=== 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: 9fa292db88c5773cf53ea201dab2b0cc1895d578 ################# 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/01 job id: 51881078 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/02 job id: 51881086 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/03 job id: 51881087 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/04 job id: 51881088 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/05 job id: 51881089 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/06 job id: 51881091 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/07 job id: 51881093 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/08 job id: 51881094 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/09 job id: 51881095 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/10 job id: 51881096 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/11 job id: 51881097 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/12 job id: 51881098 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/13 job id: 51881099 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/14 job id: 51881100 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/15 job id: 51881101 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/16 job id: 51881102 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/17 job id: 51881103 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/18 job id: 51881106 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/19 job id: 51881107 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/20 job id: 51881109 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/21 job id: 51881110 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/22 job id: 51881111 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/23 job id: 51881112 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/24 job id: 51881113 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/25 job id: 51881114 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/26 job id: 51881115 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/27 job id: 51881116 --- 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/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 0.021 0.033 140.451 140.452 farming_run 1 2.0 139.289 139.290 140.420 140.425 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.536807E+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 0.023 120.196 120.197 qs_energies 1 2.0 0.008 0.013 119.999 120.002 mp2_main 1 3.0 0.000 0.000 117.910 117.920 mp2_gpw_main 1 4.0 0.029 0.038 116.833 116.836 mp2_ri_gpw_compute_in 1 5.0 0.183 0.189 98.024 98.152 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.004 57.346 57.473 mp2_eri_3c_integrate_gpw 272 7.0 0.149 0.159 43.793 48.629 get_2c_integrals 1 6.0 0.008 0.010 39.791 40.490 integrate_v_rspace 273 8.0 0.439 0.454 25.509 29.754 fft_wrap_pw1pw2 5465 10.4 0.068 0.073 28.043 28.797 fft_wrap_pw1pw2_100 2178 11.4 1.344 1.440 25.480 26.259 grid_integrate_task_list 273 9.0 20.747 25.917 20.747 25.917 compute_2c_integrals 1 7.0 0.002 0.002 21.603 21.614 compute_2c_integrals_loop_lm 1 8.0 0.002 0.004 20.829 21.348 mp2_eri_2c_integrate_gpw 1 9.0 2.394 2.446 20.827 21.345 cp_fm_cholesky_decompose 12 8.2 18.124 18.844 18.124 18.844 rpa_ri_compute_en 1 5.0 0.020 0.034 18.704 18.832 cholesky_decomp 1 7.0 0.000 0.000 17.028 17.747 fft3d_s 5443 12.4 16.791 17.233 16.812 17.254 ao_to_mo_and_store_B_mult_1 272 7.0 10.767 15.304 10.767 15.304 calculate_wavefunction 272 8.0 5.470 5.539 12.958 13.920 calc_potential_gpw 544 9.5 0.004 0.005 10.935 11.709 rpa_num_int 1 6.0 0.001 0.014 10.705 10.714 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.654 10.683 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 9.806 10.674 potential_pw2rs 545 10.0 0.106 0.108 8.417 10.268 calc_mat_Q 8 8.0 0.000 0.000 9.515 9.621 contract_S_to_Q 8 9.0 0.000 0.000 8.940 9.044 parallel_gemm_fm 14 9.1 0.000 0.000 8.519 8.643 parallel_gemm_fm_cosma 14 10.1 8.519 8.643 8.519 8.643 collocate_single_gaussian 272 10.0 0.040 0.043 7.811 8.603 create_integ_mat 1 6.0 0.018 0.032 7.768 7.778 array2fm 1 7.0 0.000 0.000 6.758 7.226 pw_poisson_solve 545 10.5 0.009 0.011 4.664 4.939 pw_gather_s 2722 12.2 3.887 4.779 3.887 4.779 pw_scatter_s 2720 12.7 4.516 4.767 4.516 4.767 pw_poisson_set 548 11.5 0.017 0.018 3.353 3.558 array2fm_buffer_send 1 8.0 2.979 3.146 2.979 3.146 pw_copy 4911 11.6 2.620 2.732 2.620 2.732 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=116.833006, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2798.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 0.027 0.036 529.871 529.873 farming_run 1 2.0 528.357 528.406 529.825 529.831 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.271652E+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 19728442. 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 0.033 324.748 324.749 qs_energies 1 2.0 0.004 0.017 324.514 324.515 mp2_main 1 3.0 0.002 0.018 209.486 209.490 mp2_gpw_main 1 4.0 0.058 0.128 208.623 208.630 mp2_ri_gpw_compute_en 1 5.0 0.037 0.041 128.478 130.568 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.728 1.834 126.835 126.842 scf_env_do_scf 1 3.0 0.000 0.000 114.737 114.738 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.834 113.842 rebuild_ks_matrix 4 6.0 0.000 0.000 113.832 113.841 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.062 113.832 113.841 hfx_ks_matrix 4 8.0 0.001 0.001 113.487 113.492 integrate_four_center 4 9.0 0.152 0.466 113.487 113.491 mp2_ri_gpw_compute_en_expansio 172 7.0 0.463 0.495 110.964 112.116 local_gemm 172 8.0 110.500 111.650 110.500 111.650 integrate_four_center_main 4 10.0 0.133 0.545 102.120 105.432 integrate_four_center_bin 267 11.0 101.987 104.896 101.987 104.896 init_scf_loop 1 4.0 0.000 0.000 96.731 96.731 mp2_ri_gpw_compute_in 1 5.0 0.068 0.089 79.945 80.968 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.012 56.710 57.732 mp2_eri_3c_integrate_gpw 91 7.0 0.146 0.162 44.361 49.376 integrate_v_rspace 95 8.0 0.397 0.566 27.990 32.825 fft_wrap_pw1pw2 1868 10.4 0.029 0.034 30.497 31.005 fft_wrap_pw1pw2_100 730 11.4 0.628 0.695 28.205 28.659 grid_integrate_task_list 95 9.0 23.296 28.353 23.296 28.353 ao_to_mo_and_store_B_mult_1 91 7.0 10.677 27.522 10.677 27.522 get_2c_integrals 1 6.0 0.001 0.009 23.136 23.172 compute_2c_integrals 1 7.0 0.004 0.012 22.120 22.128 compute_2c_integrals_loop_lm 1 8.0 0.002 0.009 21.724 21.977 mp2_eri_2c_integrate_gpw 1 9.0 1.730 1.876 21.722 21.975 fft3d_s 1823 12.4 19.224 19.910 19.238 19.923 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.004 18.004 calc_potential_gpw 182 9.5 0.002 0.002 12.366 13.127 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.013 10.713 11.386 calculate_wavefunction 91 8.0 2.028 2.071 10.048 10.898 potential_pw2rs 186 10.0 0.033 0.034 9.034 10.421 mp2_ri_gpw_compute_en_comm 22 7.0 0.504 0.522 8.028 9.060 collocate_single_gaussian 91 10.0 0.018 0.031 8.286 8.905 mp_sync 37 10.5 3.359 8.387 3.359 8.387 mp_sendrecv_dm3 2068 8.0 6.053 7.105 6.053 7.105 integrate_four_center_load 4 10.0 0.000 0.000 6.751 6.755 hfx_load_balance 1 11.0 0.000 0.000 6.751 6.755 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=208.620262, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1525.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 447.352832E+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 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 0.026 0.083 58.411 58.412 qs_mol_dyn_low 1 2.0 0.008 0.047 57.719 57.727 qs_forces 11 3.9 0.023 0.034 57.625 57.630 qs_energies 11 4.9 0.007 0.020 56.023 56.058 scf_env_do_scf 11 5.9 0.000 0.001 49.588 49.588 scf_env_do_scf_inner_loop 108 6.5 0.003 0.009 47.335 47.335 qs_scf_new_mos 108 7.5 0.000 0.001 36.776 37.093 qs_scf_loop_do_ot 108 8.5 0.000 0.001 36.775 37.093 dbcsr_multiply_generic 2286 12.5 0.100 0.103 35.818 36.354 ot_scf_mini 108 9.5 0.002 0.002 34.984 35.191 multiply_cannon 2286 13.5 0.177 0.183 27.128 28.985 velocity_verlet 10 3.0 0.002 0.005 28.292 28.293 multiply_cannon_loop 2286 14.5 1.805 1.907 26.382 28.266 ot_mini 108 10.5 0.001 0.001 21.348 21.588 qs_ot_get_derivative 108 11.5 0.001 0.001 18.074 18.263 mp_waitall_1 245248 16.5 9.676 15.829 9.676 15.829 multiply_cannon_metrocomm3 54864 15.5 0.072 0.078 6.671 13.435 multiply_cannon_multrec 54864 15.5 3.573 5.643 7.509 10.830 qs_ot_get_p 119 10.4 0.001 0.001 8.749 9.015 mp_sum_l 7287 12.8 6.492 8.343 6.492 8.343 rebuild_ks_matrix 119 8.3 0.000 0.000 8.166 8.321 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.022 8.165 8.320 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.196 7.345 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.335 6.464 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.913 6.369 multiply_cannon_sync_h2d 54864 15.5 5.086 6.127 5.086 6.127 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.537 5.574 dbcsr_mm_accdrv_process 76910 16.1 1.837 2.898 3.851 5.518 init_scf_run 11 5.9 0.000 0.001 4.944 4.944 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.944 4.944 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.721 4.721 sum_up_and_integrate 119 10.3 0.001 0.002 4.642 4.657 integrate_v_rspace 119 11.3 0.002 0.003 4.631 4.647 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.426 4.572 calculate_rho_elec 119 8.7 0.011 0.016 4.426 4.571 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.474 4.475 cp_fm_redistribute_end 50 14.0 2.278 4.440 2.288 4.445 cp_fm_diag_elpa_base 50 14.0 2.149 4.322 2.154 4.331 multiply_cannon_metrocomm1 54864 15.5 0.056 0.062 2.202 3.489 calculate_dm_sparse 119 9.5 0.000 0.000 3.050 3.216 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.009 3.207 apply_single 119 13.6 0.000 0.000 3.009 3.207 ot_diis_step 108 11.5 0.006 0.006 2.887 2.887 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.668 2.725 acc_transpose_blocks 54864 15.5 0.219 0.238 2.197 2.718 calculate_first_density_matrix 1 7.0 0.000 0.001 2.572 2.580 jit_kernel_multiply 13 15.8 1.951 2.565 1.951 2.565 density_rs2pw 119 9.7 0.003 0.004 2.349 2.427 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.003 2.306 2.307 wfi_extrapolate 11 7.9 0.001 0.001 2.291 2.292 mp_sum_d 4139 12.0 1.667 2.232 1.667 2.232 init_scf_loop 11 6.9 0.000 0.000 2.229 2.229 grid_integrate_task_list 119 12.3 2.009 2.111 2.009 2.111 potential_pw2rs 119 12.3 0.003 0.004 1.990 2.015 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.955 2.015 make_m2s 4572 13.5 0.053 0.055 1.890 1.942 fft_wrap_pw1pw2 1201 11.6 0.011 0.013 1.798 1.881 make_images 4572 14.5 0.132 0.137 1.808 1.860 mp_alltoall_d11v 2130 13.8 1.507 1.696 1.507 1.696 transfer_rs2pw 487 10.6 0.006 0.007 1.537 1.676 fft3d_ps 1201 13.6 0.368 0.469 1.574 1.652 mp_waitany 12084 13.8 1.307 1.515 1.307 1.515 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.434 1.472 transfer_pw2rs 487 13.2 0.005 0.006 1.450 1.470 fft_wrap_pw1pw2_140 487 12.2 0.048 0.053 1.370 1.454 acc_transpose_blocks_sync 164592 16.5 1.199 1.434 1.199 1.434 grid_collocate_task_list 119 9.7 1.350 1.419 1.350 1.419 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.953 1.360 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=58.412000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=426.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 484.372480E+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 1819394. 