0.063 0.068 14.688 15.614 make_images 4992 14.6 2.149 2.573 14.383 15.304 qs_ot_p2m_diag 82 11.4 0.489 0.495 14.214 14.230 sum_up_and_integrate 127 10.3 0.002 0.002 13.794 13.847 integrate_v_rspace 127 11.3 0.004 0.004 13.733 13.789 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.490 13.508 calculate_rho_elec 127 8.7 0.255 0.266 13.489 13.507 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.048 13.049 multiply_cannon_sync_h2d 9984 15.6 10.645 11.154 10.645 11.154 init_scf_run 11 5.9 0.000 0.001 10.408 10.408 scf_env_initial_rho_setup 11 6.9 0.003 0.003 10.407 10.408 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.160 10.170 cp_fm_diag_elpa_base 82 14.4 9.928 9.999 10.155 10.165 make_images_data 4992 15.6 0.055 0.066 8.400 9.799 hybrid_alltoall_any 5177 16.4 0.841 3.767 8.230 9.402 fft_wrap_pw1pw2 1281 11.7 0.017 0.019 8.093 8.121 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 8.009 8.074 grid_integrate_task_list 127 12.3 7.722 8.055 7.722 8.055 cp_fm_cholesky_decompose 22 10.9 7.910 8.011 7.910 8.011 dbcsr_mm_accdrv_process 20634 16.1 3.105 4.451 7.225 7.915 wfi_extrapolate 11 7.9 0.001 0.001 7.319 7.319 multiply_cannon_metrocomm1 9984 15.6 0.029 0.030 4.279 7.165 fft_wrap_pw1pw2_140 519 12.2 0.431 0.437 7.106 7.125 density_rs2pw 127 9.7 0.005 0.005 6.182 6.358 calculate_dm_sparse 127 9.5 0.001 0.001 6.078 6.155 fft3d_ps 1281 13.7 3.087 3.208 6.062 6.082 grid_collocate_task_list 127 9.7 5.519 5.721 5.519 5.721 dbcsr_complete_redistribute 393 12.7 2.100 2.201 5.326 5.715 multiply_cannon_metrocomm4 7488 15.6 0.025 0.028 1.850 5.386 qs_energies_init_hamiltonians 11 5.9 0.048 0.077 5.354 5.355 mp_alltoall_d11v 2401 14.1 4.767 5.337 4.767 5.337 mp_irecv_dv 28706 15.9 1.809 5.294 1.809 5.294 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.190 5.205 mp_allgather_i34 2496 14.6 1.282 4.573 1.282 4.573 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.084 4.129 potential_pw2rs 127 12.3 0.026 0.027 4.027 4.035 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.580 3.873 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.452 3.760 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.640 3.724 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.636 3.676 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.505 3.582 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.459 3.471 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.355000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1488.909091, yerr=20.065595 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 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 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 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 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.222086E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+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 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3949 59902. MP_Allreduce 10950 1507. MP_Sync 86 MP_Alltoall 1700 36954374. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.087 0.117 292.452 292.453 qs_mol_dyn_low 1 2.0 0.003 0.003 291.645 291.663 qs_forces 11 3.9 0.023 0.062 290.949 290.951 qs_energies 11 4.9 0.007 0.008 282.166 282.191 scf_env_do_scf 11 5.9 0.001 0.001 259.634 259.642 velocity_verlet 10 3.0 0.028 0.035 209.267 209.350 scf_env_do_scf_inner_loop 116 6.6 0.004 0.008 133.049 133.051 init_scf_loop 11 6.9 0.000 0.000 126.307 126.311 prepare_preconditioner 11 7.9 0.000 0.000 119.764 119.784 make_preconditioner 11 8.9 0.000 0.000 119.763 119.784 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.690 116.899 qs_scf_new_mos 116 7.6 0.001 0.001 90.556 90.644 qs_scf_loop_do_ot 116 8.6 0.001 0.001 90.555 90.643 ot_scf_mini 116 9.6 0.004 0.004 85.702 85.737 dbcsr_multiply_generic 2485 12.5 0.243 0.254 82.305 82.897 cp_fm_upper_to_full 104 14.8 53.021 76.086 53.021 76.086 multiply_cannon 2485 13.5 0.681 0.733 58.759 60.070 multiply_cannon_loop 2485 14.5 1.039 1.054 54.638 56.078 ot_mini 116 10.6 0.001 0.001 44.700 44.757 dbcsr_complete_redistribute 393 12.7 3.997 4.030 29.987 42.899 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.496 39.368 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 24.032 36.700 mp_alltoall_i22 712 14.1 21.779 34.759 21.779 34.759 rebuild_ks_matrix 127 8.3 0.001 0.001 33.627 33.685 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.018 33.626 33.684 cp_fm_cholesky_invert 11 10.9 33.503 33.509 33.503 33.509 mp_waitall_1 102768 16.8 28.194 32.413 28.194 32.413 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.226 31.285 qs_ot_get_p 127 10.4 0.680 0.680 27.410 27.504 qs_ot_get_derivative 116 11.6 0.002 0.002 24.790 24.827 qs_ot_p2m_diag 82 11.4 0.868 0.873 22.004 22.032 make_m2s 4970 13.5 0.073 0.074 18.843 20.396 cp_dbcsr_syevd 82 12.4 0.006 0.006 20.217 20.220 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 18.771 20.195 make_images 4970 14.5 3.056 3.260 18.364 19.918 ot_diis_step 116 11.6 0.022 0.022 19.870 19.870 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.235 19.390 apply_single 127 13.6 0.001 0.001 19.235 19.389 multiply_cannon_multrec 9940 15.5 10.240 12.060 17.994 18.061 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.027 17.028 cp_fm_diag_elpa_base 82 14.4 12.646 14.277 17.023 17.023 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.798 15.824 calculate_rho_elec 127 8.7 0.479 0.480 15.798 15.824 sum_up_and_integrate 127 10.3 0.002 0.002 15.636 15.728 integrate_v_rspace 127 11.3 0.004 0.004 15.573 15.667 multiply_cannon_sync_h2d 9940 15.5 14.208 14.244 14.208 14.244 hybrid_alltoall_any 5155 16.4 1.307 3.042 10.866 12.995 make_images_data 4970 15.5 0.061 0.066 10.703 12.922 init_scf_run 11 5.9 0.000 0.001 12.255 12.256 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.255 12.256 fft_wrap_pw1pw2 1281 11.7 0.020 0.021 10.012 10.041 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.696 9.719 dbcsr_mm_accdrv_process 20590 16.0 4.342 6.131 7.506 9.374 cp_fm_cholesky_decompose 22 10.9 8.952 8.976 8.952 8.976 fft_wrap_pw1pw2_140 519 12.2 0.818 0.819 8.848 8.880 wfi_extrapolate 11 7.9 0.001 0.001 8.825 8.825 grid_integrate_task_list 127 12.3 8.498 8.701 8.498 8.701 qs_energies_init_hamiltonians 11 5.9 0.003 0.004 8.149 8.166 fft3d_ps 1281 13.7 3.944 3.968 7.314 7.332 mp_alltoall_d11v 2401 14.1 6.872 7.027 6.872 7.027 density_rs2pw 127 9.7 0.005 0.005 6.884 6.929 calculate_dm_sparse 127 9.5 0.001 0.001 6.711 6.742 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.457 6.488 grid_collocate_task_list 127 9.7 6.366 6.446 6.366 6.446 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.146 6.200 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=292.453000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2923.090909, yerr=161.272471 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 1.283158E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 18885162. 