=== 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: 15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c ################# 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, SIRIUS 7.4.3, SPGLIB 1.16.2 # # 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: 04.10.2023 # # \ 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-libvdwxc --with-pexsi --with-plumed; \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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_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_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/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/01 job id: 49847712 --- 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/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/02 job id: 49847713 --- 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] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/03 job id: 49847714 --- 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] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/04 job id: 49847716 --- 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] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/05 job id: 49847717 --- 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] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/06 job id: 49847718 --- 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] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/07 job id: 49847720 --- 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] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/08 job id: 49847721 --- 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/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/09 job id: 49847723 --- 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/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/10 job id: 49847725 --- 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/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/11 job id: 49847727 --- 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/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/12 job id: 49847728 --- 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/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/13 job id: 49847730 --- 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/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/14 job id: 49847731 --- 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/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/15 job id: 49847733 --- 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/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/16 job id: 49847735 --- 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/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/17 job id: 49847736 --- 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/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/18 job id: 49847737 --- 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/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/19 job id: 49847738 --- 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/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/20 job id: 49847740 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/21 job id: 49847741 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/22 job id: 49847743 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/23 job id: 49847744 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/24 job id: 49847745 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/25 job id: 49847747 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/26 job id: 49847748 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/27 job id: 49847749 --- 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/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_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 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.035 134.717 134.718 farming_run 1 2.0 133.908 133.911 134.673 134.677 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.508590E+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 0.007 0.025 114.539 114.540 qs_energies 1 2.0 0.000 0.000 114.349 114.352 mp2_main 1 3.0 0.000 0.000 112.163 112.166 mp2_gpw_main 1 4.0 0.029 0.035 111.084 111.086 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 92.180 92.471 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.758 55.049 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.164 41.168 46.195 get_2c_integrals 1 6.0 0.008 0.009 36.534 37.250 integrate_v_rspace 273 8.0 0.435 0.449 24.722 29.554 pw_transfer 6555 10.6 0.379 0.398 26.768 27.279 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 25.432 25.961 grid_integrate_task_list 273 9.0 20.615 25.927 20.615 25.927 fft_wrap_pw1pw2_100 2178 12.4 1.123 1.290 22.987 23.495 rpa_ri_compute_en 1 5.0 0.020 0.023 18.797 19.114 compute_2c_integrals 1 7.0 0.002 0.002 18.890 18.891 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.467 18.565 mp2_eri_2c_integrate_gpw 1 9.0 2.382 2.433 18.465 18.562 cp_fm_cholesky_decompose 12 8.2 17.628 18.338 17.628 18.338 cholesky_decomp 1 7.0 0.000 0.000 16.478 17.203 fft3d_s 5443 13.4 16.095 16.449 16.118 16.472 ao_to_mo_and_store_B_mult_1 272 7.0 10.766 15.314 10.766 15.314 calculate_wavefunction 272 8.0 5.421 5.571 12.367 13.029 rpa_num_int 1 6.0 0.000 0.001 10.692 10.692 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.673 10.680 calc_mat_Q 8 8.0 0.000 0.000 9.480 9.573 contract_S_to_Q 8 9.0 0.000 0.000 8.900 8.995 calc_potential_gpw 544 9.5 0.005 0.005 8.190 8.614 parallel_gemm_fm 14 9.1 0.000 0.000 8.485 8.601 parallel_gemm_fm_cosma 14 10.1 8.485 8.601 8.485 8.601 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.007 8.231 potential_pw2rs 545 10.0 0.107 0.109 7.463 8.078 create_integ_mat 1 6.0 0.028 0.037 7.722 7.722 collocate_single_gaussian 272 10.0 0.040 0.042 7.273 7.488 array2fm 1 7.0 0.000 0.000 6.554 7.073 pw_scatter_s 2720 13.7 4.363 4.533 4.363 4.533 pw_gather_s 2722 13.2 3.454 3.790 3.454 3.790 array2fm_buffer_send 1 8.0 2.875 3.059 2.875 3.059 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.086246, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2825.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_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 205321. 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 0.023 0.037 421.628 421.629 farming_run 1 2.0 420.517 420.521 421.594 421.596 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.235190E+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 30146148. 