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 0.038 0.135 39.695 39.696 qs_mol_dyn_low 1 2.0 0.011 0.037 39.191 39.198 qs_forces 11 3.9 0.015 0.025 39.015 39.022 qs_energies 11 4.9 0.002 0.002 37.327 37.347 scf_env_do_scf 11 5.9 0.000 0.001 32.111 32.111 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.662 29.663 dbcsr_multiply_generic 2286 12.5 0.101 0.104 22.139 22.487 qs_scf_new_mos 108 7.5 0.001 0.001 21.015 21.236 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.014 21.236 ot_scf_mini 108 9.5 0.002 0.003 20.075 20.233 velocity_verlet 10 3.0 0.027 0.103 18.593 18.598 multiply_cannon 2286 13.5 0.207 0.216 16.893 18.464 multiply_cannon_loop 2286 14.5 1.181 1.240 15.680 17.132 ot_mini 108 10.5 0.001 0.001 12.285 12.508 mp_waitall_1 200699 16.5 5.611 10.752 5.611 10.752 qs_ot_get_derivative 108 11.5 0.001 0.001 9.895 10.054 multiply_cannon_metrocomm3 27432 15.5 0.072 0.076 4.047 9.462 multiply_cannon_multrec 27432 15.5 1.816 4.279 6.124 8.965 rebuild_ks_matrix 119 8.3 0.000 0.000 6.799 6.945 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.798 6.945 dbcsr_mm_accdrv_process 47894 16.0 3.448 6.042 4.230 6.306 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.986 6.117 qs_ot_get_p 119 10.4 0.001 0.002 4.853 5.063 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.742 4.592 mp_sum_l 7287 12.8 2.103 4.137 2.103 4.137 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.012 4.063 apply_single 119 13.6 0.000 0.000 3.012 4.063 init_scf_run 11 5.9 0.000 0.001 4.001 4.001 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.001 4.001 sum_up_and_integrate 119 10.3 0.001 0.002 3.681 3.686 integrate_v_rspace 119 11.3 0.002 0.003 3.667 3.672 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.604 3.637 calculate_rho_elec 119 8.7 0.021 0.024 3.603 3.637 qs_ot_p2m_diag 50 11.0 0.008 0.012 3.156 3.174 make_m2s 4572 13.5 0.051 0.053 2.716 3.051 make_images 4572 14.5 0.205 0.244 2.627 2.964 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.754 2.755 calculate_first_density_matrix 1 7.0 0.000 0.001 2.468 2.470 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.382 2.466 init_scf_loop 11 6.9 0.000 0.000 2.427 2.427 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.394 2.394 cp_fm_redistribute_end 50 14.0 1.210 2.357 1.214 2.361 ot_diis_step 108 11.5 0.011 0.011 2.339 2.340 cp_fm_diag_elpa_base 50 14.0 1.113 2.260 1.142 2.302 multiply_cannon_sync_h2d 27432 15.5 1.687 2.211 1.687 2.211 calculate_dm_sparse 119 9.5 0.000 0.001 2.078 2.161 jit_kernel_multiply 11 16.2 0.722 2.070 0.722 2.070 density_rs2pw 119 9.7 0.003 0.004 1.899 1.984 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.964 1.966 grid_integrate_task_list 119 12.3 1.843 1.958 1.843 1.958 acc_transpose_blocks 27432 15.5 0.113 0.120 1.605 1.939 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.680 1.721 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.678 1.712 make_images_data 4572 15.5 0.047 0.054 1.263 1.650 prepare_preconditioner 11 7.9 0.000 0.000 1.541 1.567 make_preconditioner 11 8.9 0.000 0.000 1.541 1.567 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.442 1.498 hybrid_alltoall_any 4725 16.4 0.054 0.115 1.118 1.491 wfi_extrapolate 11 7.9 0.001 0.001 1.474 1.474 potential_pw2rs 119 12.3 0.006 0.006 1.422 1.430 fft3d_ps 1201 13.6 0.511 0.563 1.397 1.428 mp_alltoall_d11v 2130 13.8 1.227 1.361 1.227 1.361 fft_wrap_pw1pw2_140 487 12.2 0.047 0.051 1.315 1.347 grid_collocate_task_list 119 9.7 1.293 1.343 1.293 1.343 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.217 1.263 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.249 1.259 transfer_rs2pw 487 10.6 0.005 0.006 1.157 1.251 mp_allgather_i34 2286 14.5 0.639 1.041 0.639 1.041 mp_sum_d 4139 12.0 0.571 1.035 0.571 1.035 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.950 0.969 transfer_pw2rs 487 13.2 0.004 0.005 0.936 0.942 acc_transpose_blocks_sync 82296 16.5 0.812 0.935 0.812 0.935 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.869 0.884 acc_transpose_blocks_kernels 27432 16.5 0.188 0.277 0.655 0.862 make_images_sizes 4572 15.5 0.005 0.005 0.585 0.840 mp_alltoall_i44 4572 16.5 0.580 0.835 0.580 0.835 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.696000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=461.000000, yerr=1.954017 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 516.038656E+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 0.011 0.032 32.174 32.175 qs_mol_dyn_low 1 2.0 0.003 0.003 31.834 31.841 qs_forces 11 3.9 0.017 0.038 31.743 31.746 qs_energies 11 4.9 0.007 0.009 30.200 30.218 scf_env_do_scf 11 5.9 0.000 0.001 25.509 25.509 scf_env_do_scf_inner_loop 108 6.5 0.006 0.019 22.982 22.982 dbcsr_multiply_generic 2286 12.5 0.095 0.097 16.963 17.046 qs_scf_new_mos 108 7.5 0.001 0.001 15.433 15.454 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.432 15.453 velocity_verlet 10 3.0 0.026 0.039 15.240 15.241 ot_scf_mini 108 9.5 0.002 0.002 14.665 14.686 multiply_cannon 2286 13.5 0.190 0.196 13.711 14.455 multiply_cannon_loop 2286 14.5 0.859 0.895 12.947 13.714 ot_mini 108 10.5 0.001 0.001 9.073 9.096 qs_ot_get_derivative 108 11.5 0.001 0.001 7.571 7.594 multiply_cannon_multrec 18288 15.5 1.866 2.899 6.966 7.307 rebuild_ks_matrix 119 8.3 0.000 0.000 5.905 5.924 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.905 5.924 dbcsr_mm_accdrv_process 38222 16.0 4.944 5.733 5.009 5.800 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.199 5.216 mp_waitall_1 158411 16.6 2.678 3.514 2.678 3.514 init_scf_run 11 5.9 0.001 0.010 3.502 3.502 scf_env_initial_rho_setup 11 6.9 0.002 0.010 3.501 3.502 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.889 3.467 sum_up_and_integrate 119 10.3 0.001 0.001 3.445 3.452 integrate_v_rspace 119 11.3 0.003 0.003 3.433 3.441 qs_ot_get_p 119 10.4 0.002 0.004 3.344 3.363 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.171 3.178 calculate_rho_elec 119 8.7 0.031 0.031 3.170 3.177 init_scf_loop 11 6.9 0.000 0.000 2.511 2.512 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.986 2.364 apply_single 119 13.6 0.000 0.000 1.986 2.364 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.495 2.309 calculate_first_density_matrix 1 7.0 0.000 0.000 2.294 2.296 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.230 2.238 make_m2s 4572 13.5 0.043 0.044 1.827 1.953 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.936 1.937 grid_integrate_task_list 119 12.3 1.806 1.911 1.806 1.911 make_images 4572 14.5 0.191 0.203 1.742 1.867 calculate_dm_sparse 119 9.5 0.000 0.001 1.852 1.862 density_rs2pw 119 9.7 0.003 0.003 1.738 1.839 prepare_preconditioner 11 7.9 0.000 0.000 1.763 1.767 make_preconditioner 11 8.9 0.000 0.000 1.763 1.767 acc_transpose_blocks 18288 15.5 0.078 0.080 1.639 1.718 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.706 1.715 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.618 1.700 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.697 1.698 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.678 1.690 cp_fm_diag_elpa_base 50 14.0 1.655 1.669 1.677 1.688 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.634 1.642 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.548 1.555 ot_diis_step 108 11.5 0.011 0.011 1.484 1.484 mp_sum_l 7287 12.8 1.030 1.433 1.030 1.433 fft3d_ps 1201 13.6 0.534 0.552 1.319 1.330 grid_collocate_task_list 119 9.7 1.234 1.317 1.234 1.317 fft_wrap_pw1pw2_140 487 12.2 0.060 0.063 1.290 1.299 potential_pw2rs 119 12.3 0.007 0.007 1.244 1.247 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.180 1.184 multiply_cannon_sync_h2d 18288 15.5 1.042 1.172 1.042 1.172 wfi_extrapolate 11 7.9 0.001 0.001 1.156 1.156 transfer_rs2pw 487 10.6 0.005 0.005 0.965 1.101 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.959 0.975 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 0.950 0.964 make_images_data 4572 15.5 0.047 0.051 0.804 0.954 hybrid_alltoall_any 4725 16.4 0.059 0.117 0.704 0.892 mp_alltoall_d11v 2130 13.8 0.672 0.869 0.672 0.869 acc_transpose_blocks_kernels 18288 16.5 0.216 0.224 0.796 0.812 acc_transpose_blocks_sync 54864 16.5 0.744 0.808 0.744 0.808 transfer_pw2rs 487 13.2 0.004 0.004 0.796 0.798 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.777 0.780 cp_fm_cholesky_invert 11 10.9 0.724 0.728 0.724 0.728 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.671 0.721 mp_alltoall_z22v 1201 15.6 0.626 0.698 0.626 0.698 mp_waitany 9880 13.7 0.530 0.667 0.530 0.667 transfer_rs2pw_140 130 11.5 0.115 0.119 0.522 0.657 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.175000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=490.818182, yerr=1.991718 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 555.544576E+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 0.048 0.059 36.274 36.275 qs_mol_dyn_low 1 2.0 0.003 0.003 35.912 35.920 qs_forces 11 3.9 0.045 0.051 35.853 35.853 qs_energies 11 4.9 0.001 0.001 34.101 34.143 scf_env_do_scf 11 5.9 0.000 0.001 29.165 29.166 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.773 25.773 dbcsr_multiply_generic 2286 12.5 0.101 0.104 19.806 19.947 velocity_verlet 10 3.0 0.001 0.002 18.287 18.289 qs_scf_new_mos 108 7.5 0.001 0.001 17.869 17.920 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.868 17.920 ot_scf_mini 108 9.5 0.002 0.003 16.836 16.883 multiply_cannon 2286 13.5 0.215 0.224 16.022 16.478 multiply_cannon_loop 2286 14.5 1.514 1.569 15.078 15.361 ot_mini 108 10.5 0.001 0.001 10.308 10.357 multiply_cannon_multrec 27432 15.5 2.449 3.116 9.003 9.301 qs_ot_get_derivative 108 11.5 0.001 0.001 8.421 8.470 dbcsr_mm_accdrv_process 47916 15.9 5.939 7.660 6.454 7.859 rebuild_ks_matrix 119 8.3 0.000 0.000 6.294 6.334 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.293 6.334 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.569 5.609 qs_ot_get_p 119 10.4 0.002 0.004 3.670 3.739 init_scf_run 11 5.9 0.000 0.001 3.595 3.596 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.595 3.595 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.048 3.528 sum_up_and_integrate 119 10.3 0.001 0.001 3.414 3.423 integrate_v_rspace 119 11.3 0.003 0.003 3.403 3.411 init_scf_loop 11 6.9 0.000 0.000 3.373 3.374 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.223 3.257 calculate_rho_elec 119 8.7 0.040 0.046 3.222 3.257 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.205 2.654 apply_single 119 13.6 0.000 0.000 2.205 2.653 prepare_preconditioner 11 7.9 0.000 0.000 2.533 2.541 make_preconditioner 11 8.9 0.000 0.000 2.533 2.541 acc_transpose_blocks 27432 15.5 0.116 0.119 2.401 2.484 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.142 2.469 make_m2s 4572 13.5 0.053 0.055 2.289 2.388 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.324 2.351 make_images 4572 14.5 0.272 0.335 2.181 2.278 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.253 2.263 calculate_first_density_matrix 1 7.0 0.000 0.000 2.219 2.222 calculate_dm_sparse 119 9.5 0.000 0.000 2.140 2.199 mp_waitall_1 137007 16.6 1.709 2.185 1.709 2.185 grid_integrate_task_list 119 12.3 1.830 1.959 1.830 1.959 ot_diis_step 108 11.5 0.012 0.012 1.848 1.848 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.846 1.847 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.787 1.788 density_rs2pw 119 9.7 0.003 0.003 1.686 1.783 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.704 1.733 mp_sum_l 7287 12.8 1.064 1.694 1.064 1.694 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.619 1.628 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.607 1.617 cp_fm_diag_elpa_base 50 14.0 1.574 1.589 1.605 1.615 acc_transpose_blocks_sync 82296 16.5 1.428 1.517 1.428 1.517 fft_wrap_pw1pw2_140 487 12.2 0.071 0.073 1.406 1.442 fft3d_ps 1201 13.6 0.576 0.638 1.365 1.387 grid_collocate_task_list 119 9.7 1.248 1.355 1.248 1.355 wfi_extrapolate 11 7.9 0.001 0.001 1.324 1.324 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.289 1.306 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.236 1.247 potential_pw2rs 119 12.3 0.008 0.009 1.219 1.222 multiply_cannon_metrocomm3 27432 15.5 0.041 0.043 0.711 1.195 cp_fm_upper_to_full 72 14.2 0.823 1.167 0.823 1.167 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 1.083 1.125 jit_kernel_multiply 6 16.1 0.448 1.092 0.448 1.092 dbcsr_complete_redistribute 329 12.2 0.116 0.146 0.744 1.026 transfer_rs2pw 487 10.6 0.005 0.005 0.907 1.004 make_images_data 4572 15.5 0.047 0.051 0.854 0.964 hybrid_alltoall_any 4725 16.4 0.067 0.158 0.734 0.901 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.875 mp_alltoall_d11v 2130 13.8 0.716 0.856 0.716 0.856 acc_transpose_blocks_kernels 27432 16.5 0.271 0.280 0.827 0.850 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.558 0.834 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.812 0.819 cp_fm_cholesky_invert 11 10.9 0.744 0.747 0.744 0.747 transfer_pw2rs 487 13.2 0.004 0.004 0.725 0.728 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.275000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.363636, yerr=2.226809 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 612.302848E+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 0.009 0.025 29.613 29.613 qs_mol_dyn_low 1 2.0 0.003 0.003 29.236 29.246 qs_forces 11 3.9 0.021 0.022 29.150 29.151 qs_energies 11 4.9 0.001 0.001 27.433 27.435 scf_env_do_scf 11 5.9 0.000 0.001 22.005 22.005 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.492 19.493 velocity_verlet 10 3.0 0.002 0.002 14.641 14.646 dbcsr_multiply_generic 2286 12.5 0.094 0.099 13.320 13.447 qs_scf_new_mos 108 7.5 0.001 0.001 12.024 12.060 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.024 12.060 ot_scf_mini 108 9.5 0.002 0.002 11.311 11.340 multiply_cannon 2286 13.5 0.223 0.230 10.692 11.225 multiply_cannon_loop 2286 14.5 0.642 0.656 9.758 9.965 ot_mini 108 10.5 0.001 0.001 6.608 6.640 multiply_cannon_multrec 9144 15.5 1.743 2.090 6.076 6.415 rebuild_ks_matrix 119 8.3 0.000 0.000 5.724 5.742 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.723 5.742 qs_ot_get_derivative 108 11.5 0.001 0.001 5.289 5.319 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.097 5.114 dbcsr_mm_accdrv_process 12550 15.8 3.592 4.278 4.225 4.321 init_scf_run 11 5.9 0.006 0.100 3.985 3.985 scf_env_initial_rho_setup 11 6.9 0.040 0.099 3.979 3.985 sum_up_and_integrate 119 10.3 0.001 0.001 3.325 3.330 integrate_v_rspace 119 11.3 0.003 0.003 3.315 3.320 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.211 3.216 calculate_rho_elec 119 8.7 0.060 0.061 3.211 3.216 calculate_first_density_matrix 1 7.0 0.000 0.000 2.808 2.854 qs_ot_get_p 119 10.4 0.001 0.001 2.806 2.842 init_scf_loop 11 6.9 0.000 0.000 2.493 2.494 mp_waitall_1 115863 16.7 1.499 2.038 1.499 2.038 make_m2s 4572 13.5 0.033 0.035 1.836 1.998 grid_integrate_task_list 119 12.3 1.836 1.945 1.836 1.945 make_images 4572 14.5 0.269 0.302 1.746 1.906 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.814 1.816 calculate_dm_sparse 119 9.5 0.000 0.000 1.774 1.794 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.752 1.774 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.747 1.756 prepare_preconditioner 11 7.9 0.000 0.000 1.737 1.740 