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 0.093 0.151 85.169 85.170 qs_energies 1 2.0 0.000 0.000 84.522 84.529 ls_scf 1 3.0 0.000 0.000 83.635 83.643 dbcsr_multiply_generic 111 6.7 0.017 0.022 72.694 72.842 multiply_cannon 111 7.7 0.017 0.020 55.784 57.066 multiply_cannon_loop 111 8.7 0.227 0.249 52.326 53.908 ls_scf_main 1 4.0 0.000 0.000 52.433 52.433 density_matrix_trs4 2 5.0 0.002 0.003 46.850 46.932 ls_scf_init_scf 1 4.0 0.000 0.000 28.164 28.165 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.021 27.069 mp_waitall_1 11031 10.9 22.326 25.150 22.326 25.150 matrix_sqrt_Newton_Schulz 2 6.5 0.004 0.008 24.933 24.953 multiply_cannon_multrec 2664 9.7 8.138 8.802 15.587 17.266 multiply_cannon_sync_h2d 2664 9.7 13.599 15.590 13.599 15.590 make_m2s 222 7.7 0.008 0.011 13.198 13.621 make_images 222 8.7 0.098 0.109 13.176 13.602 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.573 11.748 make_images_data 222 9.7 0.004 0.005 7.728 8.234 hybrid_alltoall_any 227 10.6 0.215 1.840 6.589 8.193 dbcsr_mm_accdrv_process 4760 10.4 0.585 0.705 7.066 8.020 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.360 7.822 dbcsr_mm_accdrv_process_sort 4760 11.4 6.282 7.158 6.282 7.158 calculate_norms 4752 9.8 5.506 6.112 5.506 6.112 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.046 5.208 mp_sum_l 887 5.1 3.181 5.101 3.181 5.101 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.364 3.748 multiply_cannon_metrocomm4 2442 9.7 0.011 0.015 2.050 3.623 mp_irecv_dv 6231 10.9 2.034 3.596 2.034 3.596 make_images_sizes 222 9.7 0.000 0.000 0.763 3.288 mp_alltoall_i44 222 10.7 0.763 3.287 0.763 3.287 arnoldi_extremal 4 6.8 0.000 0.000 3.181 3.212 arnoldi_normal_ev 4 7.8 0.001 0.003 3.180 3.212 build_subspace 16 8.4 0.009 0.012 3.088 3.090 ls_scf_post 1 4.0 0.000 0.000 3.039 3.045 ls_scf_store_result 1 5.0 0.000 0.000 2.855 2.895 dbcsr_special_finalize 555 9.7 0.005 0.006 2.345 2.841 dbcsr_merge_single_wm 555 10.7 0.467 0.596 2.337 2.832 make_images_pack 222 9.7 2.214 2.638 2.216 2.639 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.314 2.554 dbcsr_sort_data 658 11.4 2.129 2.524 2.129 2.524 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.457 2.069 2.459 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.322 2.401 buffer_matrices_ensure_size 222 8.7 1.771 2.090 1.771 2.090 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.828 1.829 rebuild_ks_matrix 3 7.3 0.000 0.000 1.818 1.819 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.818 1.819 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.170000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1135.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 2.176586E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.122 0.169 90.011 90.013 qs_energies 1 2.0 0.000 0.000 89.278 89.286 ls_scf 1 3.0 0.000 0.000 87.845 87.854 dbcsr_multiply_generic 111 6.7 0.015 0.016 73.951 74.310 multiply_cannon 111 7.7 0.029 0.072 52.623 56.587 ls_scf_main 1 4.0 0.000 0.000 54.324 54.331 multiply_cannon_loop 111 8.7 0.135 0.148 49.875 52.882 density_matrix_trs4 2 5.0 0.002 0.003 48.633 48.860 ls_scf_init_scf 1 4.0 0.000 0.001 29.852 29.854 mp_waitall_1 9105 10.9 20.590 29.651 20.590 29.651 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.657 28.728 multiply_cannon_multrec 1332 9.7 13.318 17.667 22.667 28.172 