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 0.013 0.031 216.240 216.241 qs_energies 1 2.0 0.000 0.000 216.013 216.019 scf_env_do_scf 1 3.0 0.000 0.000 114.241 114.241 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.292 113.299 rebuild_ks_matrix 4 6.0 0.000 0.000 113.291 113.297 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.062 113.291 113.297 hfx_ks_matrix 4 8.0 0.001 0.001 112.941 112.946 integrate_four_center 4 9.0 0.153 0.466 112.940 112.945 integrate_four_center_main 4 10.0 0.111 0.743 101.582 104.776 integrate_four_center_bin 262 11.0 101.471 104.774 101.471 104.774 mp2_main 1 3.0 0.000 0.000 101.479 101.485 mp2_gpw_main 1 4.0 0.045 0.067 100.554 100.566 init_scf_loop 1 4.0 0.000 0.000 96.461 96.462 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 73.898 74.925 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.716 54.743 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.159 41.532 46.316 integrate_v_rspace 95 8.0 0.394 0.560 28.010 32.679 pw_transfer 2240 10.6 0.143 0.159 29.535 29.928 fft_wrap_pw1pw2 1868 11.4 0.017 0.019 28.538 28.960 ao_to_mo_and_store_B_mult_1 91 7.0 10.504 28.511 10.504 28.511 grid_integrate_task_list 95 9.0 23.438 28.286 23.438 28.286 mp2_ri_gpw_compute_en 1 5.0 0.055 0.062 26.510 28.037 fft_wrap_pw1pw2_100 730 12.4 1.313 1.462 26.286 26.671 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.839 1.891 24.824 24.834 get_2c_integrals 1 6.0 0.000 0.000 20.088 20.117 compute_2c_integrals 1 7.0 0.002 0.003 19.069 19.074 fft3d_s 1823 13.4 18.496 18.933 18.510 18.946 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.708 18.937 mp2_eri_2c_integrate_gpw 1 9.0 1.725 1.894 18.707 18.937 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.778 17.779 calculate_wavefunction 91 8.0 2.018 2.047 9.619 9.855 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.582 8.781 9.312 potential_pw2rs 186 10.0 0.033 0.034 8.461 8.951 local_gemm 172 8.0 8.223 8.733 8.223 8.733 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.206 8.588 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.517 7.826 8.244 calc_potential_gpw 182 9.5 0.002 0.002 7.908 8.059 collocate_single_gaussian 91 10.0 0.017 0.022 7.784 8.034 integrate_four_center_load 4 10.0 0.000 0.000 6.755 6.759 hfx_load_balance 1 11.0 0.000 0.000 6.755 6.758 mp2_ri_gpw_compute_en_ener 172 7.0 6.333 6.400 6.333 6.400 mp_sendrecv_dm3 2068 8.0 5.861 6.273 5.861 6.273 mp_sync 37 10.5 3.342 5.847 3.342 5.847 pw_gather_s 912 13.2 4.469 4.980 4.469 4.980 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.550922, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1495.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_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 452.038656E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.007 0.029 55.149 55.151 qs_mol_dyn_low 1 2.0 0.003 0.003 54.943 54.951 qs_forces 11 3.9 0.002 0.002 54.788 54.789 qs_energies 11 4.9 0.001 0.001 53.285 53.304 scf_env_do_scf 11 5.9 0.001 0.001 46.782 46.782 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.631 44.632 dbcsr_multiply_generic 2286 12.5 0.093 0.097 34.136 34.653 qs_scf_new_mos 108 7.5 0.000 0.001 34.298 34.575 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.298 34.574 ot_scf_mini 108 9.5 0.002 0.002 32.654 32.845 multiply_cannon 2286 13.5 0.192 0.203 26.816 28.473 multiply_cannon_loop 2286 14.5 1.823 1.934 26.072 27.748 velocity_verlet 10 3.0 0.001 0.002 26.707 26.708 ot_mini 108 10.5 0.001 0.001 19.491 19.738 qs_ot_get_derivative 108 11.5 0.001 0.001 16.477 16.661 mp_waitall_1 245248 16.5 8.588 15.182 8.588 15.182 multiply_cannon_metrocomm3 54864 15.5 0.073 0.079 6.077 13.138 multiply_cannon_multrec 54864 15.5 3.630 5.653 8.129 11.719 qs_ot_get_p 119 10.4 0.001 0.001 8.502 8.761 rebuild_ks_matrix 119 8.3 0.000 0.000 7.978 8.109 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.012 7.978 8.108 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.027 7.145 mp_sum_l 7287 12.8 5.223 6.899 5.223 6.899 multiply_cannon_sync_h2d 54864 15.5 5.150 6.300 5.150 6.300 dbcsr_mm_accdrv_process 76910 16.1 1.845 2.889 4.412 6.256 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.471 5.940 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.377 5.506 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.474 5.502 init_scf_run 11 5.9 0.000 0.001 5.213 5.214 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.213 5.213 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.670 4.670 sum_up_and_integrate 119 10.3 0.001 0.002 4.571 4.577 integrate_v_rspace 119 11.3 0.002 0.002 4.559 4.567 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.377 4.490 calculate_rho_elec 119 8.7 0.011 0.016 4.377 4.490 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.432 4.433 cp_fm_redistribute_end 50 14.0 2.257 4.401 2.267 4.405 cp_fm_diag_elpa_base 50 14.0 2.131 4.287 2.136 4.297 jit_kernel_multiply 13 15.8 2.503 3.365 2.503 3.365 calculate_dm_sparse 119 9.5 0.000 0.000 3.193 3.331 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.951 3.158 apply_single 119 13.6 0.000 0.000 2.951 3.157 calculate_first_density_matrix 1 7.0 0.000 0.001 2.913 2.919 multiply_cannon_metrocomm1 54864 15.5 0.057 0.063 1.756 2.823 ot_diis_step 108 11.5 0.006 0.006 2.759 2.760 acc_transpose_blocks 54864 15.5 0.231 0.259 2.217 2.742 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.446 2.491 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.432 2.435 density_rs2pw 119 9.7 0.004 0.004 2.346 2.435 wfi_extrapolate 11 7.9 0.001 0.001 2.227 2.227 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.125 2.199 init_scf_loop 11 6.9 0.000 0.000 2.128 2.129 grid_integrate_task_list 119 12.3 2.018 2.120 2.018 2.120 mp_sum_d 4135 12.0 1.367 1.956 1.367 1.956 potential_pw2rs 119 12.3 0.004 0.004 1.931 1.952 pw_transfer 1439 11.6 0.052 0.056 1.806 1.889 make_m2s 4572 13.5 0.053 0.056 1.784 1.826 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.731 1.816 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.720 1.760 make_images 4572 14.5 0.134 0.140 1.702 1.744 mp_alltoall_d11v 2130 13.8 1.424 1.543 1.424 1.543 transfer_rs2pw 487 10.6 0.005 0.006 1.465 1.533 fft3d_ps 1201 14.6 0.372 0.480 1.447 1.528 acc_transpose_blocks_sync 164592 16.5 1.203 1.441 1.203 1.441 mp_waitany 12084 13.8 1.247 1.435 1.247 1.435 grid_collocate_task_list 119 9.7 1.353 1.432 1.353 1.432 transfer_pw2rs 487 13.2 0.006 0.007 1.404 1.424 fft_wrap_pw1pw2_140 487 13.2 0.140 0.153 1.338 1.421 dbcsr_dot_sd 1205 11.9 0.048 0.059 0.734 1.181 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.151000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.363636, yerr=0.881396 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_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 490.795008E+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 1124175. 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 0.009 0.027 42.362 42.363 qs_mol_dyn_low 1 2.0 0.003 0.003 42.170 42.178 qs_forces 11 3.9 0.002 0.002 42.079 42.081 qs_energies 11 4.9 0.001 0.001 40.335 40.340 scf_env_do_scf 11 5.9 0.001 0.001 34.774 34.775 scf_env_do_scf_inner_loop 108 6.5 0.005 0.015 31.989 31.990 dbcsr_multiply_generic 2286 12.5 0.102 0.104 23.529 23.969 qs_scf_new_mos 108 7.5 0.001 0.001 22.501 22.729 qs_scf_loop_do_ot 108 8.5 0.001 0.001 22.501 22.729 ot_scf_mini 108 9.5 0.002 0.003 21.538 21.701 velocity_verlet 10 3.0 0.001 0.001 20.010 20.011 multiply_cannon 2286 13.5 0.210 0.217 17.765 19.306 multiply_cannon_loop 2286 14.5 1.187 1.248 16.457 18.110 ot_mini 108 10.5 0.001 0.001 13.220 13.455 mp_waitall_1 200699 16.5 6.286 11.755 6.286 11.755 qs_ot_get_derivative 108 11.5 0.001 0.001 10.791 10.959 multiply_cannon_metrocomm3 27432 15.5 0.071 0.073 4.451 10.166 multiply_cannon_multrec 27432 15.5 1.851 4.325 6.529 9.597 rebuild_ks_matrix 119 8.3 0.000 0.000 7.245 7.413 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.244 7.412 dbcsr_mm_accdrv_process 47894 16.0 3.683 6.193 4.598 6.938 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.384 6.537 qs_ot_get_p 119 10.4 0.001 0.001 5.285 5.504 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.130 4.955 init_scf_run 11 5.9 0.000 0.001 4.313 4.313 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.313 4.313 mp_sum_l 7287 12.8 2.325 4.243 2.325 4.243 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.165 4.201 calculate_rho_elec 119 8.7 0.021 0.024 4.164 4.200 sum_up_and_integrate 119 10.3 0.001 0.001 4.103 4.110 integrate_v_rspace 