make_preconditioner 11 8.9 0.000 0.000 1.737 1.740 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.626 1.650 density_rs2pw 119 9.7 0.003 0.003 1.582 1.645 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.572 1.573 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.571 1.572 acc_transpose_blocks 9144 15.5 0.041 0.043 1.442 1.462 fft_wrap_pw1pw2_140 487 12.2 0.094 0.095 1.449 1.460 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.409 1.417 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.371 1.387 fft3d_ps 1201 13.6 0.635 0.652 1.359 1.369 grid_collocate_task_list 119 9.7 1.296 1.357 1.296 1.357 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.315 1.323 cp_fm_diag_elpa_base 50 14.0 1.288 1.306 1.314 1.322 ot_diis_step 108 11.5 0.013 0.013 1.305 1.306 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.244 1.267 apply_single 119 13.6 0.000 0.000 1.244 1.267 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.216 1.216 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.159 1.164 potential_pw2rs 119 12.3 0.010 0.010 1.132 1.135 jit_kernel_multiply 6 15.3 0.594 1.083 0.594 1.083 wfi_extrapolate 11 7.9 0.001 0.001 1.081 1.081 hybrid_alltoall_any 4725 16.4 0.066 0.177 0.761 0.983 mp_alltoall_d11v 2130 13.8 0.818 0.965 0.818 0.965 make_images_data 4572 15.5 0.042 0.046 0.797 0.965 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.878 0.928 parallel_gemm_fm 81 9.0 0.000 0.000 0.899 0.900 parallel_gemm_fm_cosma 81 10.0 0.899 0.900 0.899 0.900 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.428 0.891 cp_fm_cholesky_invert 11 10.9 0.877 0.879 0.877 0.879 make_basis_sm 11 9.8 0.000 0.000 0.873 0.874 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.821 0.826 transfer_rs2pw 487 10.6 0.004 0.005 0.763 0.819 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.754 0.755 acc_transpose_blocks_sync 27432 16.5 0.717 0.734 0.717 0.734 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.664 0.712 mp_allgather_i34 2286 14.5 0.242 0.682 0.242 0.682 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.668 0.673 transfer_pw2rs 487 13.2 0.003 0.004 0.641 0.642 mp_alltoall_z22v 1201 15.6 0.600 0.637 0.600 0.637 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.613000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=579.818182, yerr=5.424158 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 782.344192E+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 0.069 0.081 42.868 42.869 qs_mol_dyn_low 1 2.0 0.003 0.003 42.368 42.376 qs_forces 11 3.9 0.002 0.003 42.301 42.302 qs_energies 11 4.9 0.015 0.029 40.321 40.323 scf_env_do_scf 11 5.9 0.000 0.001 34.299 34.299 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.538 26.539 velocity_verlet 10 3.0 0.002 0.002 23.801 23.807 dbcsr_multiply_generic 2286 12.5 0.102 0.103 19.067 19.342 qs_scf_new_mos 108 7.5 0.001 0.001 17.348 17.446 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.347 17.445 ot_scf_mini 108 9.5 0.002 0.002 16.175 16.272 multiply_cannon 2286 13.5 0.298 0.307 15.083 16.151 multiply_cannon_loop 2286 14.5 0.859 0.878 13.785 14.892 ot_mini 108 10.5 0.001 0.001 9.822 9.937 multiply_cannon_multrec 9144 15.5 3.377 4.642 8.812 8.948 qs_ot_get_derivative 108 11.5 0.001 0.001 7.723 7.818 init_scf_loop 11 6.9 0.000 0.000 7.735 7.738 rebuild_ks_matrix 119 8.3 0.000 0.000 6.978 7.118 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.978 7.118 prepare_preconditioner 11 7.9 0.000 0.000 6.779 6.792 make_preconditioner 11 8.9 0.000 0.000 6.779 6.792 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.347 6.670 dbcsr_mm_accdrv_process 12550 15.8 4.747 6.571 5.304 6.604 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.295 6.422 cp_fm_upper_to_full 72 14.2 3.183 4.558 3.183 4.558 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.898 3.901 calculate_rho_elec 119 8.7 0.118 0.122 3.897 3.901 init_scf_run 11 5.9 0.000 0.001 3.834 3.834 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.833 3.833 sum_up_and_integrate 119 10.3 0.001 0.001 3.670 3.675 integrate_v_rspace 119 11.3 0.003 0.003 3.659 3.665 qs_ot_get_p 119 10.4 0.002 0.002 3.399 3.544 mp_waitall_1 94719 16.7 2.268 3.307 2.268 3.307 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.807 3.230 dbcsr_complete_redistribute 329 12.2 0.284 0.287 1.961 2.764 make_m2s 4572 13.5 0.037 0.037 2.384 2.551 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.252 2.540 apply_single 119 13.6 0.000 0.000 2.252 2.540 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.661 2.463 make_images 4572 14.5 0.351 0.383 2.264 2.432 calculate_first_density_matrix 1 7.0 0.000 0.000 2.400 2.403 fft_wrap_pw1pw2 1201 11.6 0.014 0.014 2.388 2.392 calculate_dm_sparse 119 9.5 0.000 0.000 2.246 2.258 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.303 2.222 mp_alltoall_i22 627 13.8 1.404 2.222 1.404 2.222 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.426 2.219 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.063 2.115 grid_integrate_task_list 119 12.3 2.070 2.090 2.070 2.090 ot_diis_step 108 11.5 0.014 0.015 2.066 2.067 fft_wrap_pw1pw2_140 487 12.2 0.179 0.180 2.033 2.038 qs_ot_p2m_diag 50 11.0 0.042 0.043 2.002 2.004 density_rs2pw 119 9.7 0.003 0.003 1.930 1.944 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.886 1.887 acc_transpose_blocks 9144 15.5 0.044 0.044 1.837 1.873 qs_energies_init_hamiltonians 11 5.9 0.032 0.049 1.861 1.863 fft3d_ps 1201 13.6 0.858 0.878 1.840 1.844 mp_sum_l 7287 12.8 1.056 1.793 1.056 1.793 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.635 1.676 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.652 1.653 grid_collocate_task_list 119 9.7 1.514 1.529 1.514 1.529 cp_fm_cholesky_invert 11 10.9 1.423 1.426 1.423 1.426 wfi_extrapolate 11 7.9 0.001 0.001 1.373 1.373 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.366 1.366 cp_fm_diag_elpa_base 50 14.0 1.221 1.272 1.364 1.364 hybrid_alltoall_any 4725 16.4 0.091 0.151 1.086 1.350 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.311 1.336 make_images_data 4572 15.5 0.045 0.048 1.031 1.256 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.234 1.240 potential_pw2rs 119 12.3 0.013 0.014 1.191 1.192 mp_alltoall_d11v 2130 13.8 1.157 1.178 1.157 1.178 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.152 1.171 acc_transpose_blocks_sync 27432 16.5 1.094 1.127 1.094 1.127 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.998 1.055 jit_kernel_multiply 6 15.7 0.528 1.027 0.528 1.027 qs_create_task_list 11 7.9 0.001 0.002 0.929 0.941 generate_qs_task_list 11 8.9 0.368 0.389 0.928 0.940 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.894 0.907 mp_alltoall_z22v 1201 15.6 0.846 0.876 0.846 0.876 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.869000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=736.181818, yerr=10.928014 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 499.486720E+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 1383689. 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 0.031 0.077 88.379 88.380 qs_mol_dyn_low 1 2.0 0.014 0.050 87.774 87.783 qs_forces 11 3.9 0.003 0.003 87.548 87.554 qs_energies 11 4.9 0.003 0.008 84.574 84.601 scf_env_do_scf 11 5.9 0.000 0.001 74.987 74.990 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 68.620 68.620 dbcsr_multiply_generic 2055 12.4 0.108 0.112 53.076 53.359 qs_scf_new_mos 99 7.5 0.000 0.001 50.313 50.468 qs_scf_loop_do_ot 99 8.5 0.000 0.001 50.312 50.467 ot_scf_mini 99 9.5 0.002 0.002 47.827 47.927 velocity_verlet 10 3.0 0.025 0.091 46.424 46.431 multiply_cannon 2055 13.4 0.177 0.183 43.372 44.368 multiply_cannon_loop 2055 14.4 1.781 1.822 42.126 43.055 ot_mini 99 10.5 0.001 0.001 28.006 28.103 qs_ot_get_derivative 99 11.5 0.001 0.001 21.066 21.162 multiply_cannon_multrec 49320 15.4 11.267 11.865 17.289 18.109 rebuild_ks_matrix 110 8.3 0.000 0.000 15.300 15.420 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.011 15.300 15.420 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.468 13.567 mp_waitall_1 220248 16.4 12.235 13.067 12.235 13.067 qs_ot_get_p 110 10.4 0.001 0.001 10.927 11.027 multiply_cannon_sync_h2d 49320 15.4 9.487 10.014 9.487 10.014 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.996 8.558 multiply_cannon_metrocomm3 49320 15.4 0.083 0.086 6.887 7.900 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.322 7.849 apply_single 110 13.6 0.000 0.001 7.322 7.848 sum_up_and_integrate 110 10.3 0.001 0.003 7.515 7.530 integrate_v_rspace 110 11.3 0.003 0.003 7.490 7.511 qs_ot_p2m_diag 48 11.0 0.012 0.018 7.374 7.402 init_scf_run 11 5.9 0.000 0.002 7.324 7.324 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.324 7.324 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.952 7.114 calculate_rho_elec 110 8.6 0.021 0.025 6.952 7.114 ot_diis_step 99 11.5 0.006 0.006 6.670 6.670 cp_dbcsr_syevd 48 12.0 0.002 0.003 6.447 6.447 init_scf_loop 11 6.9 0.000 0.000 6.336 6.337 dbcsr_mm_accdrv_process 87628 16.1 3.087 3.168 5.894 6.161 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.835 5.894 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.758 5.781 cp_fm_diag_elpa_base 48 14.0 5.741 5.765 5.756 5.779 mp_sum_l 6594 12.7 4.470 5.417 4.470 5.417 make_m2s 4110 13.4 0.061 0.065 4.795 4.959 make_images 4110 14.4 0.178 0.192 4.697 4.865 multiply_cannon_metrocomm1 49320 15.4 0.064 0.068 2.803 4.468 wfi_extrapolate 11 7.9 0.001 0.001 4.293 4.293 prepare_preconditioner 11 7.9 0.000 0.000 4.107 4.128 make_preconditioner 11 8.9 0.000 0.000 4.107 4.128 density_rs2pw 110 9.6 0.004 0.004 3.917 4.109 calculate_dm_sparse 110 9.5 0.001 0.001 3.869 4.012 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.817 3.868 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 3.750 3.754 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.533 3.606 grid_integrate_task_list 110 12.3 3.236 3.397 3.236 3.397 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.298 3.347 fft_wrap_pw1pw2 1111 11.6 0.014 0.017 3.208 3.319 potential_pw2rs 110 12.3 0.005 0.006 3.026 3.070 calculate_first_density_matrix 1 7.0 0.000 0.001 2.918 2.931 fft3d_ps 1111 13.6 0.732 0.833 2.750 2.862 mp_alltoall_d11v 2046 13.8 2.473 2.851 2.473 2.851 fft_wrap_pw1pw2_140 451 12.1 0.092 0.098 2.604 2.713 transfer_rs2pw 451 10.6 0.006 0.007 2.413 2.658 acc_transpose_blocks 49320 15.4 0.206 0.212 2.574 2.639 jit_kernel_multiply 13 15.9 2.527 2.540 2.527 2.540 make_images_data 4110 15.4 0.043 0.046 2.097 2.426 grid_collocate_task_list 110 9.6 2.157 2.283 2.157 2.283 hybrid_alltoall_any 4261 16.3 0.085 0.481 1.783 2.250 mp_waitany 14300 13.8 1.946 2.237 1.946 2.237 cp_fm_cholesky_invert 11 10.9 2.231 2.236 2.231 2.236 transfer_pw2rs 451 13.1 0.006 0.006 2.133 2.168 mp_sum_d 3893 11.9 1.471 2.022 1.471 2.022 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.972 1.995 mp_alltoall_z22v 1111 15.6 1.658 1.858 1.658 1.858 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=88.380000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=474.000000, yerr=3.357488 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 590.729216E+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 2253677. 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 0.106 0.236 71.295 71.296 qs_mol_dyn_low 1 2.0 0.047 0.354 70.183 70.226 qs_forces 11 3.9 0.005 0.024 70.013 70.061 qs_energies 11 4.9 0.008 0.052 66.659 66.710 scf_env_do_scf 11 5.9 0.001 0.002 56.648 56.651 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.974 48.974 dbcsr_multiply_generic 2055 12.4 0.116 0.121 38.021 38.161 velocity_verlet 10 3.0 0.006 0.043 36.124 36.125 qs_scf_new_mos 99 7.5 0.001 0.001 33.220 33.387 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.219 33.386 multiply_cannon 2055 13.4 0.223 0.245 31.391 32.387 ot_scf_mini 99 9.5 0.003 0.004 31.551 31.717 multiply_cannon_loop 2055 14.4 1.155 1.177 30.153 31.003 ot_mini 99 10.5 0.001 0.001 18.368 18.535 multiply_cannon_multrec 24660 15.4 7.032 8.433 14.130 15.433 rebuild_ks_matrix 110 8.3 0.000 0.000 13.664 13.780 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.019 13.664 13.779 qs_ot_get_derivative 99 11.5 0.001 0.001 12.581 12.746 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.048 12.150 mp_waitall_1 176588 16.5 7.467 9.913 7.467 9.913 multiply_cannon_metrocomm3 24660 15.4 0.072 0.076 5.053 7.798 multiply_cannon_sync_h2d 24660 15.4 6.485 7.705 6.485 7.705 init_scf_loop 11 6.9 0.002 0.011 7.638 7.639 dbcsr_mm_accdrv_process 52282 16.1 5.419 6.251 6.929 7.241 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.480 7.180 apply_single 110 13.6 0.000 0.001 6.480 7.180 init_scf_run 11 5.9 0.001 0.016 7.119 7.120 scf_env_initial_rho_setup 11 6.9 0.015 0.021 7.119 7.120 qs_ot_get_p 110 10.4 0.003 0.008 6.532 6.743 sum_up_and_integrate 110 10.3 0.002 0.003 6.319 6.330 integrate_v_rspace 110 11.3 0.002 0.003 6.289 6.305 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.751 5.759 calculate_rho_elec 110 8.6 0.039 0.047 5.750 5.758 ot_diis_step 99 11.5 0.010 0.010 5.739 5.739 prepare_preconditioner 11 7.9 0.000 0.000 5.521 5.543 make_preconditioner 11 8.9 0.000 0.001 5.521 5.543 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.727 5.446 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.086 5.241 make_m2s 4110 13.4 0.057 0.059 4.222 4.634 