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.007 26.286 26.298 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.390 20.469 make_m2s 222 7.7 0.006 0.008 15.060 15.618 make_images 222 8.7 1.374 1.699 15.030 15.588 dbcsr_mm_accdrv_process 4041 10.4 0.338 0.538 8.943 10.505 dbcsr_mm_accdrv_process_sort 4041 11.4 8.461 9.971 8.461 9.971 make_images_data 222 9.7 0.004 0.005 8.711 9.676 hybrid_alltoall_any 227 10.6 0.541 2.540 8.110 9.286 mp_sum_l 887 5.1 4.823 7.985 4.823 7.985 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.198 7.903 mp_irecv_dv 3311 11.0 3.177 7.847 3.177 7.847 calculate_norms 2376 9.8 6.071 6.840 6.071 6.840 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.624 6.292 multiply_cannon_sync_h2d 1332 9.7 4.921 6.277 4.921 6.277 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.922 5.160 arnoldi_extremal 4 6.8 0.000 0.000 4.655 4.670 arnoldi_normal_ev 4 7.8 0.001 0.005 4.655 4.670 build_subspace 16 8.4 0.014 0.021 4.410 4.413 ls_scf_post 1 4.0 0.002 0.012 3.669 3.677 ls_scf_store_result 1 5.0 0.000 0.000 3.380 3.490 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.163 3.411 dbcsr_matrix_vector_mult_local 304 10.0 2.761 3.250 2.763 3.251 mp_allgather_i34 111 8.7 0.712 2.868 0.712 2.868 ls_scf_dm_to_ks 2 5.0 0.000 0.007 2.616 2.719 dbcsr_data_new 4174 10.1 2.141 2.430 2.141 2.430 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.137 2.385 make_images_pack 222 9.7 1.823 2.139 1.825 2.141 dbcsr_sort_data 436 11.2 1.844 2.131 1.844 2.131 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.881 1.883 rebuild_ks_matrix 3 7.3 0.000 0.000 1.868 1.870 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.868 1.870 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.013000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1777.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 2.869322E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.153 0.229 93.863 93.864 qs_energies 1 2.0 0.000 0.000 92.820 92.824 ls_scf 1 3.0 0.000 0.000 91.392 91.395 dbcsr_multiply_generic 111 6.7 0.016 0.016 76.045 76.283 ls_scf_main 1 4.0 0.000 0.000 57.220 57.223 multiply_cannon 111 7.7 0.034 0.101 52.449 56.820 multiply_cannon_loop 111 8.7 0.116 0.131 49.666 53.974 density_matrix_trs4 2 5.0 0.002 0.003 51.257 51.428 mp_waitall_1 7281 11.0 24.116 34.493 24.116 34.493 ls_scf_init_scf 1 4.0 0.000 0.001 30.459 30.462 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.277 29.356 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.008 26.912 26.923 multiply_cannon_multrec 888 9.7 12.643 15.710 21.297 24.321 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.473 23.732 make_m2s 222 7.7 0.006 0.007 16.690 17.433 make_images 222 8.7 1.585 1.851 16.653 17.392 make_images_data 222 9.7 0.004 0.004 9.778 10.910 hybrid_alltoall_any 227 10.6 0.641 2.945 9.296 10.568 dbcsr_mm_accdrv_process 3754 10.4 0.310 0.487 8.162 9.396 mp_sum_l 887 5.1 5.267 9.272 5.267 9.272 dbcsr_mm_accdrv_process_sort 3754 11.4 7.729 8.909 7.729 8.909 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.084 7.358 multiply_cannon_sync_h2d 888 9.7 6.035 7.224 6.035 7.224 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.478 6.894 mp_irecv_dv 2335 11.1 2.462 6.826 2.462 6.826 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.508 6.724 arnoldi_extremal 4 6.8 0.000 0.000 5.160 5.175 arnoldi_normal_ev 4 7.8 0.010 0.023 5.160 5.175 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.783 