119 11.3 0.002 0.003 4.088 4.095 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.003 4.094 apply_single 119 13.6 0.000 0.000 3.003 4.094 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.380 3.406 make_m2s 4572 13.5 0.052 0.054 2.993 3.364 make_images 4572 14.5 0.206 0.247 2.904 3.279 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.946 2.947 init_scf_loop 11 6.9 0.000 0.000 2.754 2.755 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.561 2.647 calculate_first_density_matrix 1 7.0 0.000 0.000 2.580 2.584 density_rs2pw 119 9.7 0.004 0.004 2.443 2.574 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.493 2.494 cp_fm_redistribute_end 50 14.0 1.259 2.446 1.263 2.448 ot_diis_step 108 11.5 0.011 0.011 2.370 2.371 cp_fm_diag_elpa_base 50 14.0 1.148 2.333 1.178 2.369 calculate_dm_sparse 119 9.5 0.000 0.001 2.183 2.264 jit_kernel_multiply 9 16.2 0.856 2.245 0.856 2.245 multiply_cannon_sync_h2d 27432 15.5 1.658 2.234 1.658 2.234 pw_transfer 1439 11.6 0.065 0.069 2.122 2.191 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.184 2.187 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.030 2.102 acc_transpose_blocks 27432 15.5 0.114 0.119 1.630 2.017 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.915 1.963 grid_integrate_task_list 119 12.3 1.841 1.946 1.841 1.946 prepare_preconditioner 11 7.9 0.000 0.000 1.805 1.832 make_preconditioner 11 8.9 0.000 0.000 1.805 1.832 potential_pw2rs 119 12.3 0.006 0.006 1.806 1.826 make_images_data 4572 15.5 0.047 0.053 1.335 1.784 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.699 1.753 fft3d_ps 1201 14.6 0.519 0.575 1.667 1.734 mp_alltoall_d11v 2130 13.8 1.437 1.681 1.437 1.681 wfi_extrapolate 11 7.9 0.001 0.001 1.662 1.662 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.190 1.661 transfer_rs2pw 487 10.6 0.005 0.005 1.453 1.581 fft_wrap_pw1pw2_140 487 13.2 0.158 0.168 1.498 1.561 grid_collocate_task_list 119 9.7 1.288 1.361 1.288 1.361 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.346 1.357 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.286 1.334 transfer_pw2rs 487 13.2 0.004 0.005 1.173 1.196 mp_allgather_i34 2286 14.5 0.728 1.191 0.728 1.191 mp_sum_d 4135 12.0 0.633 1.123 0.633 1.123 make_images_sizes 4572 15.5 0.005 0.005 0.792 1.059 mp_alltoall_i44 4572 16.5 0.788 1.055 0.788 1.055 acc_transpose_blocks_kernels 27432 16.5 0.188 0.280 0.678 0.953 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.949 0.950 acc_transpose_blocks_sync 82296 16.5 0.812 0.940 0.812 0.940 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.922 0.938 mp_alltoall_z22v 1201 16.6 0.852 0.934 0.852 0.934 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=42.363000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=466.181818, yerr=1.799908 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 520.437760E+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 0.014 0.037 38.179 38.181 qs_mol_dyn_low 1 2.0 0.003 0.003 37.941 37.949 qs_forces 11 3.9 0.002 0.002 37.860 37.861 qs_energies 11 4.9 0.001 0.001 36.158 36.166 scf_env_do_scf 11 5.9 0.000 0.001 30.894 30.894 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 27.708 27.709 dbcsr_multiply_generic 2286 12.5 0.100 0.101 19.213 19.501 qs_scf_new_mos 108 7.5 0.001 0.001 18.388 18.406 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.388 18.406 velocity_verlet 10 3.0 0.001 0.001 18.238 18.240 ot_scf_mini 108 9.5 0.002 0.002 17.502 17.515 multiply_cannon 2286 13.5 0.195 0.198 14.682 15.398 multiply_cannon_loop 2286 14.5 0.858 0.898 13.628 14.459 ot_mini 108 10.5 0.001 0.001 10.598 10.624 qs_ot_get_derivative 108 11.5 0.001 0.001 8.893 8.918 multiply_cannon_multrec 18288 15.5 1.851 2.754 7.271 7.524 rebuild_ks_matrix 119 8.3 0.000 0.000 6.924 7.014 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.924 7.013 dbcsr_mm_accdrv_process 38222 16.0 5.222 6.142 5.329 6.209 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.108 6.186 mp_waitall_1 158411 16.6 4.037 5.419 4.037 5.419 qs_ot_get_p 119 10.4 0.001 0.001 4.388 4.428 sum_up_and_integrate 119 10.3 0.001 0.001 4.393 4.397 integrate_v_rspace 119 11.3 0.003 0.003 4.380 4.385 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.218 4.240 calculate_rho_elec 119 8.7 0.031 0.031 4.218 4.240 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.429 4.037 init_scf_run 11 5.9 0.000 0.001 4.014 4.014 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.013 4.013 multiply_cannon_metrocomm3 18288 15.5 0.047 0.048 1.849 3.317 init_scf_loop 11 6.9 0.000 0.000 3.157 3.157 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.897 2.904 density_rs2pw 119 9.7 0.004 0.004 2.700 2.767 make_m2s 4572 13.5 0.044 0.045 2.507 2.734 make_images 4572 14.5 0.194 0.207 2.422 2.649 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.549 2.549 calculate_first_density_matrix 1 7.0 0.000 0.000 2.489 2.489 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.087 2.405 apply_single 119 13.6 0.000 0.000 2.087 2.405 prepare_preconditioner 11 7.9 0.000 0.000 2.275 2.279 make_preconditioner 11 8.9 0.000 0.000 2.274 2.279 pw_transfer 1439 11.6 0.065 0.069 2.182 2.208 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.099 2.196 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.175 2.184 cp_fm_diag_elpa_base 50 14.0 2.147 2.165 2.173 2.182 potential_pw2rs 119 12.3 0.007 0.008 2.139 2.162 mp_sum_l 7287 12.8 1.635 2.118 1.635 2.118 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.089 2.115 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.104 2.111 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.078 2.080 calculate_dm_sparse 119 9.5 0.000 0.000 1.980 1.994 grid_integrate_task_list 119 12.3 1.791 1.903 1.791 1.903 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.850 1.878 acc_transpose_blocks 18288 15.5 0.081 0.083 1.667 1.825 transfer_rs2pw 487 10.6 0.005 0.005 1.676 1.765 fft3d_ps 1201 14.6 0.529 0.545 1.660 1.690 ot_diis_step 108 11.5 0.012 0.012 1.612 1.612 fft_wrap_pw1pw2_140 487 13.2 0.209 0.216 1.515 1.538 transfer_pw2rs 487 13.2 0.004 0.004 1.474 1.495 wfi_extrapolate 11 7.9 0.001 0.001 1.456 1.457 make_images_data 4572 15.5 0.048 0.051 1.093 1.420 grid_collocate_task_list 119 9.7 1.235 1.313 1.235 1.313 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.157 1.229 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.210 1.223 hybrid_alltoall_any 4725 16.4 0.058 0.117 0.920 1.183 mp_alltoall_d11v 2130 13.8 0.990 1.131 0.990 1.131 multiply_cannon_sync_h2d 18288 15.5 0.977 1.110 0.977 1.110 cp_fm_cholesky_invert 11 10.9 1.071 1.078 1.071 1.078 mp_alltoall_z22v 1201 16.6 0.932 1.022 0.932 1.022 acc_transpose_blocks_kernels 18288 16.5 0.218 0.226 0.828 0.961 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.956 0.957 mp_allgather_i34 2286 14.5 0.504 0.949 0.504 0.949 mp_sendrecv_dv 11067 12.7 0.918 0.944 0.918 0.944 make_images_sizes 4572 15.5 0.005 0.005 0.653 0.910 mp_alltoall_i44 4572 16.5 0.648 0.904 0.648 0.904 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.890 0.901 acc_transpose_blocks_sync 54864 16.5 0.738 0.794 0.738 0.794 dbcsr_complete_redistribute 329 12.2 0.093 0.099 0.650 0.769 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=38.181000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.000000, yerr=1.595448 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_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 556.908544E+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 0.008 0.030 37.889 37.891 qs_mol_dyn_low 1 2.0 0.003 0.003 37.407 37.677 qs_forces 