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.573 4.592 make_images 4110 14.4 0.407 0.465 4.110 4.518 cp_dbcsr_syevd 48 12.0 0.003 0.006 4.110 4.110 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.566 3.575 cp_fm_diag_elpa 48 13.0 0.000 0.001 3.541 3.555 cp_fm_diag_elpa_base 48 14.0 3.493 3.513 3.537 3.549 wfi_extrapolate 11 7.9 0.002 0.011 3.532 3.533 calculate_first_density_matrix 1 7.0 0.000 0.004 3.443 3.447 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.285 3.376 grid_integrate_task_list 110 12.3 3.164 3.308 3.164 3.308 fft_wrap_pw1pw2 1111 11.6 0.015 0.017 3.147 3.284 density_rs2pw 110 9.6 0.004 0.006 2.989 3.218 calculate_dm_sparse 110 9.5 0.001 0.001 3.030 3.061 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.915 2.955 hybrid_alltoall_any 4261 16.3 0.106 0.458 2.021 2.802 fft_wrap_pw1pw2_140 451 12.1 0.110 0.116 2.664 2.797 make_images_data 4110 15.4 0.048 0.052 2.301 2.748 fft3d_ps 1111 13.6 1.070 1.279 2.525 2.651 cp_fm_cholesky_invert 11 10.9 2.529 2.536 2.529 2.536 mp_sum_l 6594 12.7 1.817 2.405 1.817 2.405 grid_collocate_task_list 110 9.6 2.173 2.345 2.173 2.345 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.271 2.322 potential_pw2rs 110 12.3 0.008 0.009 2.186 2.199 transfer_rs2pw 451 10.6 0.007 0.008 1.853 2.097 mp_alltoall_d11v 2046 13.8 1.794 2.052 1.794 2.052 acc_transpose_blocks 24660 15.4 0.115 0.119 1.991 2.017 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.965 1.982 jit_kernel_multiply 10 16.4 1.139 1.978 1.139 1.978 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.793 1.805 dbcsr_complete_redistribute 325 12.2 0.298 0.623 1.467 1.728 multiply_cannon_metrocomm4 22605 15.4 0.078 0.082 0.745 1.648 mp_waitany 10164 13.8 1.426 1.642 1.426 1.642 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.597 1.606 transfer_rs2pw_140 121 11.5 0.374 0.427 1.318 1.562 mp_allgather_i34 2055 14.4 0.547 1.554 0.547 1.554 mp_irecv_dv 57340 16.2 0.615 1.525 0.615 1.525 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.389 1.491 make_basis_sm 11 9.8 0.001 0.004 1.482 1.485 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.356 1.476 parallel_gemm_fm 81 9.0 0.000 0.000 1.455 1.471 parallel_gemm_fm_cosma 81 10.0 1.454 1.471 1.454 1.471 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.296000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=558.727273, yerr=7.097899 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 666.460160E+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 0.011 0.025 61.125 61.126 qs_mol_dyn_low 1 2.0 0.009 0.051 60.578 60.588 qs_forces 11 3.9 0.026 0.055 60.439 60.445 qs_energies 11 4.9 0.001 0.002 57.239 57.272 scf_env_do_scf 11 5.9 0.000 0.001 48.725 48.725 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 40.031 40.032 velocity_verlet 10 3.0 0.002 0.011 32.803 32.809 dbcsr_multiply_generic 2055 12.4 0.110 0.114 29.208 29.521 qs_scf_new_mos 99 7.5 0.001 0.001 25.559 25.664 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.558 25.663 ot_scf_mini 99 9.5 0.002 0.003 24.289 24.398 multiply_cannon 2055 13.4 0.211 0.218 22.836 24.238 multiply_cannon_loop 2055 14.4 0.813 0.847 21.682 22.997 ot_mini 99 10.5 0.001 0.001 13.798 13.907 rebuild_ks_matrix 110 8.3 0.000 0.000 12.172 12.357 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.171 12.356 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.726 10.895 multiply_cannon_multrec 16440 15.4 3.615 4.494 9.824 10.643 mp_waitall_1 139946 16.5 7.020 10.205 7.020 10.205 qs_ot_get_derivative 99 11.5 0.001 0.001 9.361 9.472 init_scf_loop 11 6.9 0.000 0.000 8.658 8.660 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.396 7.358 prepare_preconditioner 11 7.9 0.000 0.000 6.864 6.882 make_preconditioner 11 8.9 0.000 0.000 6.864 6.882 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.142 6.489 dbcsr_mm_accdrv_process 34862 16.1 5.345 5.683 6.055 6.194 sum_up_and_integrate 110 10.3 0.001 0.002 6.178 6.193 integrate_v_rspace 110 11.3 0.003 0.003 6.152 6.167 init_scf_run 11 5.9 0.000 0.001 5.905 5.905 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.904 5.905 qs_ot_get_p 110 10.4 0.005 0.009 5.639 5.778 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.539 5.550 calculate_rho_elec 110 8.6 0.058 0.059 5.538 5.550 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.891 5.372 apply_single 110 13.6 0.000 0.000 4.891 5.372 make_m2s 4110 13.4 0.049 0.050 4.265 4.620 make_images 4110 14.4 0.396 0.521 4.149 4.502 ot_diis_step 99 11.5 0.011 0.011 4.406 4.406 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.940 3.943 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.243 3.847 multiply_cannon_sync_h2d 16440 15.4 3.217 3.705 3.217 3.705 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.568 3.568 grid_integrate_task_list 110 12.3 3.199 3.431 3.199 3.431 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.144 3.146 fft_wrap_pw1pw2 1111 11.6 0.014 0.017 3.095 3.103 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.013 3.022 cp_fm_diag_elpa_base 48 14.0 2.945 2.973 3.011 3.021 wfi_extrapolate 11 7.9 0.001 0.001 3.007 3.007 density_rs2pw 110 9.6 0.004 0.004 2.778 2.936 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.804 2.876 hybrid_alltoall_any 4261 16.3 0.109 0.382 2.160 2.827 make_images_data 4110 15.4 0.045 0.050 2.394 2.825 calculate_first_density_matrix 1 7.0 0.000 0.000 2.809 2.810 cp_fm_cholesky_invert 11 10.9 2.657 2.664 2.657 2.664 fft_wrap_pw1pw2_140 451 12.1 0.127 0.130 2.632 2.641 calculate_dm_sparse 110 9.5 0.001 0.001 2.544 2.577 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.465 2.525 multiply_cannon_metrocomm4 14385 15.4 0.048 0.052 0.895 2.524 fft3d_ps 1111 13.6 1.088 1.100 2.407 2.420 grid_collocate_task_list 110 9.6 2.223 2.391 2.223 2.391 mp_irecv_dv 48980 15.7 0.820 2.386 0.820 2.386 mp_alltoall_d11v 2046 13.8 1.852 2.241 1.852 2.241 mp_sum_l 6594 12.7 1.534 2.160 1.534 2.160 qs_energies_init_hamiltonians 11 5.9 0.023 0.091 2.134 2.156 potential_pw2rs 110 12.3 0.009 0.010 2.037 2.046 acc_transpose_blocks 16440 15.4 0.076 0.078 1.937 2.003 dbcsr_complete_redistribute 325 12.2 0.322 0.346 1.503 1.954 cp_fm_upper_to_full 70 14.2 1.383 1.753 1.383 1.753 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.646 1.663 cp_fm_cholesky_decompose 22 10.9 1.610 1.636 1.610 1.636 transfer_rs2pw 451 10.6 0.005 0.006 1.402 1.609 mp_allgather_i34 2055 14.4 0.469 1.558 0.469 1.558 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.008 1.368 1.486 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.461 1.474 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.026 1.464 mp_waitany 17072 13.8 1.233 1.440 1.233 1.440 qs_env_update_s_mstruct 11 6.9 0.005 0.018 1.194 1.296 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.283 1.292 rs_gather_matrices 110 12.3 0.137 0.150 0.865 1.254 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.126000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.363636, yerr=8.412898 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 744.914944E+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 0.063 0.096 66.580 66.590 qs_mol_dyn_low 1 2.0 0.004 0.012 65.858 65.881 qs_forces 11 3.9 0.041 0.053 65.789 65.800 qs_energies 11 4.9 0.001 0.002 62.388 62.430 scf_env_do_scf 11 5.9 0.000 0.001 53.060 53.068 scf_env_do_scf_inner_loop 99 6.5 0.004 0.009 41.499 41.505 velocity_verlet 10 3.0 0.002 0.006 36.811 36.819 dbcsr_multiply_generic 2055 12.4 0.117 0.121 30.813 31.009 qs_scf_new_mos 99 7.5 0.001 0.001 27.257 27.362 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.257 27.361 ot_scf_mini 99 9.5 0.002 0.003 25.558 25.649 multiply_cannon 2055 13.4 0.237 0.250 23.910 25.021 multiply_cannon_loop 2055 14.4 1.411 1.471 22.527 23.067 ot_mini 99 10.5 0.001 0.001 14.695 14.812 multiply_cannon_multrec 24660 15.4 4.111 6.849 13.083 14.313 rebuild_ks_matrix 110 8.3 0.000 0.000 11.835 11.916 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.835 11.915 init_scf_loop 11 6.9 0.000 0.000 11.521 11.523 qs_ot_get_derivative 99 11.5 0.001 0.001 10.539 10.634 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.463 10.534 prepare_preconditioner 11 7.9 0.000 0.000 9.797 9.812 make_preconditioner 11 8.9 0.000 0.000 9.796 9.812 dbcsr_mm_accdrv_process 52304 16.0 7.717 8.923 8.816 9.713 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.008 9.468 init_scf_run 11 5.9 0.001 0.023 6.476 6.477 scf_env_initial_rho_setup 11 6.9 0.003 0.023 6.475 6.477 sum_up_and_integrate 110 10.3 0.001 0.002 6.055 6.066 integrate_v_rspace 110 11.3 0.003 0.003 6.029 6.039 qs_ot_get_p 110 10.4 0.001 0.001 5.771 5.902 mp_waitall_1 121746 16.5 4.039 5.763 4.039 5.763 make_m2s 4110 13.4 0.059 0.061 5.318 5.670 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.535 5.543 calculate_rho_elec 110 8.6 0.077 0.081 5.534 5.543 make_images 4110 14.4 0.574 0.698 5.176 5.524 cp_fm_upper_to_full 70 14.2 3.386 4.905 3.386 4.905 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.038 4.138 apply_single 110 13.6 0.000 0.000 4.037 4.138 ot_diis_step 99 11.5 0.011 0.011 4.118 4.118 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.934 3.948 dbcsr_complete_redistribute 325 12.2 0.425 0.475 2.651 3.789 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.511 3.556 calculate_first_density_matrix 1 7.0 0.000 0.000 3.437 3.442 grid_integrate_task_list 110 12.3 3.301 3.428 3.301 3.428 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.377 3.378 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.320 3.376 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.138 3.261 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.169 3.171 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.142 3.164 multiply_cannon_metrocomm3 24660 15.4 0.038 0.040 1.324 3.060 calculate_dm_sparse 110 9.5 0.001 0.001 3.002 3.039 wfi_extrapolate 11 7.9 0.001 0.001 2.945 2.945 hybrid_alltoall_any 4261 16.3 0.123 0.470 2.260 2.938 make_images_data 4110 15.4 0.048 0.052 2.593 2.921 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.779 2.891 acc_transpose_blocks 24660 15.4 0.109 0.111 2.801 2.888 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.848 2.857 cp_fm_diag_elpa_base 48 14.0 2.701 2.759 2.845 2.855 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.802 2.830 mp_alltoall_i22 605 13.7 1.652 2.815 1.652 2.815 density_rs2pw 110 9.6 0.004 0.004 2.682 2.813 fft_wrap_pw1pw2_140 451 12.1 0.145 0.147 2.690 2.715 cp_fm_cholesky_invert 11 10.9 2.615 2.624 2.615 2.624 multiply_cannon_sync_h2d 24660 15.4 2.388 2.490 2.388 2.490 fft3d_ps 1111 13.6 1.123 1.163 2.411 2.428 grid_collocate_task_list 110 9.6 2.261 2.391 2.261 2.391 qs_energies_init_hamiltonians 11 5.9 0.003 0.009 2.325 2.364 potential_pw2rs 110 12.3 0.012 0.013 1.883 1.889 mp_alltoall_d11v 2046 13.8 1.689 1.871 1.689 1.871 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.795 1.828 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.615 1.705 cp_fm_cholesky_decompose 22 10.9 1.606 1.663 1.606 1.663 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.616 1.626 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.546 1.562 multiply_cannon_metrocomm4 20550 15.4 0.059 0.063 0.836 1.542 mp_sum_l 6594 12.7 0.920 1.523 0.920 1.523 acc_transpose_blocks_sync 73980 16.4 1.413 1.512 1.413 1.512 mp_irecv_dv 62702 16.1 0.733 1.459 0.733 1.459 mp_allgather_i34 2055 14.4 0.474 1.443 0.474 1.443 make_basis_sm 11 9.8 0.000 0.000 1.394 1.396 parallel_gemm_fm 81 9.0 0.000 0.000 1.338 1.349 parallel_gemm_fm_cosma 81 10.0 1.338 1.349 1.338 1.349 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.590000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=704.636364, yerr=10.653832 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 878.825472E+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 0.048 0.076 55.186 55.186 qs_mol_dyn_low 1 2.0 0.003 0.003 54.473 54.519 qs_forces 11 3.9 0.035 0.054 54.401 54.403 qs_energies 11 4.9 0.014 0.019 50.731 50.766 scf_env_do_scf 11 5.9 0.000 0.001 42.252 42.252 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.648 34.650 velocity_verlet 10 3.0 0.002 0.003 30.716 30.719 dbcsr_multiply_generic 2055 12.4 0.106 0.108 23.097 23.316 qs_scf_new_mos 99 7.5 0.001 0.001 20.557 20.631 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.557 20.631 ot_scf_mini 99 9.5 0.002 0.002 19.311 19.359 multiply_cannon 2055 13.4 0.238 0.249 17.647 18.731 multiply_cannon_loop 2055 14.4 0.608 0.627 16.429 16.871 rebuild_ks_matrix 110 8.3 0.000 0.000 11.436 11.503 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.435 11.503 ot_mini 99 10.5 0.001 0.001 10.700 10.749 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.176 10.237 multiply_cannon_multrec 8220 15.4 3.158 4.661 7.575 8.690 init_scf_loop 11 6.9 0.000 0.000 7.558 7.559 mp_waitall_1 103326 16.6 5.790 7.515 5.790 7.515 qs_ot_get_derivative 99 11.5 0.001 0.001 6.934 6.983 sum_up_and_integrate 110 10.3 0.001 0.002 6.026 6.038 integrate_v_rspace 110 11.3 0.003 0.003 5.999 6.011 prepare_preconditioner 11 7.9 0.000 0.000 5.927 5.933 make_preconditioner 11 8.9 0.000 0.000 5.927 5.933 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.641 5.654 calculate_rho_elec 110 8.6 0.115 0.116 5.640 5.653 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.516 5.586 init_scf_run 11 5.9 0.000 0.001 