5.003 build_subspace 16 8.4 0.014 0.020 4.851 4.856 calculate_norms 1584 9.8 4.410 4.854 4.410 4.854 mp_allgather_i34 111 8.7 0.892 3.842 0.892 3.842 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.493 3.822 ls_scf_post 1 4.0 0.000 0.000 3.713 3.717 dbcsr_matrix_vector_mult_local 304 10.0 3.047 3.633 3.049 3.635 ls_scf_store_result 1 5.0 0.000 0.000 3.463 3.541 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.879 2.990 dbcsr_data_new 4116 9.9 2.112 2.463 2.112 2.463 make_images_sizes 222 9.7 0.000 0.000 1.048 2.189 mp_alltoall_i44 222 10.7 1.048 2.189 1.048 2.189 dbcsr_sort_data 325 11.1 1.895 2.107 1.895 2.107 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.925 1.926 rebuild_ks_matrix 3 7.3 0.000 0.000 1.907 1.908 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.907 1.908 make_images_pack 222 9.7 1.624 1.875 1.627 1.878 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.864000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2230.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 3.399487E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.043 97.992 97.992 qs_energies 1 2.0 0.000 0.000 97.354 97.361 ls_scf 1 3.0 0.000 0.000 95.642 95.648 dbcsr_multiply_generic 111 6.7 0.016 0.017 79.217 79.487 ls_scf_main 1 4.0 0.000 0.000 59.567 59.568 multiply_cannon 111 7.7 0.046 0.145 52.080 56.687 density_matrix_trs4 2 5.0 0.002 0.003 53.385 53.485 multiply_cannon_loop 111 8.7 0.152 0.170 47.092 50.270 ls_scf_init_scf 1 4.0 0.000 0.000 32.783 32.784 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.599 31.662 mp_waitall_1 6369 11.0 23.239 29.548 23.239 29.548 matrix_sqrt_Newton_Schulz 2 6.5 0.010 0.021 29.017 29.028 multiply_cannon_multrec 1332 9.7 14.074 16.979 22.053 25.329 make_m2s 222 7.7 0.006 0.007 21.232 22.599 make_images 222 8.7 3.159 3.622 21.182 22.551 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.625 18.048 make_images_data 222 9.7 0.004 0.004 11.872 13.466 hybrid_alltoall_any 227 10.6 0.801 3.819 11.375 13.209 dbcsr_mm_accdrv_process 3641 10.4 0.282 0.487 7.619 9.175 dbcsr_mm_accdrv_process_sort 3641 11.4 7.174 8.674 7.174 8.674 mp_sum_l 887 5.1 4.319 8.280 4.319 8.280 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.371 6.642 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.096 6.130 mp_irecv_dv 3229 10.9 2.072 6.040 2.072 6.040 multiply_cannon_sync_h2d 1332 9.7 5.425 6.021 5.425 6.021 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.575 5.470 arnoldi_extremal 4 6.8 0.000 0.000 5.281 5.295 arnoldi_normal_ev 4 7.8 0.001 0.005 5.281 5.295 build_subspace 16 8.4 0.014 0.021 4.945 4.953 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.605 4.763 calculate_norms 2376 9.8 4.222 4.547 4.222 4.547 mp_allgather_i34 111 8.7 2.098 4.283 2.098 4.283 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 3.628 3.944 dbcsr_matrix_vector_mult_local 304 10.0 3.234 3.744 3.236 3.746 dbcsr_sort_data 658 11.4 3.133 3.407 3.133 3.407 ls_scf_post 1 4.0 0.000 0.000 3.292 3.299 dbcsr_special_finalize 555 9.7 0.006 0.007 2.890 3.177 dbcsr_merge_single_wm 555 10.7 0.546 0.671 2.882 3.170 ls_scf_store_result 1 5.0 0.000 0.000 3.026 3.094 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.031 3.090 dbcsr_data_release 10477 10.7 1.580 2.470 1.580 2.470 dbcsr_finalize 304 7.8 0.049 0.061 1.805 2.015 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.992000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2743.