11 3.9 0.002 0.002 37.220 37.220 qs_energies 11 4.9 0.001 0.001 35.489 35.493 scf_env_do_scf 11 5.9 0.000 0.001 30.375 30.376 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 26.869 26.869 dbcsr_multiply_generic 2286 12.5 0.099 0.103 20.322 20.446 velocity_verlet 10 3.0 0.001 0.002 18.972 18.975 qs_scf_new_mos 108 7.5 0.001 0.001 18.606 18.660 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.606 18.659 ot_scf_mini 108 9.5 0.002 0.003 17.561 17.618 multiply_cannon 2286 13.5 0.218 0.227 16.376 16.912 multiply_cannon_loop 2286 14.5 1.520 1.624 15.354 15.714 ot_mini 108 10.5 0.001 0.001 10.685 10.746 multiply_cannon_multrec 27432 15.5 2.498 3.165 9.189 9.499 qs_ot_get_derivative 108 11.5 0.001 0.001 8.798 8.854 dbcsr_mm_accdrv_process 47916 15.9 6.317 7.681 6.588 7.976 rebuild_ks_matrix 119 8.3 0.000 0.000 6.438 6.495 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.438 6.494 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.703 5.753 qs_ot_get_p 119 10.4 0.001 0.001 3.977 4.055 init_scf_run 11 5.9 0.000 0.001 3.763 3.763 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.763 3.763 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.248 3.631 sum_up_and_integrate 119 10.3 0.001 0.001 3.594 3.603 integrate_v_rspace 119 11.3 0.003 0.003 3.583 3.592 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.463 3.494 calculate_rho_elec 119 8.7 0.040 0.046 3.463 3.493 init_scf_loop 11 6.9 0.000 0.000 3.482 3.482 prepare_preconditioner 11 7.9 0.000 0.000 2.639 2.647 make_preconditioner 11 8.9 0.000 0.000 2.639 2.647 acc_transpose_blocks 27432 15.5 0.119 0.123 2.402 2.581 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.243 2.571 make_m2s 4572 13.5 0.054 0.056 2.439 2.537 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.462 2.475 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.430 2.459 make_images 4572 14.5 0.274 0.337 2.332 2.427 mp_waitall_1 137007 16.6 1.931 2.385 1.931 2.385 calculate_dm_sparse 119 9.5 0.000 0.000 2.264 2.319 calculate_first_density_matrix 1 7.0 0.000 0.000 2.317 2.318 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.104 2.248 apply_single 119 13.6 0.000 0.000 2.103 2.248 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.046 2.046 pw_transfer 1439 11.6 0.065 0.070 2.007 2.045 density_rs2pw 119 9.7 0.004 0.004 1.924 2.013 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.915 1.956 grid_integrate_task_list 119 12.3 1.832 1.897 1.832 1.897 ot_diis_step 108 11.5 0.012 0.012 1.845 1.845 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.820 1.822 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.756 1.766 cp_fm_diag_elpa_base 50 14.0 1.721 1.739 1.754 1.764 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.636 1.650 fft_wrap_pw1pw2_140 487 13.2 0.245 0.256 1.544 1.587 acc_transpose_blocks_sync 82296 16.5 1.411 1.550 1.411 1.550 fft3d_ps 1201 14.6 0.553 0.603 1.435 1.460 potential_pw2rs 119 12.3 0.008 0.009 1.393 1.397 wfi_extrapolate 11 7.9 0.001 0.001 1.391 1.391 mp_sum_l 7287 12.8 1.081 1.371 1.081 1.371 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.351 1.359 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.324 1.347 grid_collocate_task_list 119 9.7 1.248 1.345 1.248 1.345 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.772 1.290 cp_fm_upper_to_full 72 14.2 0.817 1.162 0.817 1.162 dbcsr_complete_redistribute 329 12.2 0.119 0.153 0.827 1.108 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.078 1.078 transfer_rs2pw 487 10.6 0.004 0.005 0.987 1.078 make_images_data 4572 15.5 0.047 0.051 0.918 1.029 hybrid_alltoall_any 4725 16.4 0.065 0.155 0.790 0.958 acc_transpose_blocks_kernels 27432 16.5 0.270 0.279 0.842 0.914 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.801 0.877 mp_alltoall_d11v 2130 13.8 0.796 0.869 0.796 0.869 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.586 0.855 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.826 0.831 transfer_pw2rs 487 13.2 0.004 0.004 0.809 0.812 cp_fm_cholesky_invert 11 10.9 0.808 0.812 0.808 0.812 mp_alltoall_z22v 1201 16.6 0.765 0.792 0.765 0.792 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.891000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.909091, yerr=3.260077 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_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 624.967680E+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 0.035 31.034 31.042 qs_mol_dyn_low 1 2.0 0.003 0.003 30.862 30.869 qs_forces 11 3.9 0.004 0.006 30.347 30.349 qs_energies 11 4.9 0.001 0.001 28.626 28.629 scf_env_do_scf 11 5.9 0.001 0.001 22.680 22.680 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 20.095 20.097 velocity_verlet 10 3.0 0.001 0.001 15.535 15.566 dbcsr_multiply_generic 2286 12.5 0.093 0.098 14.542 14.612 qs_scf_new_mos 108 7.5 0.001 0.001 12.330 12.360 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.330 12.359 multiply_cannon 2286 13.5 0.224 0.231 11.718 12.136 ot_scf_mini 108 9.5 0.002 0.002 11.611 11.635 multiply_cannon_loop 2286 14.5 0.637 0.663 10.760 10.966 multiply_cannon_multrec 9144 15.5 1.686 1.939 6.743 7.157 ot_mini 108 10.5 0.001 0.001 6.548 6.577 rebuild_ks_matrix 119 8.3 0.000 0.000 5.849 5.869 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.849 5.869 qs_ot_get_derivative 108 11.5 0.001 0.001 5.228 5.252 dbcsr_mm_accdrv_process 12550 15.8 4.150 5.097 4.946 5.251 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.206 5.223 init_scf_run 11 5.9 0.000 0.001 4.485 4.486 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.485 4.485 sum_up_and_integrate 119 10.3 0.001 0.001 3.500 3.505 integrate_v_rspace 119 11.3 0.003 0.003 3.490 3.495 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.410 3.415 calculate_rho_elec 119 8.7 0.060 0.061 3.410 3.414 calculate_first_density_matrix 1 7.0 0.000 0.000 3.288 3.289 qs_ot_get_p 119 10.4 0.001 0.001 3.113 3.140 calculate_dm_sparse 119 9.5 0.000 0.000 2.721 2.739 init_scf_loop 11 6.9 0.000 0.000 2.558 2.558 mp_waitall_1 115863 16.7 1.707 2.414 1.707 2.414 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.107 2.114 make_m2s 4572 13.5 0.034 0.035 1.932 2.075 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.042 2.045 pw_transfer 1439 11.6 0.066 0.068 2.035 2.045 make_images 4572 14.5 0.270 0.300 1.843 1.983 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.942 1.953 grid_integrate_task_list 119 12.3 1.876 1.950 1.876 1.950 density_rs2pw 119 9.7 0.003 0.004 1.804 1.845 prepare_preconditioner 11 7.9 0.000 0.000 1.816 1.821 make_preconditioner 11 8.9 0.000 0.000 1.816 1.821 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.810 1.811 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.785 1.801 jit_kernel_multiply 6 15.8 0.757 1.796 0.757 1.796 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.787 1.787 acc_transpose_blocks 9144 15.5 0.042 0.043 1.684 1.743 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.705 1.728 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.630 1.637 fft_wrap_pw1pw2_140 487 13.2 0.322 0.333 1.569 1.579 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.486 1.494 cp_fm_diag_elpa_base 50 14.0 1.457 1.473 1.483 1.492 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.413 1.426 fft3d_ps 1201 14.6 0.559 0.571 1.356 1.368 grid_collocate_task_list 119 9.7 1.297 1.353 1.297 1.353 ot_diis_step 108 11.5 0.013 0.013 1.304 1.304 potential_pw2rs 119 12.3 0.010 0.011 1.297 1.302 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.238 1.264 apply_single 119 13.6 0.000 0.000 1.238 1.264 