5.221 5.221 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.220 5.221 dbcsr_mm_accdrv_process 17442 15.9 3.089 4.159 4.277 5.203 qs_ot_get_p 110 10.4 0.001 0.002 4.679 4.752 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 3.103 4.651 make_m2s 4110 13.4 0.037 0.039 4.109 4.298 make_images 4110 14.4 0.647 0.709 3.979 4.164 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.797 3.862 apply_single 110 13.6 0.000 0.000 3.796 3.862 ot_diis_step 99 11.5 0.012 0.012 3.744 3.744 grid_integrate_task_list 110 12.3 3.380 3.518 3.380 3.518 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.324 3.328 fft_wrap_pw1pw2 1111 11.6 0.014 0.015 3.317 3.323 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.985 2.986 cp_fm_cholesky_invert 11 10.9 2.868 2.871 2.868 2.871 fft_wrap_pw1pw2_140 451 12.1 0.192 0.195 2.857 2.865 qs_energies_init_hamiltonians 11 5.9 0.025 0.035 2.766 2.798 density_rs2pw 110 9.6 0.003 0.004 2.670 2.765 wfi_extrapolate 11 7.9 0.001 0.001 2.662 2.662 calculate_dm_sparse 110 9.5 0.001 0.001 2.504 2.547 grid_collocate_task_list 110 9.6 2.369 2.512 2.369 2.512 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.509 2.512 hybrid_alltoall_any 4261 16.3 0.203 0.862 2.123 2.505 multiply_cannon_sync_h2d 8220 15.4 2.353 2.489 2.353 2.489 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.479 2.487 cp_fm_diag_elpa_base 48 14.0 2.424 2.448 2.477 2.485 fft3d_ps 1111 13.6 1.270 1.281 2.476 2.485 calculate_first_density_matrix 1 7.0 0.000 0.000 2.465 2.468 make_images_data 4110 15.4 0.041 0.047 2.173 2.433 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.141 2.163 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.001 2.040 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.793 2.017 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.900 1.927 potential_pw2rs 110 12.3 0.015 0.015 1.828 1.831 mp_alltoall_d11v 2046 13.8 1.577 1.791 1.577 1.791 cp_fm_cholesky_decompose 22 10.9 1.651 1.668 1.651 1.668 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.556 1.661 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.590 1.596 jit_kernel_multiply 9 15.5 0.867 1.541 0.867 1.541 dbcsr_complete_redistribute 325 12.2 0.559 0.581 1.421 1.500 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.461 1.474 acc_transpose_blocks 8220 15.4 0.039 0.040 1.387 1.428 mp_allgather_i34 2055 14.4 0.390 1.402 0.390 1.402 qs_create_task_list 11 7.9 0.000 0.001 1.212 1.312 generate_qs_task_list 11 8.9 0.376 0.444 1.212 1.311 multiply_cannon_metrocomm4 6165 15.4 0.018 0.020 0.484 1.304 transfer_rs2pw 451 10.6 0.005 0.005 1.169 1.273 mp_irecv_dv 24056 15.7 0.457 1.260 0.457 1.260 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.705 1.225 mp_waitany 9240 13.8 1.098 1.219 1.098 1.219 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.108 1.144 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.186000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=824.000000, yerr=16.868475 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 1.373897E+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 0.111 0.114 88.476 88.476 qs_mol_dyn_low 1 2.0 0.003 0.003 87.522 87.556 qs_forces 11 3.9 0.003 0.003 87.448 87.449 qs_energies 11 4.9 0.002 0.002 83.339 83.341 scf_env_do_scf 11 5.9 0.000 0.001 72.301 72.301 velocity_verlet 10 3.0 0.005 0.009 55.520 55.527 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.954 43.955 dbcsr_multiply_generic 2055 12.4 0.121 0.125 29.995 30.043 init_scf_loop 11 6.9 0.000 0.000 28.276 28.280 qs_scf_new_mos 99 7.5 0.001 0.001 27.012 27.053 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.011 27.052 prepare_preconditioner 11 7.9 0.000 0.000 26.283 26.290 make_preconditioner 11 8.9 0.000 0.000 26.283 26.289 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.409 25.751 ot_scf_mini 99 9.5 0.002 0.002 25.205 25.243 multiply_cannon 2055 13.4 0.338 0.366 22.755 23.487 multiply_cannon_loop 2055 14.4 0.826 0.847 20.919 21.304 cp_fm_upper_to_full 70 14.2 12.768 18.362 12.768 18.362 ot_mini 99 10.5 0.001 0.001 14.243 14.273 rebuild_ks_matrix 110 8.3 0.000 0.000 13.477 13.525 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.477 13.525 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.222 12.265 dbcsr_complete_redistribute 325 12.2 1.026 1.047 7.418 10.748 multiply_cannon_multrec 8220 15.4 4.062 4.237 9.943 10.032 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.419 9.757 qs_ot_get_derivative 99 11.5 0.001 0.001 9.498 9.535 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.858 9.168 mp_alltoall_i22 605 13.7 5.471 8.825 5.471 8.825 mp_waitall_1 84994 16.7 7.675 8.521 7.675 8.521 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.831 6.871 calculate_rho_elec 110 8.6 0.227 0.227 6.831 6.870 sum_up_and_integrate 110 10.3 0.002 0.002 6.588 6.603 integrate_v_rspace 110 11.3 0.003 0.004 6.560 6.575 init_scf_run 11 5.9 0.000 0.001 6.503 6.503 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.503 6.503 make_m2s 4110 13.4 0.042 0.043 5.495 6.034 dbcsr_mm_accdrv_process 11614 15.7 3.838 4.114 5.733 5.941 make_images 4110 14.4 0.877 0.917 5.306 5.845 qs_ot_get_p 110 10.4 0.002 0.002 5.629 5.660 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.874 5.315 apply_single 110 13.6 0.000 0.000 4.874 5.314 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.848 5.303 cp_fm_cholesky_invert 11 10.9 5.198 5.202 5.198 5.202 ot_diis_step 99 11.5 0.015 0.015 4.720 4.721 fft_wrap_pw1pw2 1111 11.6 0.016 0.016 4.475 4.486 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.051 4.061 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.930 3.931 fft_wrap_pw1pw2_140 451 12.1 0.364 0.365 3.905 3.921 hybrid_alltoall_any 4261 16.3 0.264 0.566 2.980 3.782 grid_integrate_task_list 110 12.3 3.679 3.747 3.679 3.747 make_images_data 4110 15.4 0.044 0.047 2.992 3.660 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.534 3.534 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.000 3.453 calculate_dm_sparse 110 9.5 0.001 0.001 3.318 3.343 fft3d_ps 1111 13.6 1.832 1.850 3.333 3.341 density_rs2pw 110 9.6 0.004 0.004 3.296 3.310 wfi_extrapolate 11 7.9 0.001 0.001 3.286 3.286 multiply_cannon_sync_h2d 8220 15.4 3.126 3.151 3.126 3.151 calculate_first_density_matrix 1 7.0 0.000 0.000 3.107 3.108 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.051 3.056 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.941 2.941 cp_fm_diag_elpa_base 48 14.0 2.396 2.599 2.939 2.939 grid_collocate_task_list 110 9.6 2.668 2.677 2.668 2.677 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.657 2.677 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.560 2.575 qs_env_update_s_mstruct 11 6.9 0.001 0.002 2.446 2.502 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.097 2.211 potential_pw2rs 110 12.3 0.021 0.021 2.111 2.116 mp_alltoall_d11v 2046 13.8 2.010 2.066 2.010 2.066 qs_create_task_list 11 7.9 0.036 0.037 1.922 1.968 cp_fm_cholesky_decompose 22 10.9 1.947 1.962 1.947 1.962 jit_kernel_multiply 10 15.3 1.697 1.945 1.697 1.945 generate_qs_task_list 11 8.9 0.732 0.787 1.886 1.932 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.856 1.861 acc_transpose_blocks 8220 15.4 0.041 0.041 1.812 1.846 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.810 1.825 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.476000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1247.727273, yerr=46.896580 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 632.524800E+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 2529110. 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 0.038 0.138 207.565 207.567 qs_mol_dyn_low 1 2.0 0.006 0.026 206.851 206.865 qs_forces 11 3.9 0.005 0.005 206.572 206.574 qs_energies 11 4.9 0.001 0.002 200.996 201.017 scf_env_do_scf 11 5.9 0.001 0.001 184.488 184.492 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 163.934 163.936 dbcsr_multiply_generic 2507 12.6 0.178 0.185 126.044 126.605 qs_scf_new_mos 117 7.6 0.001 0.001 124.981 125.132 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.981 125.131 velocity_verlet 10 3.0 0.008 0.057 123.256 123.264 ot_scf_mini 117 9.6 0.003 0.003 118.254 118.429 multiply_cannon 2507 13.6 0.237 0.245 101.671 102.846 multiply_cannon_loop 2507 14.6 2.381 2.441 99.544 100.712 ot_mini 117 10.6 0.001 0.001 66.568 66.741 multiply_cannon_multrec 60168 15.6 31.675 33.332 41.458 43.103 qs_ot_get_derivative 117 11.6 0.001 0.001 41.763 41.916 rebuild_ks_matrix 128 8.3 0.001 0.001 33.595 33.868 qs_ks_build_kohn_sham_matrix 128 9.3 0.014 0.017 33.595 33.868 mp_waitall_1 267128 16.5 29.063 31.975 29.063 31.975 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.183 30.448 qs_ot_get_p 128 10.4 0.001 0.001 29.435 29.740 multiply_cannon_sync_h2d 60168 15.6 26.315 27.852 26.315 27.852 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.282 25.091 apply_single 128 13.6 0.001 0.001 24.282 25.091 ot_diis_step 117 11.6 0.008 0.008 24.523 24.524 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.843 22.942 init_scf_loop 11 6.9 0.000 0.000 20.478 20.479 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.325 20.478 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.012 20.012 multiply_cannon_metrocomm3 60168 15.6 0.118 0.122 16.258 18.035 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.052 17.087 cp_fm_diag_elpa_base 83 14.4 16.986 17.021 17.048 17.081 prepare_preconditioner 11 7.9 0.000 0.000 15.911 15.956 make_preconditioner 11 8.9 0.000 0.000 15.911 15.956 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.140 15.284 make_m2s 5014 13.6 0.105 0.114 13.697 14.139 make_images 5014 14.6 0.398 0.417 13.511 13.966 sum_up_and_integrate 128 10.3 0.002 0.004 13.717 13.735 integrate_v_rspace 128 11.3 0.003 0.004 13.658 13.673 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.599 12.791 calculate_rho_elec 128 8.7 0.045 0.064 12.598 12.790 init_scf_run 11 5.9 0.000 0.002 12.320 12.320 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.320 12.320 mp_sum_l 7950 12.9 9.436 10.575 9.436 10.575 dbcsr_mm_accdrv_process 124484 16.2 4.725 4.902 9.356 9.838 cp_fm_cholesky_invert 11 10.9 9.156 9.164 9.156 9.164 wfi_extrapolate 11 7.9 0.001 0.001 9.067 9.067 calculate_dm_sparse 128 9.5 0.001 0.001 8.534 8.648 multiply_cannon_metrocomm1 60168 15.6 0.095 0.099 6.435 8.495 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.218 8.311 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.104 8.221 make_images_data 5014 15.6 0.065 0.071 6.680 7.565 grid_integrate_task_list 128 12.3 7.080 7.473 7.080 7.473 hybrid_alltoall_any 5200 16.5 0.295 2.275 5.860 7.211 density_rs2pw 128 9.7 0.005 0.006 6.187 6.854 fft_wrap_pw1pw2 1291 11.7 0.018 0.025 6.479 6.756 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.639 6.651 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.790 5.903 fft_wrap_pw1pw2_140 523 12.2 0.233 0.242 5.639 5.837 fft3d_ps 1291 13.7 2.091 2.574 5.303 5.513 grid_collocate_task_list 128 9.7 4.848 5.266 4.848 5.266 mp_alltoall_d11v 2415 14.1 4.322 5.253 4.322 5.253 cp_fm_cholesky_decompose 22 10.9 4.561 4.574 4.561 4.574 potential_pw2rs 128 12.3 0.009 0.010 4.213 4.232 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.567000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.909091, yerr=7.573333 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 843.563008E+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 4003 58199. MP_Allreduce 11086 960. MP_Sync 86 MP_Alltoall 1955 5616370. 