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_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 4.832817E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.054 93.456 93.456 qs_energies 1 2.0 0.000 0.000 92.544 92.548 ls_scf 1 3.0 0.000 0.000 90.568 90.572 dbcsr_multiply_generic 111 6.7 0.017 0.019 71.869 72.048 ls_scf_main 1 4.0 0.000 0.000 57.465 57.485 multiply_cannon 111 7.7 0.083 0.129 53.167 56.395 multiply_cannon_loop 111 8.7 0.089 0.095 50.535 52.187 density_matrix_trs4 2 5.0 0.002 0.003 50.456 50.541 ls_scf_init_scf 1 4.0 0.000 0.000 29.642 29.660 mp_waitall_1 5436 11.0 25.402 29.424 25.402 29.424 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.397 28.434 matrix_sqrt_Newton_Schulz 2 6.5 0.008 0.018 26.075 26.086 multiply_cannon_multrec 444 9.7 13.607 16.110 20.734 22.080 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.168 16.110 make_m2s 222 7.7 0.005 0.005 13.908 14.891 make_images 222 8.7 2.043 2.479 13.841 14.821 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.437 14.559 hybrid_alltoall_any 227 10.6 0.802 3.823 8.585 10.395 make_images_data 222 9.7 0.003 0.004 8.613 9.934 multiply_cannon_sync_h2d 444 9.7 6.707 8.117 6.707 8.117 dbcsr_mm_accdrv_process 3003 10.4 0.316 0.397 6.822 7.957 dbcsr_mm_accdrv_process_sort 3003 11.4 6.461 7.559 6.461 7.559 arnoldi_extremal 4 6.8 0.000 0.000 5.902 5.914 arnoldi_normal_ev 4 7.8 0.002 0.005 5.902 5.914 build_subspace 16 8.4 0.015 0.019 5.497 5.506 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.481 4.686 mp_sum_l 887 5.1 2.807 4.564 2.807 4.564 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.219 4.419 dbcsr_matrix_vector_mult_local 304 10.0 3.750 4.207 3.752 4.209 mp_allgather_i34 111 8.7 1.203 3.848 1.203 3.848 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.542 3.781 mp_irecv_dv 1241 11.2 1.529 3.739 1.529 3.739 calculate_norms 792 9.8 3.603 3.717 3.603 3.717 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.505 3.558 make_images_sizes 222 9.7 0.000 0.000 0.901 3.506 mp_alltoall_i44 222 10.7 0.901 3.506 0.901 3.506 ls_scf_post 1 4.0 0.000 0.000 3.461 3.465 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.975 3.319 ls_scf_store_result 1 5.0 0.000 0.000 3.249 3.285 dbcsr_data_new 4608 9.7 1.798 2.305 1.798 2.305 dbcsr_finalize 304 7.8 0.062 0.077 2.200 2.276 dbcsr_merge_all 275 8.9 0.480 0.520 2.056 2.112 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.063 2.080 rebuild_ks_matrix 3 7.3 0.000 0.000 2.030 2.047 qs_ks_build_kohn_sham_matrix 3 8.3 0.020 0.037 2.030 2.047 qs_energies_init_hamiltonians 1 3.0 0.038 0.064 1.960 1.960 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=93.456000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3757.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/26/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 30.358840E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 339931 0.0% 0.0% 100.0% average stack size 0.0 0.0 28341.7 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 9.045955E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.219 0.394 111.225 111.226 qs_energies 1 2.0 0.000 0.000 109.341 109.346 ls_scf 1 3.0 0.000 0.000 106.398 106.405 dbcsr_multiply_generic 111 6.7 0.024 0.026 79.722 79.857 ls_scf_main 1 4.0 0.000 0.000 64.363 64.364 multiply_cannon 111 7.7 0.141 0.301 53.444 55.801 density_matrix_trs4 2 5.0 0.002 0.003 55.489 55.541 multiply_cannon_loop 111 8.7 0.104 0.135 