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.230 1.232 wfi_extrapolate 11 7.9 0.001 0.001 1.141 1.141 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.813 1.025 make_images_data 4572 15.5 0.042 0.046 0.842 1.004 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.915 0.959 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.929 cp_fm_cholesky_invert 11 10.9 0.923 0.926 0.923 0.926 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.480 0.913 transfer_rs2pw 487 10.6 0.004 0.004 0.845 0.894 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.864 0.872 mp_alltoall_d11v 2130 13.8 0.813 0.864 0.813 0.864 jit_kernel_transpose 5 15.6 0.798 0.839 0.798 0.839 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.767 0.769 acc_transpose_blocks_sync 27432 16.5 0.711 0.753 0.711 0.753 mp_allgather_i34 2286 14.5 0.275 0.740 0.275 0.740 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.678 0.731 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 0.362 0.722 mp_alltoall_z22v 1201 16.6 0.670 0.706 0.670 0.706 transfer_pw2rs 487 13.2 0.003 0.004 0.700 0.703 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=31.042000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=591.818182, yerr=6.506195 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_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 756.908032E+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 0.013 0.031 44.348 44.349 qs_mol_dyn_low 1 2.0 0.003 0.003 44.142 44.150 qs_forces 11 3.9 0.002 0.002 44.008 44.009 qs_energies 11 4.9 0.001 0.001 41.992 41.996 scf_env_do_scf 11 5.9 0.001 0.001 35.634 35.634 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.746 27.748 velocity_verlet 10 3.0 0.001 0.002 24.374 24.388 dbcsr_multiply_generic 2286 12.5 0.102 0.106 20.390 20.649 qs_scf_new_mos 108 7.5 0.001 0.001 18.208 18.317 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.208 18.316 multiply_cannon 2286 13.5 0.298 0.304 16.236 17.286 ot_scf_mini 108 9.5 0.002 0.002 17.037 17.147 multiply_cannon_loop 2286 14.5 0.863 0.880 14.926 16.016 ot_mini 108 10.5 0.001 0.001 10.420 10.550 multiply_cannon_multrec 9144 15.5 3.620 5.062 9.550 9.767 qs_ot_get_derivative 108 11.5 0.001 0.001 8.293 8.401 init_scf_loop 11 6.9 0.000 0.000 7.856 7.860 rebuild_ks_matrix 119 8.3 0.000 0.001 7.192 7.353 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.191 7.353 dbcsr_mm_accdrv_process 12550 15.8 4.915 7.180 5.789 7.213 prepare_preconditioner 11 7.9 0.000 0.000 6.888 6.904 make_preconditioner 11 8.9 0.000 0.000 6.888 6.904 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.436 6.782 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.479 6.626 cp_fm_upper_to_full 72 14.2 3.170 4.568 3.170 4.568 init_scf_run 11 5.9 0.000 0.001 4.305 4.305 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.305 4.305 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.102 4.110 calculate_rho_elec 119 8.7 0.118 0.121 4.101 4.109 sum_up_and_integrate 119 10.3 0.001 0.001 3.838 3.843 mp_waitall_1 94719 16.7 2.621 3.843 2.621 3.843 integrate_v_rspace 119 11.3 0.004 0.004 3.827 3.833 qs_ot_get_p 119 10.4 0.001 0.001 3.617 3.772 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.011 3.428 calculate_dm_sparse 119 9.5 0.000 0.000 2.879 2.894 dbcsr_complete_redistribute 329 12.2 0.284 0.289 2.016 2.837 calculate_first_density_matrix 1 7.0 0.000 0.000 2.809 2.811 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.528 2.676 make_m2s 4572 13.5 0.038 0.038 2.492 2.673 pw_transfer 1439 11.6 0.069 0.069 2.654 2.658 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.273 2.580 apply_single 119 13.6 0.000 0.000 2.273 2.580 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.555 2.560 make_images 4572 14.5 0.353 0.387 2.371 2.553 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.693 2.513 mp_alltoall_i22 627 13.8 1.438 2.304 1.438 2.304 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.447 2.258 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.133 2.191 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.164 2.166 density_rs2pw 119 9.7 0.003 0.003 2.106 2.151 fft_wrap_pw1pw2_140 487 13.2 0.576 0.580 2.140 2.146 grid_integrate_task_list 119 12.3 2.105 2.145 2.105 2.145 ot_diis_step 108 11.5 0.014 0.014 2.099 2.099 acc_transpose_blocks 9144 15.5 0.044 0.044 1.972 1.996 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.931 1.932 mp_sum_l 7287 12.8 1.120 1.913 1.120 1.913 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.872 1.880 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.796 1.797 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.765 1.767 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.667 1.714 fft3d_ps 1201 14.6 0.596 0.608 1.652 1.656 grid_collocate_task_list 119 9.7 1.540 1.578 1.540 1.578 cp_fm_cholesky_invert 11 10.9 1.507 1.510 1.507 1.510 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.484 1.485 cp_fm_diag_elpa_base 50 14.0 1.336 1.391 1.482 1.483 wfi_extrapolate 11 7.9 0.001 0.001 1.431 1.431 hybrid_alltoall_any 4725 16.4 0.090 0.151 1.118 1.394 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.339 1.360 make_images_data 4572 15.5 0.046 0.050 1.089 1.323 potential_pw2rs 119 12.3 0.014 0.014 1.320 1.321 mp_alltoall_d11v 2130 13.8 1.178 1.227 1.178 1.227 jit_kernel_multiply 5 15.4 0.845 1.215 0.845 1.215 acc_transpose_blocks_sync 27432 16.5 1.101 1.125 1.101 1.125 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.099 1.116 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.997 1.038 qs_create_task_list 11 7.9 0.000 0.001 0.934 0.945 generate_qs_task_list 11 8.9 0.367 0.387 0.933 0.944 mp_alltoall_z22v 1201 16.6 0.920 0.942 0.920 0.942 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.924 0.938 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.349000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=713.454545, yerr=9.576141 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_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 501.731328E+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 1520677. 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 0.010 0.028 83.052 83.054 qs_mol_dyn_low 1 2.0 0.003 0.005 82.764 82.773 qs_forces 11 3.9 0.003 0.003 82.695 82.696 qs_energies 11 4.9 0.001 0.001 79.805 79.831 scf_env_do_scf 11 5.9 0.000 0.001 70.499 70.501 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.920 64.921 dbcsr_multiply_generic 2055 12.4 0.106 0.108 51.181 51.451 qs_scf_new_mos 99 7.5 0.001 0.001 47.439 47.602 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.439 47.602 ot_scf_mini 99 9.5 0.002 0.002 45.098 45.227 multiply_cannon 2055 13.4 0.187 0.192 42.957 43.876 velocity_verlet 10 3.0 0.001 0.001 43.231 43.232 multiply_cannon_loop 2055 14.4 1.779 1.832 41.898 42.903 ot_mini 99 10.5 0.001 0.001 26.532 26.655 qs_ot_get_derivative 99 11.5 0.001 0.001 19.699 19.817 multiply_cannon_multrec 49320 15.4 11.444 12.333 17.827 18.577 rebuild_ks_matrix 110 8.3 0.000 0.000 14.738 14.899 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.737 14.899 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.928 13.065 mp_waitall_1 220248 16.4 10.727 11.718 10.727 11.718 multiply_cannon_sync_h2d 49320 15.4 9.660 10.382 9.660 10.382 qs_ot_get_p 110 10.4 0.001 0.001 10.184 10.338 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.598 8.132 multiply_cannon_metrocomm3 49320 15.4 0.083 0.088 6.346 7.892 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.266 7.873 apply_single 110 13.6 0.000 0.000 7.265 7.873 init_scf_run 11 5.9 0.000 0.001 