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 0.048 0.113 186.841 186.843 qs_mol_dyn_low 1 2.0 0.009 0.028 186.241 186.255 qs_forces 11 3.9 0.004 0.005 186.037 186.041 qs_energies 11 4.9 0.001 0.002 179.258 179.271 scf_env_do_scf 11 5.9 0.001 0.001 163.002 163.012 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 130.462 130.465 velocity_verlet 10 3.0 0.023 0.095 117.021 117.030 dbcsr_multiply_generic 2485 12.5 0.185 0.190 96.076 97.430 qs_scf_new_mos 116 7.6 0.001 0.001 92.793 93.399 qs_scf_loop_do_ot 116 8.6 0.001 0.001 92.793 93.398 ot_scf_mini 116 9.6 0.004 0.004 88.042 88.686 multiply_cannon 2485 13.5 0.498 0.557 76.114 80.382 multiply_cannon_loop 2485 14.5 1.545 1.615 73.284 76.191 ot_mini 116 10.6 0.001 0.001 49.607 50.185 mp_waitall_1 212858 16.6 23.792 37.227 23.792 37.227 multiply_cannon_multrec 29820 15.5 20.737 25.667 31.313 36.618 rebuild_ks_matrix 127 8.3 0.001 0.001 32.012 32.763 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 32.011 32.763 init_scf_loop 11 6.9 0.000 0.000 32.451 32.453 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.782 29.478 multiply_cannon_metrocomm3 29820 15.5 0.095 0.100 15.569 28.766 qs_ot_get_derivative 116 11.6 0.001 0.002 27.823 28.460 prepare_preconditioner 11 7.9 0.000 0.000 28.144 28.222 make_preconditioner 11 8.9 0.000 0.000 28.144 28.222 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.834 27.382 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.916 23.088 apply_single 127 13.6 0.001 0.001 21.916 23.088 ot_diis_step 116 11.6 0.014 0.015 21.607 21.609 qs_ot_get_p 127 10.4 0.001 0.001 20.758 21.540 multiply_cannon_sync_h2d 29820 15.5 17.826 20.754 17.826 20.754 cp_fm_cholesky_invert 11 10.9 16.375 16.389 16.375 16.389 qs_ot_p2m_diag 82 11.4 0.186 0.214 15.912 15.945 make_m2s 4970 13.5 0.086 0.092 14.005 15.106 make_images 4970 14.5 1.156 1.356 13.790 14.892 cp_dbcsr_syevd 82 12.4 0.005 0.006 14.732 14.733 sum_up_and_integrate 127 10.3 0.002 0.004 13.577 13.610 integrate_v_rspace 127 11.3 0.003 0.004 13.517 13.551 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.693 12.721 calculate_rho_elec 127 8.7 0.087 0.105 12.692 12.721 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.615 11.647 cp_fm_diag_elpa_base 82 14.4 11.368 11.452 11.609 11.637 init_scf_run 11 5.9 0.000 0.002 11.418 11.420 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.418 11.420 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.966 11.419 multiply_cannon_metrocomm4 27335 15.5 0.105 0.116 3.841 10.965 dbcsr_mm_accdrv_process 61726 16.2 5.348 6.370 10.028 10.597 mp_irecv_dv 68888 16.3 3.632 10.561 3.632 10.561 make_images_data 4970 15.5 0.065 0.074 8.247 9.607 hybrid_alltoall_any 5155 16.4 0.347 1.522 6.964 9.162 wfi_extrapolate 11 7.9 0.001 0.001 8.247 8.247 grid_integrate_task_list 127 12.3 7.152 7.438 7.152 7.438 fft_wrap_pw1pw2 1281 11.7 0.019 0.025 7.308 7.391 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.229 6.986 cp_fm_cholesky_decompose 22 10.9 6.834 6.913 6.834 6.913 density_rs2pw 127 9.7 0.005 0.006 6.448 6.862 calculate_dm_sparse 127 9.5 0.001 0.001 6.404 6.563 fft_wrap_pw1pw2_140 519 12.2 0.249 0.261 6.430 6.498 mp_sum_l 7884 12.9 4.226 6.264 4.226 6.264 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.135 6.145 fft3d_ps 1281 13.7 2.769 2.946 5.758 5.834 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.325 5.401 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.169 5.390 grid_collocate_task_list 127 9.7 4.967 5.357 4.967 5.357 mp_alltoall_d11v 2401 14.1 4.078 4.437 4.078 4.437 mp_allgather_i34 2485 14.5 1.363 4.377 1.363 4.377 potential_pw2rs 127 12.3 0.014 0.015 4.135 4.153 dbcsr_complete_redistribute 393 12.7 0.779 0.900 3.199 4.026 mp_sum_d 4456 12.1 2.640 3.995 2.640 3.995 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=186.843000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=804.000000, yerr=1.705606 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.950886E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 4015680 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.9 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 972.943360E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1051232 MPI messages size (bytes): total size 2.737039E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.603648E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 284160 37245419520 131072 < size <= 4194304 665408 1004401590272 4194304 < size <= 16777216 66080 937889764224 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57812. MP_Allreduce 11152 997. MP_Sync 87 MP_Alltoall 1724 9394172. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.053 0.078 176.006 176.008 qs_mol_dyn_low 1 2.0 0.008 0.044 175.248 175.318 qs_forces 11 3.9 0.021 0.031 175.036 175.040 qs_energies 11 4.9 0.013 0.098 168.411 168.443 scf_env_do_scf 11 5.9 0.001 0.002 152.566 152.568 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 117.870 117.870 velocity_verlet 10 3.0 0.010 0.077 112.625 112.631 dbcsr_multiply_generic 2527 12.6 0.185 0.190 82.517 83.688 qs_scf_new_mos 118 7.6 0.001 0.001 80.863 81.195 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.863 81.194 ot_scf_mini 118 9.6 0.004 0.004 76.569 76.963 multiply_cannon 2527 13.6 0.511 0.529 62.706 67.208 multiply_cannon_loop 2527 14.6 1.136 1.194 59.592 62.320 ot_mini 118 10.6 0.001 0.001 42.710 43.119 init_scf_loop 11 6.9 0.001 0.005 34.595 34.596 mp_waitall_1 171878 16.6 24.737 33.958 24.737 33.958 rebuild_ks_matrix 129 8.3 0.001 0.001 30.600 31.065 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.022 30.600 31.064 prepare_preconditioner 11 7.9 0.000 0.000 30.555 30.603 make_preconditioner 11 8.9 0.000 0.001 30.555 30.603 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.125 29.522 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.613 28.039 multiply_cannon_multrec 20216 15.6 13.223 15.874 22.666 25.373 multiply_cannon_metrocomm3 20216 15.6 0.061 0.064 15.300 25.139 qs_ot_get_derivative 118 11.6 0.001 0.002 23.013 23.413 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.742 20.691 apply_single 129 13.6 0.001 0.001 19.742 20.690 ot_diis_step 118 11.6 0.018 0.023 19.584 19.584 qs_ot_get_p 129 10.4 0.001 0.002 18.975 19.450 make_m2s 5054 13.6 0.078 0.085 14.803 16.217 make_images 5054 14.6 1.157 1.247 14.565 15.972 multiply_cannon_sync_h2d 20216 15.6 13.721 15.167 13.721 15.167 qs_ot_p2m_diag 83 11.4 0.265 0.272 14.689 14.699 cp_fm_cholesky_invert 11 10.9 14.131 14.139 14.131 14.139 sum_up_and_integrate 129 10.3 0.002 0.003 13.709 13.734 integrate_v_rspace 129 11.3 0.003 0.004 13.648 13.675 cp_dbcsr_syevd 83 12.4 0.009 0.033 13.635 13.636 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.096 13.125 calculate_rho_elec 129 8.7 0.133 0.147 13.095 13.125 make_images_data 5054 15.6 0.062 0.072 8.886 10.681 init_scf_run 11 5.9 0.001 0.003 10.500 10.500 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.499 10.500 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.394 10.423 cp_fm_diag_elpa_base 83 14.4 9.971 10.134 10.379 10.406 hybrid_alltoall_any 5240 16.5 0.453 2.045 7.695 10.220 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.027 9.306 multiply_cannon_metrocomm4 17689 15.6 0.067 0.077 3.505 9.250 dbcsr_mm_accdrv_process 41830 16.2 5.672 5.959 8.897 9.026 mp_irecv_dv 50625 16.2 3.373 8.990 3.373 8.990 grid_integrate_task_list 129 12.3 7.361 7.764 7.361 7.764 fft_wrap_pw1pw2 1301 11.7 0.019 0.025 7.598 7.703 wfi_extrapolate 11 7.9 0.004 0.028 7.415 7.415 cp_fm_upper_to_full 105 14.8 5.795 7.323 5.795 7.323 cp_fm_cholesky_decompose 22 10.9 7.138 7.244 7.138 7.244 fft_wrap_pw1pw2_140 527 12.2 0.289 0.296 6.696 6.813 density_rs2pw 129 9.7 0.006 0.009 6.399 6.810 dbcsr_complete_redistribute 395 12.7 1.172 1.238 4.715 6.504 calculate_dm_sparse 129 9.5 0.001 0.001 5.885 5.976 fft3d_ps 1301 13.7 2.801 3.045 5.874 5.949 grid_collocate_task_list 129 9.7 5.209 5.626 5.209 5.626 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.511 5.519 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.722 5.368 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.555 5.342 mp_allgather_i34 2527 14.6 1.606 4.910 1.606 4.910 mp_alltoall_d11v 2423 14.1 4.357 4.880 4.357 4.880 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.562 4.693 mp_sum_l 8010 12.9 3.174 4.661 3.174 4.661 transfer_fm_to_dbcsr 11 9.9 0.017 0.024 2.411 4.163 potential_pw2rs 129 12.3 0.020 0.022 4.089 4.099 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.938 3.976 mp_alltoall_i22 718 14.1 1.978 3.950 1.978 3.950 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.798 3.817 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.503 3.542 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.008000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=920.272727, yerr=8.180808 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.176859E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58189. MP_Allreduce 11085 1083. MP_Sync 86 MP_Alltoall 1700 12496376. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.102 0.152 187.047 187.048 qs_mol_dyn_low 1 2.0 0.003 0.003 186.268 186.303 qs_forces 11 3.9 0.004 0.004 185.605 185.616 qs_energies 11 4.9 0.005 0.031 178.529 178.539 scf_env_do_scf 11 5.9 0.001 0.001 161.811 161.825 velocity_verlet 10 3.0 0.006 0.009 123.432 123.457 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 115.502 115.503 dbcsr_multiply_generic 2485 12.5 0.188 0.193 79.749 80.492 qs_scf_new_mos 116 7.6 0.001 0.001 79.965 80.285 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.965 80.284 ot_scf_mini 116 9.6 0.003 0.003 75.415 75.755 multiply_cannon 2485 13.5 0.545 0.572 55.518 57.790 multiply_cannon_loop 2485 14.5 1.805 1.861 51.689 53.407 init_scf_loop 11 6.9 0.000 0.000 46.187 46.187 ot_mini 116 10.6 0.001 0.001 42.507 42.815 prepare_preconditioner 11 7.9 0.000 0.000 42.107 42.130 make_preconditioner 11 8.9 0.000 0.000 42.107 42.130 make_full_inverse_cholesky 11 9.9 0.011 0.023 35.747 40.815 multiply_cannon_multrec 29820 15.5 13.567 18.627 26.468 31.136 rebuild_ks_matrix 127 8.3 0.001 0.001 29.046 29.310 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 29.045 29.310 mp_waitall_1 146592 16.7 16.668 27.486 16.668 27.486 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.200 26.434 qs_ot_get_derivative 116 11.6 0.001 0.002 23.018 23.365 make_m2s 4970 13.5 0.091 0.097 19.923 21.154 make_images 4970 14.5 1.924 2.237 19.616 20.853 qs_ot_get_p 127 10.4 0.002 0.002 19.175 19.497 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.930 19.487 apply_single 127 13.6 0.001 0.001 18.930 19.487 ot_diis_step 116 11.6 0.017 0.018 19.357 19.358 cp_fm_upper_to_full 104 14.8 11.620 17.094 11.620 17.094 cp_fm_cholesky_invert 11 10.9 15.965 15.974 15.965 15.974 qs_ot_p2m_diag 82 11.4 0.338 0.385 15.066 15.114 multiply_cannon_metrocomm3 29820 15.5 0.050 0.053 6.147 14.961 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.685 13.686 sum_up_and_integrate 127 10.3 0.002 0.003 13.549 13.578 integrate_v_rspace 127 11.3 0.003 0.004 13.489 13.520 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.193 13.244 calculate_rho_elec 127 8.7 0.173 0.189 13.192 13.243 dbcsr_mm_accdrv_process 61748 16.2 8.358 9.087 12.471 13.032 dbcsr_complete_redistribute 393 12.7 1.483 1.603 9.084 12.850 make_images_data 4970 15.5 0.064 0.070 10.650 12.601 hybrid_alltoall_any 5155 16.4 0.525 2.194 9.505 11.862 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.719 11.512 multiply_cannon_sync_h2d 29820 15.5 10.443 11.229 10.443 11.229 init_scf_run 11 5.9 0.001 0.008 10.740 10.741 scf_env_initial_rho_setup 11 6.9 0.002 0.007 10.739 10.741 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.508 10.524 cp_fm_diag_elpa_base 82 14.4 9.542 9.854 10.501 10.516 transfer_fm_to_dbcsr 11 9.9 0.001 0.007 6.337 10.078 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.603 9.856 mp_alltoall_i22 712 14.1 5.586 9.409 5.586 9.409 grid_integrate_task_list 127 12.3 7.482 7.837 7.482 7.837 fft_wrap_pw1pw2 1281 11.7 0.019 0.024 7.737 7.818 wfi_extrapolate 11 7.9 0.001 0.001 7.646 7.646 cp_fm_cholesky_decompose 22 10.9 7.501 7.596 7.501 7.596 fft_wrap_pw1pw2_140 519 12.2 0.325 0.333 6.839 6.957 multiply_cannon_metrocomm4 24850 15.5 0.078 0.087 2.775 6.786 density_rs2pw 127 9.7 0.005 0.005 6.320 6.652 mp_irecv_dv 75445 16.2 2.623 6.514 2.623 6.514 calculate_dm_sparse 127 9.5 0.001 0.001 6.228 6.330 fft3d_ps 1281 13.7 2.917 2.988 5.932 5.978 grid_collocate_task_list 127 9.7 5.246 5.667 5.246 5.667 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.443 5.508 mp_alltoall_d11v 2401 14.1 4.834 5.393 4.834 5.393 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.440 4.537 qs_energies_init_hamiltonians 11 5.9 0.013 0.041 4.425 4.454 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.355 4.447 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.205 4.278 potential_pw2rs 127 12.3 0.022 0.022 3.909 3.920 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.048000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1097.909091, yerr=21.660414 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.557520E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532237E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58346. MP_Allreduce 11057 1167. MP_Sync 87 MP_Alltoall 1712 18838210. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.046 0.080 169.198 169.199 qs_mol_dyn_low 1 2.0 0.003 0.003 168.528 168.546 qs_forces 11 3.9 0.038 0.052 168.436 168.438 qs_energies 11 4.9 0.028 0.049 160.982 161.000 scf_env_do_scf 11 5.9 0.001 0.001 143.085 143.098 velocity_verlet 10 3.0 0.011 0.011 110.751 110.754 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 107.368 107.369 dbcsr_multiply_generic 2507 12.6 0.183 0.191 71.889 72.162 qs_scf_new_mos 117 7.6 0.001 0.001 71.074 71.186 qs_scf_loop_do_ot 117 8.6 0.001 0.001 71.074 71.186 ot_scf_mini 117 9.6 0.003 0.004 66.648 66.730 multiply_cannon 2507 13.6 0.562 0.595 53.317 56.528 multiply_cannon_loop 2507 14.6 0.805 0.833 50.353 51.040 ot_mini 117 10.6 0.001 0.001 37.150 37.236 init_scf_loop 11 6.9 0.000 0.000 35.569 35.570 prepare_preconditioner 11 7.9 0.000 0.000 31.727 31.755 make_preconditioner 11 8.9 0.000 0.000 31.727 31.755 mp_waitall_1 125778 16.7 24.326 30.863 24.326 30.863 make_full_inverse_cholesky 11 9.9 0.014 0.027 29.611 29.874 rebuild_ks_matrix 128 8.3 0.001 0.001 29.127 29.220 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 29.127 29.219 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.571 26.663 multiply_cannon_multrec 10028 15.6 10.434 14.564 18.094 20.955 qs_ot_get_derivative 117 11.6 0.002 0.002 20.240 20.326 multiply_cannon_metrocomm3 10028 15.6 0.025 0.027 12.407 19.534 cp_fm_cholesky_invert 11 10.9 18.146 18.152 18.146 18.152 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 16.941 17.046 apply_single 128 13.6 0.001 0.001 16.941 17.046 qs_ot_get_p 128 10.4 0.002 0.002 16.812 16.901 ot_diis_step 117 11.6 0.020 0.020 16.841 16.841 make_m2s 5014 13.6 0.063 0.069 14.682 15.592 make_images 5014 14.6 2.163 2.597 14.375 15.277 sum_up_and_integrate 128 10.3 0.002 0.002 13.766 13.809 integrate_v_rspace 128 11.3 0.004 0.004 13.705 13.750 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.418 13.450 calculate_rho_elec 128 8.7 0.257 0.268 13.417 13.449 qs_ot_p2m_diag 83 11.4 0.496 0.501 13.101 13.116 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.943 11.944 multiply_cannon_sync_h2d 10028 15.6 10.927 11.231 10.927 11.231 init_scf_run 11 5.9 0.000 0.001 10.968 10.968 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.968 10.968 hybrid_alltoall_any 5200 16.5 0.845 3.795 8.273 9.743 make_images_data 5014 15.6 0.055 0.066 8.423 9.713 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.068 9.079 cp_fm_diag_elpa_base 83 14.4 8.831 8.910 9.065 9.075 fft_wrap_pw1pw2 1291 11.7 0.018 0.019 8.094 8.126 cp_fm_cholesky_decompose 22 10.9 7.894 8.032 7.894 8.032 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.945 8.009 dbcsr_mm_accdrv_process 20762 16.1 3.115 4.490 7.306 7.989 grid_integrate_task_list 128 12.3 7.773 7.989 7.773 7.989 wfi_extrapolate 11 7.9 0.001 0.001 7.289 7.289 fft_wrap_pw1pw2_140 523 12.2 0.435 0.439 7.142 7.179 multiply_cannon_metrocomm1 10028 15.6 0.030 0.031 4.272 7.164 density_rs2pw 128 9.7 0.005 0.005 6.161 6.465 calculate_dm_sparse 128 9.5 0.001 0.001 6.086 6.145 fft3d_ps 1291 13.7 3.131 3.243 6.054 6.085 grid_collocate_task_list 128 9.7 5.566 5.799 5.566 5.799 dbcsr_complete_redistribute 395 12.7 2.096 2.170 5.217 5.617 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.372 5.397 qs_energies_init_hamiltonians 11 5.9 0.042 0.062 5.305 5.325 mp_alltoall_d11v 2415 14.1 4.535 4.912 4.535 4.912 mp_allgather_i34 2507 14.6 1.153 4.558 1.153 4.558 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.287 4.346 potential_pw2rs 128 12.3 0.026 0.027 3.958 3.964 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.594 3.906 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.369 3.693 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.558 3.637 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.561 3.584 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.423 3.539 multiply_cannon_metrocomm4 7521 15.6 0.025 0.027 1.701 3.495 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.463 3.473 calculate_first_density_matrix 1 7.0 0.000 0.000 3.444 3.446 mp_irecv_dv 28860 15.9 1.663 3.428 1.663 3.428 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=169.199000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1464.363636, yerr=19.850682 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 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 3.104412E+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 11.342275E+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 716108638608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 59207. MP_Allreduce 11070 1504. MP_Sync 86 MP_Alltoall 1700 36954383. 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 0.070 0.074 300.129 300.129 qs_mol_dyn_low 1 2.0 0.003 0.003 299.038 299.057 qs_forces 11 3.9 0.004 0.004 298.929 298.932 qs_energies 11 4.9 0.002 0.007 289.988 289.993 scf_env_do_scf 11 5.9 0.001 0.001 266.286 266.297 velocity_verlet 10 3.0 0.018 0.019 214.205 214.213 scf_env_do_scf_inner_loop 116 6.6 0.010 0.015 139.108 139.110 init_scf_loop 11 6.9 0.000 0.000 126.877 126.879 prepare_preconditioner 11 7.9 0.000 0.000 122.033 122.062 make_preconditioner 11 8.9 0.000 0.000 122.033 122.062 make_full_inverse_cholesky 11 9.9 0.038 0.039 98.083 119.119 qs_scf_new_mos 116 7.6 0.001 0.001 95.444 95.623 qs_scf_loop_do_ot 116 8.6 0.001 0.001 95.443 95.622 ot_scf_mini 116 9.6 0.004 0.004 90.408 90.561 dbcsr_multiply_generic 2485 12.5 0.217 0.230 85.291 86.295 cp_fm_upper_to_full 104 14.8 52.536 75.519 52.536 75.519 multiply_cannon 2485 13.5 0.677 0.710 60.616 61.358 multiply_cannon_loop 2485 14.5 1.034 1.071 56.230 58.747 ot_mini 116 10.6 0.001 0.001 46.504 46.691 dbcsr_complete_redistribute 393 12.7 3.982 4.024 30.080 42.719 copy_fm_to_dbcsr 208 11.6 0.001 0.001 26.456 39.013 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.907 36.357 mp_waitall_1 102768 16.8 29.863 36.122 29.863 36.122 cp_fm_cholesky_invert 11 10.9 35.554 35.562 35.554 35.562 rebuild_ks_matrix 127 8.3 0.001 0.001 34.279 34.511 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 34.278 34.511 mp_alltoall_i22 712 14.1 21.696 34.211 21.696 34.211 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.863 32.074 qs_ot_get_p 127 10.4 0.002 0.002 28.489 28.597 qs_ot_get_derivative 116 11.6 0.002 0.002 26.265 26.406 qs_ot_p2m_diag 82 11.4 0.869 0.873 24.046 24.094 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 20.078 22.624 cp_dbcsr_syevd 82 12.4 0.006 0.006 22.117 22.120 make_m2s 4970 13.5 0.074 0.076 19.347 21.153 make_images 4970 14.5 3.032 3.225 18.869 20.674 ot_diis_step 116 11.6 0.022 0.025 20.160 20.160 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.299 19.681 apply_single 127 13.6 0.001 0.001 19.299 19.680 cp_fm_diag_elpa 82 13.4 0.000 0.000 18.735 18.736 cp_fm_diag_elpa_base 82 14.4 14.312 15.990 18.730 18.730 multiply_cannon_multrec 9940 15.5 10.265 12.059 18.059 18.184 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.354 16.378 calculate_rho_elec 127 8.7 0.478 0.479 16.353 16.377 sum_up_and_integrate 127 10.3 0.002 0.002 16.039 16.133 integrate_v_rspace 127 11.3 0.004 0.004 15.976 16.071 multiply_cannon_sync_h2d 9940 15.5 14.213 14.233 14.213 14.233 make_images_data 4970 15.5 0.062 0.067 10.888 13.692 hybrid_alltoall_any 5155 16.4 1.304 3.046 10.811 13.557 init_scf_run 11 5.9 0.000 0.001 13.137 13.138 scf_env_initial_rho_setup 11 6.9 0.004 0.006 13.137 13.138 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.688 10.825 fft_wrap_pw1pw2 1281 11.7 0.021 0.021 10.373 10.387 cp_fm_cholesky_decompose 22 10.9 9.831 9.850 9.831 9.850 dbcsr_mm_accdrv_process 20590 16.0 4.195 5.922 7.544 9.463 wfi_extrapolate 11 7.9 0.001 0.001 9.284 9.284 fft_wrap_pw1pw2_140 519 12.2 0.817 0.819 9.089 9.107 grid_integrate_task_list 127 12.3 8.496 8.685 8.496 8.685 qs_energies_init_hamiltonians 11 5.9 0.060 0.097 8.271 8.307 fft3d_ps 1281 13.7 3.961 3.977 7.694 7.722 mp_alltoall_d11v 2401 14.1 7.298 7.439 7.298 7.439 density_rs2pw 127 9.7 0.005 0.005 7.330 7.378 calculate_dm_sparse 127 9.5 0.001 0.001 6.905 6.973 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.811 6.860 grid_collocate_task_list 127 9.7 6.366 6.440 6.366 6.440 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.312 6.387 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=300.129000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2770.454545, yerr=172.068360 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 1.268158E+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 29299874. 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 0.022 0.077 85.701 85.702 qs_energies 1 2.0 0.000 0.000 84.715 84.735 ls_scf 1 3.0 0.000 0.000 83.795 83.815 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.697 72.848 multiply_cannon 111 7.7 0.017 0.022 55.849 57.165 multiply_cannon_loop 111 8.7 0.226 0.242 52.446 53.803 ls_scf_main 1 4.0 0.000 0.000 52.387 52.387 density_matrix_trs4 2 5.0 0.002 0.003 46.837 46.909 ls_scf_init_scf 1 4.0 0.000 0.000 28.365 28.367 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.225 27.283 mp_waitall_1 11031 10.9 22.530 25.250 22.530 25.250 matrix_sqrt_Newton_Schulz 2 6.5 0.010 0.014 25.090 25.109 multiply_cannon_multrec 2664 9.7 8.132 8.880 15.647 17.370 multiply_cannon_sync_h2d 2664 9.7 13.435 14.531 13.435 14.531 make_m2s 222 7.7 0.008 0.011 13.242 13.709 make_images 222 8.7 0.099 0.111 13.220 13.689 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.673 13.144 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.464 8.706 make_images_data 222 9.7 0.004 0.005 7.742 8.301 hybrid_alltoall_any 227 10.6 0.216 1.834 6.712 8.142 dbcsr_mm_accdrv_process 4760 10.4 0.598 0.713 7.134 8.109 dbcsr_mm_accdrv_process_sort 4760 11.4 6.337 7.234 6.337 7.234 calculate_norms 4752 9.8 5.513 6.191 5.513 6.191 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.018 5.138 mp_sum_l 887 5.1 3.101 4.428 3.101 4.428 multiply_cannon_metrocomm4 2442 9.7 0.011 0.015 2.042 3.996 mp_irecv_dv 6231 10.9 2.026 3.961 2.026 3.961 make_images_sizes 222 9.7 0.000 0.000 0.758 3.729 mp_alltoall_i44 222 10.7 0.757 3.728 0.757 3.728 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.313 3.282 arnoldi_extremal 4 6.8 0.000 0.000 3.235 3.255 arnoldi_normal_ev 4 7.8 0.033 0.043 3.235 3.255 build_subspace 16 8.4 0.009 0.012 3.106 3.107 ls_scf_post 1 4.0 0.000 0.000 3.043 3.064 dbcsr_special_finalize 555 9.7 0.005 0.006 2.387 2.916 dbcsr_merge_single_wm 555 10.7 0.457 0.587 2.378 2.906 ls_scf_store_result 1 5.0 0.000 0.000 2.853 2.900 dbcsr_sort_data 658 11.4 2.178 2.650 2.178 2.650 make_images_pack 222 9.7 2.209 2.626 2.211 2.628 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.311 2.546 dbcsr_matrix_vector_mult_local 304 10.0 2.066 2.450 2.068 2.452 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.284 2.352 buffer_matrices_ensure_size 222 8.7 1.755 2.050 1.755 2.050 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.808 1.810 rebuild_ks_matrix 3 7.3 0.000 0.000 1.797 1.799 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.797 1.799 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.702000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 2.158617E+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 0.057 0.121 91.906 91.908 qs_energies 1 2.0 0.002 0.013 90.757 90.797 ls_scf 1 3.0 0.000 0.000 89.231 89.268 dbcsr_multiply_generic 111 6.7 0.016 0.020 75.006 75.402 multiply_cannon 111 7.7 0.027 0.040 52.976 57.414 ls_scf_main 1 4.0 0.001 0.012 55.240 55.248 multiply_cannon_loop 111 8.7 0.134 0.148 50.206 53.790 density_matrix_trs4 2 5.0 0.003 0.018 49.439 49.646 mp_waitall_1 9105 10.9 21.448 31.637 21.448 31.637 ls_scf_init_scf 1 4.0 0.000 0.001 30.273 30.318 ls_scf_init_matrix_S 1 5.0 0.000 0.004 29.007 29.099 multiply_cannon_multrec 1332 9.7 13.241 16.986 22.651 27.592 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.623 26.634 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.730 21.709 make_m2s 222 7.7 0.006 0.008 15.473 16.014 make_images 222 8.7 1.370 1.695 15.442 15.984 dbcsr_mm_accdrv_process 4041 10.4 0.333 0.519 9.006 10.545 make_images_data 222 9.7 0.004 0.005 9.152 10.099 dbcsr_mm_accdrv_process_sort 4041 11.4 8.525 10.017 8.525 10.017 hybrid_alltoall_any 227 10.6 0.540 2.572 8.512 9.664 mp_sum_l 887 5.1 5.198 8.128 5.198 8.128 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.200 7.877 mp_irecv_dv 3311 11.0 3.180 7.826 3.180 7.826 calculate_norms 2376 9.8 6.035 6.837 6.035 6.837 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.846 6.408 multiply_cannon_sync_h2d 1332 9.7 4.879 6.203 4.879 6.203 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.959 5.187 arnoldi_extremal 4 6.8 0.000 0.000 4.859 4.880 arnoldi_normal_ev 4 7.8 0.001 0.004 4.859 4.880 build_subspace 16 8.4 0.014 0.021 4.606 4.610 ls_scf_post 1 4.0 0.001 0.008 3.718 3.737 ls_scf_store_result 1 5.0 0.000 0.000 3.435 3.541 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.210 3.475 dbcsr_matrix_vector_mult_local 304 10.0 2.748 3.227 2.750 3.228 mp_allgather_i34 111 8.7 0.751 3.092 0.751 3.092 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.664 2.776 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.203 2.574 dbcsr_data_new 4174 10.1 2.119 2.418 2.119 2.418 make_images_pack 222 9.7 1.819 2.135 1.822 2.137 dbcsr_sort_data 436 11.2 1.794 2.095 1.794 2.095 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.937 1.964 rebuild_ks_matrix 3 7.3 0.000 0.000 1.924 1.951 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.016 1.924 1.951 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.908000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1826.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 2.970956E+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 0.065 0.093 94.367 94.368 qs_energies 1 2.0 0.000 0.000 92.925 92.929 ls_scf 1 3.0 0.000 0.000 91.501 91.505 dbcsr_multiply_generic 111 6.7 0.016 0.016 76.112 76.425 multiply_cannon 111 7.7 0.039 0.117 52.781 57.830 ls_scf_main 1 4.0 0.000 0.000 56.982 56.987 multiply_cannon_loop 111 8.7 0.116 0.129 50.020 53.680 density_matrix_trs4 2 5.0 0.002 0.003 50.932 51.113 mp_waitall_1 7281 11.0 24.504 34.361 24.504 34.361 ls_scf_init_scf 1 4.0 0.000 0.000 30.754 30.762 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.533 29.626 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.172 27.183 multiply_cannon_multrec 888 9.7 12.601 14.985 21.325 24.526 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.655 23.463 make_m2s 222 7.7 0.006 0.007 16.614 17.346 make_images 222 8.7 1.586 1.850 16.576 17.307 make_images_data 222 9.7 0.004 0.005 9.832 10.881 hybrid_alltoall_any 227 10.6 0.637 2.942 9.410 10.604 dbcsr_mm_accdrv_process 3754 10.4 0.312 0.530 8.237 9.510 dbcsr_mm_accdrv_process_sort 3754 11.4 7.789 8.979 7.789 8.979 mp_sum_l 887 5.1 5.024 8.615 5.024 8.615 multiply_cannon_sync_h2d 888 9.7 5.994 7.777 5.994 7.777 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.500 7.043 mp_irecv_dv 2335 11.1 2.484 6.987 2.484 6.987 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.865 6.809 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.655 6.555 arnoldi_extremal 4 6.8 0.000 0.000 5.117 5.135 arnoldi_normal_ev 4 7.8 0.001 0.005 5.117 5.135 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.874 5.114 build_subspace 16 8.4 0.014 0.020 4.823 4.829 calculate_norms 1584 9.8 4.396 4.729 4.396 4.729 mp_allgather_i34 111 8.7 0.860 3.799 0.860 3.799 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.465 3.788 ls_scf_post 1 4.0 0.000 0.000 3.765 3.771 dbcsr_matrix_vector_mult_local 304 10.0 3.030 3.611 3.032 3.613 ls_scf_store_result 1 5.0 0.000 0.000 3.493 3.581 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.906 3.027 dbcsr_data_new 4116 9.9 2.110 2.481 2.110 2.481 make_images_sizes 222 9.7 0.000 0.000 0.984 2.357 mp_alltoall_i44 222 10.7 0.983 2.356 0.983 2.356 dbcsr_sort_data 325 11.1 1.845 2.061 1.845 2.061 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.956 1.958 rebuild_ks_matrix 3 7.3 0.000 0.000 1.938 1.940 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.938 1.940 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.368000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2281.