50.374 50.744 ls_scf_init_scf 1 4.0 0.000 0.000 38.160 38.161 ls_scf_init_matrix_S 1 5.0 0.000 0.000 36.710 36.732 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 33.991 33.997 mp_waitall_1 4527 11.1 23.231 26.394 23.231 26.394 make_m2s 222 7.7 0.005 0.005 22.650 23.818 make_images 222 8.7 3.587 3.875 22.542 23.708 multiply_cannon_multrec 444 9.7 17.817 18.486 22.501 23.135 hybrid_alltoall_any 227 10.6 1.660 3.642 12.760 15.667 make_images_data 222 9.7 0.003 0.004 12.974 15.188 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.604 15.097 multiply_cannon_sync_h2d 444 9.7 8.784 8.835 8.784 8.835 arnoldi_extremal 4 6.8 0.000 0.000 7.535 7.545 arnoldi_normal_ev 4 7.8 0.003 0.009 7.535 7.544 build_subspace 16 8.4 0.026 0.037 6.993 7.004 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.598 5.740 dbcsr_matrix_vector_mult_local 304 10.0 5.150 5.425 5.152 5.428 compute_matrix_preconditioner 1 6.0 0.002 0.002 5.271 5.277 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.960 5.201 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.957 5.043 dbcsr_mm_accdrv_process 1814 10.4 0.325 0.390 4.510 4.681 acc_transpose_blocks 444 9.7 0.003 0.003 3.696 4.345 dbcsr_mm_accdrv_process_sort 1814 11.4 4.162 4.291 4.162 4.291 acc_transpose_blocks_kernels 444 10.7 0.006 0.006 3.613 4.264 jit_kernel_transpose 1 13.0 3.607 4.258 3.607 4.258 ls_scf_post 1 4.0 0.000 0.000 3.875 3.880 mp_allgather_i34 111 8.7 1.184 3.721 1.184 3.721 make_images_sizes 222 9.7 0.000 0.000 1.512 3.686 mp_alltoall_i44 222 10.7 1.512 3.685 1.512 3.685 ls_scf_store_result 1 5.0 0.000 0.000 3.578 3.615 calculate_norms 792 9.8 3.235 3.271 3.235 3.271 dbcsr_finalize 304 7.8 0.082 0.089 3.087 3.170 dbcsr_merge_all 275 8.9 0.886 0.914 2.872 2.947 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.912 2.912 dbcsr_complete_redistribute 5 7.6 1.442 1.506 2.765 2.888 dbcsr_new_transposed 4 7.5 0.469 0.705 2.733 2.747 matrix_ls_to_qs 2 6.0 0.000 0.000 2.408 2.536 dbcsr_sort_data 325 11.1 2.449 2.507 2.449 2.507 dbcsr_redistribute 4 8.5 1.369 1.414 2.236 2.480 dbcsr_data_new 6591 9.6 1.861 2.392 1.861 2.392 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.322 2.323 rebuild_ks_matrix 3 7.3 0.000 0.000 2.256 2.258 qs_ks_build_kohn_sham_matrix 3 8.3 0.006 0.007 2.256 2.258 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=111.226000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7229.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 588.312576E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 4453214. 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 0.037 0.087 237.619 237.621 qs_mol_dyn_low 1 2.0 0.013 0.052 236.484 236.511 qs_forces 5 3.8 0.004 0.011 236.290 236.297 qs_energies 5 4.8 0.006 0.051 233.392 233.421 scf_env_do_scf 5 5.8 0.001 0.029 213.680 213.684 scf_env_do_scf_inner_loop 105 6.6 0.008 0.090 186.444 186.451 qs_scf_new_mos 105 7.6 0.000 0.001 145.614 145.825 qs_scf_loop_do_ot 105 8.6 0.001 0.009 145.613 145.824 dbcsr_multiply_generic 1445 12.2 0.131 0.139 135.485 135.895 ot_scf_mini 105 9.6 0.004 0.020 135.611 135.764 multiply_cannon 1445 13.2 0.273 0.290 115.664 117.715 multiply_cannon_loop 1445 14.2 2.841 2.972 113.899 115.104 velocity_verlet 4 3.0 0.031 0.120 106.350 106.361 ot_mini 105 10.6 0.001 0.006 61.376 61.484 multiply_cannon_multrec 69360 15.2 29.677 34.407 39.602 44.667 qs_ot_get_p 112 10.4 0.001 0.010 44.150 44.469 mp_waitall_1 488190 16.1 36.371 