7.245 7.245 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.244 7.245 sum_up_and_integrate 110 10.3 0.002 0.003 7.198 7.210 integrate_v_rspace 110 11.3 0.003 0.003 7.173 7.191 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.948 6.964 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.589 6.716 calculate_rho_elec 110 8.6 0.020 0.024 6.589 6.716 ot_diis_step 99 11.5 0.006 0.006 6.622 6.622 dbcsr_mm_accdrv_process 87628 16.1 3.026 3.121 6.252 6.592 cp_dbcsr_syevd 48 12.0 0.002 0.003 6.155 6.155 init_scf_loop 11 6.9 0.000 0.000 5.549 5.549 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.514 5.547 cp_fm_diag_elpa_base 48 14.0 5.498 5.533 5.512 5.545 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.333 5.396 mp_sum_l 6594 12.7 3.713 4.540 3.713 4.540 make_m2s 4110 13.4 0.060 0.066 4.074 4.195 wfi_extrapolate 11 7.9 0.001 0.001 4.117 4.117 make_images 4110 14.4 0.178 0.191 3.978 4.101 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.918 3.923 calculate_dm_sparse 110 9.5 0.001 0.001 3.796 3.884 density_rs2pw 110 9.6 0.004 0.005 3.623 3.849 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.570 3.624 multiply_cannon_metrocomm1 49320 15.4 0.066 0.068 2.353 3.565 grid_integrate_task_list 110 12.3 3.252 3.432 3.252 3.432 pw_transfer 1331 11.6 0.056 0.065 3.280 3.392 prepare_preconditioner 11 7.9 0.000 0.000 3.369 3.391 make_preconditioner 11 8.9 0.000 0.000 3.369 3.391 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.191 3.304 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.148 3.199 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.072 3.142 jit_kernel_multiply 13 15.9 2.947 3.078 2.947 3.078 calculate_first_density_matrix 1 7.0 0.000 0.000 3.017 3.023 potential_pw2rs 110 12.3 0.006 0.006 2.773 2.815 fft_wrap_pw1pw2_140 451 13.1 0.386 0.427 2.674 2.810 acc_transpose_blocks 49320 15.4 0.214 0.222 2.614 2.741 mp_alltoall_d11v 2046 13.8 2.123 2.606 2.123 2.606 fft3d_ps 1111 14.6 0.792 0.887 2.490 2.577 transfer_rs2pw 451 10.6 0.005 0.006 2.065 2.292 grid_collocate_task_list 110 9.6 2.160 2.287 2.160 2.287 mp_waitany 14300 13.8 1.818 2.083 1.818 2.083 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.940 1.965 make_images_data 4110 15.4 0.043 0.047 1.824 1.963 transfer_pw2rs 451 13.1 0.006 0.007 1.879 1.917 cp_fm_cholesky_invert 11 10.9 1.876 1.880 1.876 1.880 mp_sum_d 3889 11.9 1.349 1.879 1.349 1.879 hybrid_alltoall_any 4261 16.3 0.085 0.486 1.573 1.834 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.742 1.763 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.054000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.363636, yerr=2.568081 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_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 592.445440E+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 2709555. 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 0.008 0.030 69.328 69.329 qs_mol_dyn_low 1 2.0 0.003 0.003 69.055 69.065 qs_forces 11 3.9 0.003 0.003 68.975 68.977 qs_energies 11 4.9 0.001 0.002 65.658 65.663 scf_env_do_scf 11 5.9 0.001 0.001 56.712 56.716 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.209 49.209 dbcsr_multiply_generic 2055 12.4 0.117 0.121 38.652 38.840 velocity_verlet 10 3.0 0.001 0.002 35.873 35.875 qs_scf_new_mos 99 7.5 0.001 0.001 33.435 33.559 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.434 33.559 multiply_cannon 2055 13.4 0.224 0.244 32.048 33.086 ot_scf_mini 99 9.5 0.003 0.003 31.769 31.899 multiply_cannon_loop 2055 14.4 1.163 1.204 30.911 31.755 ot_mini 99 10.5 0.001 0.001 18.649 18.783 multiply_cannon_multrec 24660 15.4 7.025 8.864 14.633 16.438 rebuild_ks_matrix 110 8.3 0.000 0.000 13.487 13.624 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.487 13.623 qs_ot_get_derivative 99 11.5 0.001 0.001 12.871 13.003 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.862 11.987 mp_waitall_1 176588 16.5 7.736 10.057 7.736 10.057 multiply_cannon_metrocomm3 24660 15.4 0.072 0.074 5.261 7.714 dbcsr_mm_accdrv_process 52282 16.1 5.763 6.706 7.438 7.701 multiply_cannon_sync_h2d 24660 15.4 6.420 7.626 6.420 7.626 init_scf_loop 11 6.9 0.000 0.000 7.468 7.469 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.453 7.050 apply_single 110 13.6 0.000 0.001 6.453 7.049 qs_ot_get_p 110 10.4 0.001 0.001 6.458 6.600 init_scf_run 11 5.9 0.000 0.001 6.579 6.579 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.578 6.579 sum_up_and_integrate 110 10.3 0.002 0.002 6.318 6.329 integrate_v_rspace 110 11.3 0.003 0.003 6.291 6.302 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.845 5.856 calculate_rho_elec 110 8.6 0.039 0.047 5.844 5.856 ot_diis_step 99 11.5 0.010 0.010 5.730 5.730 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.768 5.495 prepare_preconditioner 11 7.9 0.000 0.000 5.462 5.481 make_preconditioner 11 8.9 0.000 0.000 5.462 5.481 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.056 5.213 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.586 4.607 make_m2s 4110 13.4 0.057 0.060 4.237 4.523 make_images 4110 14.4 0.410 0.466 4.128 4.409 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.124 4.125 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.724 3.725 pw_transfer 1331 11.6 0.066 0.074 3.518 3.660 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.552 3.563 cp_fm_diag_elpa_base 48 14.0 3.506 3.519 3.550 3.560 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.411 3.555 wfi_extrapolate 11 7.9 0.001 0.001 3.495 3.495 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.390 3.430 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.269 3.345 density_rs2pw 110 9.6 0.004 0.005 3.104 3.305 grid_integrate_task_list 110 12.3 3.166 3.290 3.166 3.290 fft_wrap_pw1pw2_140 451 13.1 0.459 0.472 2.929 3.076 calculate_dm_sparse 110 9.5 0.001 0.001 2.992 3.019 calculate_first_density_matrix 1 7.0 0.000 0.001 2.989 2.992 fft3d_ps 1111 14.6 1.110 1.332 2.509 2.664 make_images_data 4110 15.4 0.048 0.053 2.310 2.618 cp_fm_cholesky_invert 11 10.9 2.557 2.563 2.557 2.563 hybrid_alltoall_any 4261 16.3 0.106 0.465 1.983 2.549 mp_sum_l 6594 12.7 1.735 2.503 1.735 2.503 jit_kernel_multiply 11 16.2 1.311 2.406 1.311 2.406 grid_collocate_task_list 110 9.6 2.169 2.287 2.169 2.287 potential_pw2rs 110 12.3 0.008 0.009 2.203 2.214 acc_transpose_blocks 24660 15.4 0.116 0.121 2.006 2.082 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.971 1.988 mp_alltoall_d11v 2046 13.8 1.701 1.960 1.701 1.960 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.858 1.860 multiply_cannon_metrocomm4 22605 15.4 0.079 0.083 0.756 1.833 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.781 1.794 mp_irecv_dv 57340 16.2 0.626 1.613 0.626 1.613 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.562 1.575 transfer_rs2pw 451 10.6 0.007 0.007 1.422 1.564 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.390 1.492 mp_allgather_i34 2055 14.4 0.446 1.470 0.446 1.470 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.329000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.636364, yerr=6.314093 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 667.496448E+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 0.008 0.028 60.651 60.652 qs_mol_dyn_low 1 2.0 0.003 0.003 60.406 60.415 qs_forces 11 3.9 0.003 0.003 60.327 60.328 qs_energies 11 4.9 0.001 0.001 57.139 57.143 scf_env_do_scf 11 5.9 0.000 0.001 48.805 48.806 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.208 40.208 velocity_verlet 10 3.0 0.001 0.002 32.751 32.753 dbcsr_multiply_generic 2055 12.4 0.107 