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 3.376300E+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 0.028 0.047 99.114 99.115 qs_energies 1 2.0 0.000 0.000 98.215 98.252 ls_scf 1 3.0 0.000 0.000 96.511 96.550 dbcsr_multiply_generic 111 6.7 0.016 0.018 79.965 80.268 ls_scf_main 1 4.0 0.000 0.000 60.160 60.161 multiply_cannon 111 7.7 0.045 0.079 52.477 56.909 density_matrix_trs4 2 5.0 0.002 0.003 53.951 54.082 multiply_cannon_loop 111 8.7 0.153 0.172 47.419 50.603 ls_scf_init_scf 1 4.0 0.000 0.000 33.042 33.043 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.822 31.903 mp_waitall_1 6369 11.0 23.785 29.753 23.785 29.753 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 29.309 29.320 multiply_cannon_multrec 1332 9.7 14.071 17.630 22.090 24.596 make_m2s 222 7.7 0.006 0.007 21.469 22.819 make_images 222 8.7 3.148 3.597 21.419 22.771 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.701 17.532 make_images_data 222 9.7 0.004 0.004 12.094 13.731 hybrid_alltoall_any 227 10.6 0.793 3.754 11.496 13.256 dbcsr_mm_accdrv_process 3641 10.4 0.288 0.490 7.661 9.225 dbcsr_mm_accdrv_process_sort 3641 11.4 7.219 8.726 7.219 8.726 mp_sum_l 887 5.1 4.445 8.078 4.445 8.078 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.473 6.383 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.095 6.222 mp_irecv_dv 3229 10.9 2.070 6.122 2.070 6.122 multiply_cannon_sync_h2d 1332 9.7 5.385 6.023 5.385 6.023 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.813 5.531 arnoldi_extremal 4 6.8 0.000 0.000 5.395 5.414 arnoldi_normal_ev 4 7.8 0.003 0.006 5.395 5.414 build_subspace 16 8.4 0.014 0.021 5.045 5.054 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.624 4.821 calculate_norms 2376 9.8 4.210 4.558 4.210 4.558 mp_allgather_i34 111 8.7 2.160 4.419 2.160 4.419 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 3.659 3.987 dbcsr_matrix_vector_mult_local 304 10.0 3.210 3.715 3.212 3.717 dbcsr_sort_data 658 11.4 3.068 3.358 3.068 3.358 ls_scf_post 1 4.0 0.000 0.000 3.310 3.347 dbcsr_special_finalize 555 9.7 0.006 0.007 2.822 3.136 dbcsr_merge_single_wm 555 10.7 0.539 0.647 2.814 3.128 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.039 3.116 ls_scf_store_result 1 5.0 0.000 0.000 3.025 3.099 dbcsr_data_release 10477 10.7 1.568 2.465 1.568 2.465 dbcsr_finalize 304 7.8 0.049 0.061 1.804 2.026 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=99.115000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2775.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 4.818690E+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 0.079 0.083 92.522 92.522 qs_energies 1 2.0 0.000 0.000 91.429 91.435 ls_scf 1 3.0 0.000 0.000 89.471 89.478 dbcsr_multiply_generic 111 6.7 0.017 0.018 71.001 71.168 ls_scf_main 1 4.0 0.000 0.000 56.682 56.683 multiply_cannon 111 7.7 0.083 0.144 52.794 56.070 multiply_cannon_loop 111 8.7 0.093 0.142 50.253 51.305 density_matrix_trs4 2 5.0 0.002 0.003 49.728 49.818 ls_scf_init_scf 1 4.0 0.000 0.000 29.375 29.380 mp_waitall_1 5436 11.0 24.485 29.151 24.485 29.151 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.153 28.181 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.042 26.055 multiply_cannon_multrec 444 9.7 13.754 16.951 20.938 22.615 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.622 15.419 make_m2s 222 7.7 0.005 0.005 13.489 14.401 make_images 222 8.7 2.040 2.482 13.422 14.332 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.473 14.182 make_images_data 222 9.7 0.003 0.004 8.207 9.722 hybrid_alltoall_any 227 10.6 0.802 3.821 8.188 9.560 dbcsr_mm_accdrv_process 3003 10.4 0.351 0.403 6.874 8.079 multiply_cannon_sync_h2d 444 9.7 6.732 7.977 6.732 7.977 dbcsr_mm_accdrv_process_sort 3003 11.4 6.512 7.676 6.512 7.676 arnoldi_extremal 4 6.8 0.000 0.000 5.905 5.919 arnoldi_normal_ev 4 7.8 0.002 0.005 5.905 5.919 build_subspace 16 8.4 0.015 0.020 5.505 5.513 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.453 4.688 mp_sum_l 887 5.1 2.718 4.475 2.718 4.475 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.226 4.412 dbcsr_matrix_vector_mult_local 304 10.0 3.733 4.205 3.735 4.207 calculate_norms 792 9.8 3.604 3.678 3.604 3.678 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.913 3.648 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.534 3.639 mp_irecv_dv 1241 11.2 1.520 3.612 1.520 3.612 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.469 3.543 mp_allgather_i34 111 8.7 1.120 3.522 1.120 3.522 ls_scf_post 1 4.0 0.000 0.000 3.414 3.420 make_images_sizes 222 9.7 0.000 0.000 0.892 3.336 mp_alltoall_i44 222 10.7 0.892 3.335 0.892 3.335 ls_scf_store_result 1 5.0 0.000 0.000 3.203 3.238 dbcsr_finalize 304 7.8 0.062 0.077 2.204 2.269 dbcsr_data_new 4608 9.7 1.793 2.245 1.793 2.245 dbcsr_merge_all 275 8.9 0.480 0.523 2.060 2.108 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.048 2.049 rebuild_ks_matrix 3 7.3 0.000 0.000 2.013 2.014 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.013 2.014 qs_energies_init_hamiltonians 1 3.0 0.002 0.003 1.943 1.943 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.522000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3791.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 9.012724E+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 0.201 0.216 106.735 106.736 qs_energies 1 2.0 0.000 0.000 104.600 104.615 ls_scf 1 3.0 0.000 0.000 101.566 101.582 dbcsr_multiply_generic 111 6.7 0.023 0.027 75.394 75.552 ls_scf_main 1 4.0 0.000 0.000 64.134 64.134 density_matrix_trs4 2 5.0 0.002 0.003 55.321 55.387 multiply_cannon 111 7.7 0.119 0.199 48.787 51.043 multiply_cannon_loop 111 8.7 0.098 0.099 45.836 46.329 ls_scf_init_scf 1 4.0 0.000 0.000 33.625 33.626 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.201 32.229 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.502 29.508 mp_waitall_1 4527 11.1 22.446 25.703 22.446 25.703 make_m2s 222 7.7 0.005 0.005 22.829 23.871 make_images 222 8.7 3.573 3.871 22.722 23.762 multiply_cannon_multrec 444 9.7 17.753 18.371 22.468 23.122 hybrid_alltoall_any 227 10.6 1.652 3.627 12.954 15.949 make_images_data 222 9.7 0.003 0.004 13.209 15.307 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.622 10.940 multiply_cannon_sync_h2d 444 9.7 8.779 8.833 8.779 8.833 arnoldi_extremal 4 6.8 0.000 0.000 7.510 7.519 arnoldi_normal_ev 4 7.8 0.003 0.009 7.509 7.519 build_subspace 16 8.4 0.026 0.036 6.966 6.978 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.597 5.747 dbcsr_matrix_vector_mult_local 304 10.0 5.111 5.424 5.113 5.426 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.870 5.110 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.933 5.026 dbcsr_mm_accdrv_process 1814 10.4 0.244 0.359 4.529 4.649 dbcsr_mm_accdrv_process_sort 1814 11.4 4.194 4.325 4.194 4.325 ls_scf_post 1 4.0 0.000 0.000 3.808 3.823 make_images_sizes 222 9.7 0.000 0.000 1.475 3.617 mp_alltoall_i44 222 10.7 1.475 3.617 1.475 3.617 ls_scf_store_result 1 5.0 0.000 0.000 3.520 3.559 mp_allgather_i34 111 8.7 1.054 3.330 1.054 3.330 calculate_norms 792 9.8 3.242 3.323 3.242 3.323 dbcsr_finalize 304 7.8 0.082 0.090 3.081 3.167 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 3.003 3.003 dbcsr_merge_all 275 8.9 0.886 0.918 2.864 2.944 dbcsr_complete_redistribute 5 7.6 1.430 1.479 2.751 2.861 matrix_ls_to_qs 2 6.0 0.000 0.000 2.403 2.511 dbcsr_sort_data 325 11.1 2.437 2.499 2.437 2.499 dbcsr_data_new 6591 9.6 1.910 2.436 1.910 2.436 dbcsr_new_transposed 4 7.5 0.243 0.250 2.285 2.297 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.294 2.296 rebuild_ks_matrix 3 7.3 0.000 0.000 2.227 2.229 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.227 2.229 dbcsr_frobenius_norm 74 6.6 2.054 2.131 2.196 2.225 dbcsr_add_d 103 6.2 0.000 0.000 2.142 2.220 dbcsr_add_anytype 103 7.2 0.858 0.890 2.142 2.220 dbcsr_data_release 12724 10.6 1.974 2.177 1.974 2.177 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.736000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7198.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9fa292db88c5773cf53ea201dab2b0cc1895d578_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 590.061568E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 4037197. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.128 252.158 252.160 qs_mol_dyn_low 1 2.0 0.016 0.124 250.799 250.862 qs_forces 5 3.8 0.004 0.005 250.397 250.403 qs_energies 5 4.8 0.004 0.104 247.246 247.272 scf_env_do_scf 5 5.8 0.000 0.001 231.891 231.894 scf_env_do_scf_inner_loop 105 6.6 0.003 0.026 202.920 202.921 qs_scf_new_mos 105 7.6 0.000 0.001 159.708 159.840 qs_scf_loop_do_ot 105 8.6 0.001 0.001 159.707 159.840 ot_scf_mini 105 9.6 0.003 0.004 149.771 149.910 dbcsr_multiply_generic 1445 12.2 0.129 0.137 135.684 136.232 multiply_cannon 1445 13.2 0.274 0.292 115.094 117.679 multiply_cannon_loop 1445 14.2 2.837 2.976 113.044 114.550 velocity_verlet 4 3.0 0.016 0.073 109.684 109.695 ot_mini 105 10.6 0.001 0.012 62.272 62.384 qs_ot_get_p 112 10.4 0.001 0.018 57.272 57.625 qs_ot_p2m_diag 40 11.0 0.020 0.031 45.619 45.730 multiply_cannon_multrec 69360 15.2 29.676 34.325 39.627 44.523 mp_waitall_1 488190 16.1 37.122 44.517 37.122 44.517 cp_dbcsr_syevd 40 12.0 0.002 0.003 42.065 42.067 qs_ot_get_derivative 55 11.6 0.001 0.001 40.454 40.635 cp_fm_syevd 40 13.0 0.000 0.000 36.370 36.532 multiply_cannon_metrocomm3 69360 15.2 0.201 0.216 26.204 34.531 multiply_cannon_sync_h2d 69360 15.2 29.191 33.313 29.191 33.313 cp_fm_redistribute_end 40 14.0 15.713 31.361 15.721 31.364 cp_fm_syevd_base 40 14.0 15.635 31.287 15.635 31.287 rebuild_ks_matrix 110 8.4 0.000 0.000 29.972 30.154 qs_ks_build_kohn_sham_matrix 110 9.4 0.013 0.066 29.972 30.154 init_scf_loop 7 6.6 0.002 0.048 28.935 28.937 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.572 27.747 prepare_preconditioner 7 7.6 0.000 0.000 23.778 23.824 make_preconditioner 7 8.6 0.002 0.053 23.778 23.824 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.070 23.294 apply_single 62 13.6 0.000 0.000 23.070 23.294 ot_new_cg_direction 55 11.6 0.001 0.001 21.065 21.067 qs_rho_update_rho_low 110 7.6 0.000 0.001 19.449 19.810 calculate_rho_elec 110 8.6 0.029 0.042 19.449 19.810 make_full_inverse_cholesky 7 9.6 0.000 0.000 16.167 16.228 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.928 15.113 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.150 14.237 density_rs2pw 110 9.6 0.004 0.005 13.619 14.008 mp_sum_l 4764 12.2 12.787 13.706 12.787 13.706 init_scf_run 5 5.8 0.000 0.004 12.459 12.460 scf_env_initial_rho_setup 5 6.8 0.001 0.002 12.459 12.460 calculate_dm_sparse 110 9.5 0.000 0.001 11.555 11.725 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 11.502 11.575 cp_fm_cholesky_invert 7 10.6 11.346 11.354 11.346 11.354 fft_wrap_pw1pw2 1425 12.5 0.019 0.025 10.826 11.081 transfer_rs2pw 445 10.6 0.008 0.009 10.502 10.994 dbcsr_mm_accdrv_process 154766 15.8 6.259 6.474 9.817 10.636 check_diag 80 13.5 8.707 8.956 10.142 10.268 qs_vxc_create 110 10.4 0.002 0.005 9.868 9.898 sum_up_and_integrate 60 10.3 0.002 0.021 9.789 9.811 integrate_v_rspace 60 11.3 0.002 0.002 9.771 9.793 fft_wrap_pw1pw2_240 915 14.0 0.201 0.223 9.367 9.598 fft3d_pb 915 15.0 2.255 2.454 8.319 8.581 multiply_cannon_metrocomm1 69360 15.2 0.099 0.104 5.073 8.035 make_m2s 2890 13.2 0.078 0.087 7.362 7.903 make_images 2890 14.2 0.237 0.258 7.254 7.796 acc_transpose_blocks 69360 15.2 0.349 0.365 7.026 7.733 calculate_first_density_matrix 1 7.0 0.001 0.031 7.374 7.388 make_full_single_inverse 7 9.6 0.005 0.047 7.316 7.364 cp_dbcsr_sm_fm_multiply 15 9.3 0.004 0.036 7.210 7.229 xc_rho_set_and_dset_create 110 12.4 0.072 0.077 6.987 7.218 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.739 6.782 xc_vxc_pw_create 60 11.3 0.036 0.040 6.603 6.637 potential_pw2rs 60 12.3 0.002 0.003 5.968 6.045 mp_alltoall_z22v 2340 16.7 5.477 5.780 5.477 5.780 xc_pw_derive 510 13.4 0.005 0.006 5.714 5.767 transfer_rs2pw_30 110 11.6 1.281 1.355 4.606 5.446 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=252.160000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=559.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 9fa292db88c5773cf53ea201dab2b0cc1895d578 Summary: empty Status: OK