44.357 36.371 44.357 qs_ot_get_derivative 55 11.6 0.001 0.001 39.523 39.641 multiply_cannon_metrocomm3 69360 15.2 0.198 0.208 26.437 34.911 multiply_cannon_sync_h2d 69360 15.2 29.028 33.429 29.028 33.429 qs_ot_p2m_diag 40 11.0 0.021 0.043 33.139 33.236 cp_dbcsr_syevd 40 12.0 0.003 0.018 29.862 29.863 rebuild_ks_matrix 110 8.4 0.000 0.000 29.108 29.283 qs_ks_build_kohn_sham_matrix 110 9.4 0.013 0.053 29.108 29.282 init_scf_loop 7 6.6 0.001 0.029 27.199 27.202 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.853 27.011 cp_fm_syevd 40 13.0 0.001 0.024 24.560 24.731 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.089 23.285 apply_single 62 13.6 0.000 0.000 23.089 23.284 prepare_preconditioner 7 7.6 0.000 0.000 22.248 22.280 make_preconditioner 7 8.6 0.002 0.043 22.248 22.280 ot_new_cg_direction 55 11.6 0.001 0.001 21.102 21.103 cp_fm_redistribute_end 40 14.0 9.883 19.716 9.887 19.718 cp_fm_syevd_base 40 14.0 9.823 19.657 9.823 19.657 qs_rho_update_rho_low 110 7.6 0.000 0.001 18.291 18.769 calculate_rho_elec 110 8.6 0.071 0.722 18.291 18.769 init_scf_run 5 5.8 0.000 0.001 16.907 16.908 scf_env_initial_rho_setup 5 6.8 0.001 0.003 16.907 16.908 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.855 15.039 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.709 14.770 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.852 13.949 mp_sum_l 4764 12.2 12.850 13.676 12.850 13.676 density_rs2pw 110 9.6 0.004 0.060 11.549 12.103 calculate_first_density_matrix 1 7.0 0.043 0.917 11.075 11.936 calculate_dm_sparse 110 9.5 0.000 0.001 11.647 11.838 fft_wrap_pw1pw2 1425 12.5 0.019 0.026 10.970 11.288 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.969 11.046 dbcsr_mm_accdrv_process 154766 15.8 6.231 6.416 9.793 10.720 cp_fm_cholesky_invert 7 10.6 10.402 10.409 10.402 10.409 qs_vxc_create 110 10.4 0.002 0.003 10.091 10.115 fft_wrap_pw1pw2_240 915 14.0 0.199 0.238 9.627 9.833 check_diag 80 13.5 8.688 8.965 9.622 9.762 multiply_cannon_metrocomm1 69360 15.2 0.094 0.101 5.114 9.270 transfer_rs2pw 445 10.6 0.008 0.009 8.448 9.045 acc_transpose_blocks 69360 15.2 0.341 0.358 7.785 8.805 fft3d_pb 915 15.0 2.247 2.456 8.600 8.804 sum_up_and_integrate 60 10.3 0.002 0.029 8.726 8.740 integrate_v_rspace 60 11.3 0.002 0.010 8.707 8.723 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.011 8.155 8.174 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.732 7.786 xc_rho_set_and_dset_create 110 12.4 0.072 0.090 7.132 7.377 make_full_single_inverse 7 9.6 0.001 0.039 7.267 7.301 make_m2s 2890 13.2 0.079 0.088 6.507 7.114 make_images 2890 14.2 0.237 0.259 6.399 7.007 xc_vxc_pw_create 60 11.3 0.036 0.041 6.726 6.754 parallel_gemm_fm 29 8.8 0.000 0.000 6.324 6.330 parallel_gemm_fm_cosma 29 9.8 6.324 6.330 6.324 6.330 mp_alltoall_z22v 2340 16.7 5.798 6.007 5.798 6.007 xc_pw_derive 510 13.4 0.005 0.006 5.903 5.967 acc_transpose_blocks_kernels 69360 16.2 0.850 0.886 5.040 5.966 make_basis_sm 7 8.9 0.001 0.025 5.632 5.639 mp_waitany 7680 13.5 4.574 5.380 4.574 5.380 mp_alltoall_d11v 1300 13.8 4.281 5.243 4.281 5.243 jit_kernel_transpose 5 15.0 4.190 5.103 4.190 5.103 potential_pw2rs 60 12.3 0.002 0.006 4.924 4.964 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=237.621000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=558.400000, yerr=3.382307 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 7d0f9bf18b0e5cd9d2ab8bc0c692ee014390b318 Summary: empty Status: OK