0.111 29.470 29.718 qs_scf_new_mos 99 7.5 0.001 0.001 25.566 25.671 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.565 25.670 ot_scf_mini 99 9.5 0.002 0.002 24.312 24.429 multiply_cannon 2055 13.4 0.212 0.219 23.100 24.369 multiply_cannon_loop 2055 14.4 0.813 0.846 21.901 23.013 ot_mini 99 10.5 0.001 0.001 13.880 13.998 rebuild_ks_matrix 110 8.3 0.000 0.000 12.157 12.301 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.157 12.301 multiply_cannon_multrec 16440 15.4 3.592 4.380 10.266 11.053 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.703 10.830 mp_waitall_1 139946 16.5 7.163 10.194 7.163 10.194 qs_ot_get_derivative 99 11.5 0.001 0.001 9.354 9.469 init_scf_loop 11 6.9 0.000 0.000 8.561 8.561 multiply_cannon_metrocomm3 16440 15.4 0.046 0.047 4.515 7.394 prepare_preconditioner 11 7.9 0.000 0.000 6.818 6.838 make_preconditioner 11 8.9 0.000 0.000 6.818 6.838 dbcsr_mm_accdrv_process 34862 16.1 5.907 6.183 6.518 6.647 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.146 6.501 sum_up_and_integrate 110 10.3 0.001 0.002 6.278 6.293 integrate_v_rspace 110 11.3 0.003 0.003 6.252 6.267 init_scf_run 11 5.9 0.000 0.001 5.937 5.937 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.937 5.937 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.727 5.736 calculate_rho_elec 110 8.6 0.059 0.059 5.727 5.735 qs_ot_get_p 110 10.4 0.001 0.001 5.544 5.689 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.037 5.487 apply_single 110 13.6 0.000 0.000 5.036 5.487 make_m2s 4110 13.4 0.049 0.051 4.247 4.626 make_images 4110 14.4 0.401 0.528 4.133 4.512 ot_diis_step 99 11.5 0.011 0.011 4.494 4.494 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.891 3.895 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.209 3.852 multiply_cannon_sync_h2d 16440 15.4 3.211 3.814 3.211 3.814 pw_transfer 1331 11.6 0.065 0.073 3.564 3.571 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.516 3.516 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.457 3.467 grid_integrate_task_list 110 12.3 3.175 3.381 3.175 3.381 density_rs2pw 110 9.6 0.004 0.005 2.972 3.132 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.068 3.070 fft_wrap_pw1pw2_140 451 13.1 0.579 0.590 2.982 2.995 wfi_extrapolate 11 7.9 0.001 0.001 2.986 2.987 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.964 2.972 cp_fm_diag_elpa_base 48 14.0 2.899 2.929 2.962 2.970 make_images_data 4110 15.4 0.045 0.049 2.405 2.892 calculate_first_density_matrix 1 7.0 0.000 0.000 2.860 2.862 hybrid_alltoall_any 4261 16.3 0.108 0.381 2.190 2.828 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.740 2.782 calculate_dm_sparse 110 9.5 0.001 0.001 2.748 2.779 cp_fm_cholesky_invert 11 10.9 2.656 2.662 2.656 2.662 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.449 2.508 multiply_cannon_metrocomm4 14385 15.4 0.049 0.053 0.902 2.454 fft3d_ps 1111 14.6 1.090 1.100 2.399 2.409 grid_collocate_task_list 110 9.6 2.219 2.367 2.219 2.367 mp_irecv_dv 48980 15.7 0.825 2.317 0.825 2.317 potential_pw2rs 110 12.3 0.011 0.011 2.119 2.128 mp_sum_l 6594 12.7 1.499 2.125 1.499 2.125 mp_alltoall_d11v 2046 13.8 1.820 2.064 1.820 2.064 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.958 1.960 dbcsr_complete_redistribute 325 12.2 0.326 0.354 1.470 1.923 cp_fm_upper_to_full 70 14.2 1.406 1.784 1.406 1.784 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.682 1.695 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.646 1.666 mp_allgather_i34 2055 14.4 0.512 1.653 0.512 1.653 acc_transpose_blocks 16440 15.4 0.077 0.079 1.573 1.649 cp_fm_cholesky_decompose 22 10.9 1.581 1.594 1.581 1.594 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.369 1.484 transfer_rs2pw 451 10.6 0.005 0.005 1.280 1.447 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.966 1.420 mp_waitany 17072 13.8 1.129 1.316 1.129 1.316 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.300 1.305 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.652000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.363636, yerr=8.347811 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_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 730.501120E+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 0.009 0.030 67.538 67.539 qs_mol_dyn_low 1 2.0 0.003 0.003 66.933 66.942 qs_forces 11 3.9 0.003 0.003 66.866 66.867 qs_energies 11 4.9 0.001 0.001 63.458 63.463 scf_env_do_scf 11 5.9 0.000 0.001 54.249 54.252 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.585 42.585 velocity_verlet 10 3.0 0.001 0.002 37.400 37.402 dbcsr_multiply_generic 2055 12.4 0.114 0.122 31.889 32.162 qs_scf_new_mos 99 7.5 0.001 0.001 27.991 28.102 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.990 28.101 ot_scf_mini 99 9.5 0.003 0.003 26.301 26.407 multiply_cannon 2055 13.4 0.238 0.254 24.826 25.907 multiply_cannon_loop 2055 14.4 1.408 1.472 23.298 24.015 ot_mini 99 10.5 0.001 0.001 15.148 15.288 multiply_cannon_multrec 24660 15.4 4.051 6.594 13.890 15.024 rebuild_ks_matrix 110 8.3 0.000 0.000 11.957 12.071 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.956 12.071 init_scf_loop 11 6.9 0.000 0.000 11.623 11.624 qs_ot_get_derivative 99 11.5 0.001 0.001 10.931 11.043 dbcsr_mm_accdrv_process 52304 16.0 8.254 9.627 9.682 10.657 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.554 10.656 prepare_preconditioner 11 7.9 0.000 0.000 9.892 9.908 make_preconditioner 11 8.9 0.000 0.000 9.892 9.908 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.112 9.578 mp_waitall_1 121746 16.5 4.396 6.783 4.396 6.783 init_scf_run 11 5.9 0.000 0.001 6.457 6.458 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.457 6.457 sum_up_and_integrate 110 10.3 0.001 0.002 6.238 6.254 integrate_v_rspace 110 11.3 0.003 0.003 6.212 6.227 qs_ot_get_p 110 10.4 0.001 0.001 6.007 6.154 make_m2s 4110 13.4 0.060 0.062 5.403 5.841 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.829 5.837 calculate_rho_elec 110 8.6 0.078 0.081 5.828 5.837 make_images 4110 14.4 0.579 0.699 5.262 5.697 cp_fm_upper_to_full 70 14.2 3.345 4.861 3.345 4.861 ot_diis_step 99 11.5 0.011 0.012 4.176 4.176 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.085 4.173 apply_single 110 13.6 0.000 0.000 4.085 4.173 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.133 4.151 dbcsr_complete_redistribute 325 12.2 0.416 0.457 2.692 3.837 calculate_dm_sparse 110 9.5 0.001 0.001 3.771 3.799 pw_transfer 1331 11.6 0.066 0.074 3.654 3.686 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.569 3.633 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.547 3.582 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.567 3.567 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.363 3.425 grid_integrate_task_list 110 12.3 3.293 3.416 3.293 3.416 calculate_first_density_matrix 1 7.0 0.000 0.000 3.353 3.355 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.133 3.271 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.536 3.185 density_rs2pw 110 9.6 0.004 0.004 2.951 3.168 make_images_data 4110 15.4 0.048 0.053 2.652 3.135 hybrid_alltoall_any 4261 16.3 0.122 0.461 2.334 3.086 fft_wrap_pw1pw2_140 451 13.1 0.609 0.633 3.037 3.075 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.006 3.014 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.011 3.013 cp_fm_diag_elpa_base 48 14.0 2.859 2.918 3.003 3.013 wfi_extrapolate 11 7.9 0.001 0.001 3.010 3.010 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.771 2.895 mp_alltoall_i22 605 13.7 1.658 2.864 1.658 2.864 cp_fm_cholesky_invert 11 10.9 2.698 2.706 2.698 2.706 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.652 2.685 acc_transpose_blocks 24660 15.4 0.115 0.118 2.469 2.682 multiply_cannon_sync_h2d 24660 15.4 2.381 2.541 2.381 2.541 jit_kernel_multiply 8 15.8 1.082 2.480 1.082 2.480 fft3d_ps 1111 14.6 1.083 1.114 2.430 2.446 grid_collocate_task_list 110 9.6 2.261 2.380 2.261 2.380 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.325 2.341 qs_energies_init_hamiltonians 11 5.9 0.022 0.023 2.254 2.255 potential_pw2rs 110 12.3 0.012 0.013 2.049 2.053 mp_alltoall_d11v 2046 13.8 1.782 2.003 1.782 2.003 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.806 1.847 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.614 1.717 cp_fm_cholesky_decompose 22 10.9 1.630 1.663 1.630 1.663 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.642 1.653 acc_transpose_blocks_sync 73980 16.4 1.400 1.507 1.400 1.507 multiply_cannon_metrocomm4 20550 15.4 0.062 0.066 0.869 1.483 mp_sum_l 6594 12.7 0.971 1.472 0.971 1.472 transfer_rs2pw 451 10.6 0.005 0.006 1.199 1.421 mp_irecv_dv 62702 16.1 0.762 1.373 0.762 1.373 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.539000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=692.636364, yerr=8.552222 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_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 855.261184E+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 0.012 0.033 56.316 56.318 qs_mol_dyn_low 1 2.0 0.003 0.003 55.857 55.866 qs_forces 11 3.9 0.009 0.009 55.637 55.638 qs_energies 11 4.9 0.010 0.012 51.972 51.977 scf_env_do_scf 11 5.9 0.000 0.001 43.499 43.499 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.690 35.691 velocity_verlet 10 3.0 0.001 0.001 31.490 31.502 dbcsr_multiply_generic 2055 12.4 0.110 0.119 23.468 23.624 qs_scf_new_mos 99 7.5 0.001 0.001 21.209 21.273 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.208 21.273 ot_scf_mini 99 9.5 0.002 0.002 19.960 20.006 multiply_cannon 2055 13.4 0.235 0.249 17.828 19.080 multiply_cannon_loop 2055 14.4 0.598 0.626 16.550 16.819 rebuild_ks_matrix 110 8.3 0.000 0.000 11.574 11.640 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.573 11.640 ot_mini 99 10.5 0.001 0.001 10.800 10.842 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.291 10.354 multiply_cannon_multrec 8220 15.4 3.145 4.543 7.683 8.912 init_scf_loop 11 6.9 0.000 0.000 7.759 7.761 mp_waitall_1 103326 16.6 5.877 7.646 5.877 7.646 qs_ot_get_derivative 99 11.5 0.001 0.001 7.043 7.089 sum_up_and_integrate 110 10.3 0.002 0.002 6.189 6.204 integrate_v_rspace 110 11.3 0.003 0.003 6.162 6.176 prepare_preconditioner 11 7.9 0.000 0.000 6.120 6.128 make_preconditioner 11 8.9 0.000 0.000 6.120 6.128 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.947 5.959 calculate_rho_elec 110 8.6 0.114 0.114 5.946 5.959 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.700 5.774 dbcsr_mm_accdrv_process 17442 15.9 3.245 4.297 4.397 5.331 init_scf_run 11 5.9 0.000 0.001 5.322 5.322 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.321 5.322 qs_ot_get_p 110 10.4 0.001 0.001 5.150 5.206 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 3.108 4.623 make_m2s 4110 13.4 0.038 0.039 4.251 4.505 make_images 4110 14.4 0.651 0.710 4.121 4.372 pw_transfer 1331 11.6 0.066 0.070 3.832 3.842 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.788 3.832 apply_single 110 13.6 0.000 0.000 3.788 3.832 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.724 3.738 ot_diis_step 99 11.5 0.012 0.014 3.736 3.737 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.729 3.733 grid_integrate_task_list 110 12.3 3.369 3.498 3.369 3.498 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.383 3.383 fft_wrap_pw1pw2_140 451 13.1 0.776 0.786 3.208 3.223 density_rs2pw 110 9.6 0.004 0.004 2.934 3.083 cp_fm_cholesky_invert 11 10.9 2.970 2.975 2.970 2.975 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.784 2.791 cp_fm_diag_elpa_base 48 14.0 2.728 2.748 2.782 2.790 wfi_extrapolate 11 7.9 0.001 0.001 2.727 2.727 hybrid_alltoall_any 4261 16.3 0.199 0.847 2.282 2.693 qs_energies_init_hamiltonians 11 5.9 0.005 0.007 2.677 2.678 calculate_dm_sparse 110 9.5 0.001 0.001 2.618 2.662 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.637 2.639 multiply_cannon_sync_h2d 8220 15.4 2.352 2.597 2.352 2.597 make_images_data 4110 15.4 0.040 0.047 2.255 2.579 grid_collocate_task_list 110 9.6 2.364 2.512 2.364 2.512 calculate_first_density_matrix 1 7.0 0.000 0.000 2.496 2.498 fft3d_ps 1111 14.6 1.129 1.143 2.391 2.403 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.160 2.184 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.116 2.157 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.787 2.009 mp_alltoall_d11v 2046 13.8 1.782 2.000 1.782 2.000 potential_pw2rs 110 12.3 0.015 0.016 1.980 1.984 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.954 1.979 cp_fm_cholesky_decompose 22 10.9 1.689 1.705 1.689 1.705 dbcsr_complete_redistribute 325 12.2 0.653 0.678 1.560 1.653 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.610 1.618 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.499 1.616 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.575 1.587 mp_allgather_i34 2055 14.4 0.450 1.539 0.450 1.539 acc_transpose_blocks 8220 15.4 0.039 0.041 1.446 1.477 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.479 1.406 mp_irecv_dv 24056 15.7 0.452 1.358 0.452 1.358 qs_create_task_list 11 7.9 0.000 0.000 1.217 1.315 generate_qs_task_list 11 8.9 0.374 0.444 1.217 1.315 transfer_rs2pw 451 10.6 0.005 0.005 1.132 1.314 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.246 1.263 mp_waitany 9240 13.8 1.042 1.251 1.042 1.251 jit_kernel_multiply 7 15.4 0.840 1.212 0.840 1.212 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.318000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=807.090909, yerr=11.812586 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_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 1.368490E+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 0.015 0.036 88.265 88.267 qs_mol_dyn_low 1 2.0 0.003 0.003 87.963 87.973 qs_forces 11 3.9 0.006 0.007 87.884 87.885 qs_energies 11 4.9 0.001 0.001 83.747 83.751 scf_env_do_scf 11 5.9 0.000 0.001 73.488 73.488 velocity_verlet 10 3.0 0.002 0.002 56.006 56.014 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.109 45.111 dbcsr_multiply_generic 2055 12.4 0.151 0.160 30.454 30.558 init_scf_loop 11 6.9 0.000 0.000 28.306 28.308 qs_scf_new_mos 99 7.5 0.001 0.001 27.690 27.768 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.690 27.768 prepare_preconditioner 11 7.9 0.000 0.000 26.297 26.311 make_preconditioner 11 8.9 0.000 0.000 26.297 26.311 ot_scf_mini 99 9.5 0.002 0.002 25.895 25.973 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.622 25.756 multiply_cannon 2055 13.4 0.336 0.364 23.148 23.975 multiply_cannon_loop 2055 14.4 0.825 0.838 21.321 21.634 cp_fm_upper_to_full 70 14.2 12.703 18.082 12.703 18.082 ot_mini 99 10.5 0.001 0.001 14.710 14.781 rebuild_ks_matrix 110 8.3 0.001 0.001 13.699 13.768 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.698 13.768 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.425 12.487 dbcsr_complete_redistribute 325 12.2 1.022 1.052 7.272 10.467 multiply_cannon_multrec 8220 15.4 4.144 4.288 10.108 10.301 qs_ot_get_derivative 99 11.5 0.001 0.001 9.908 9.983 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.234 9.425 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.660 8.826 mp_waitall_1 84994 16.7 7.912 8.738 7.912 8.738 mp_alltoall_i22 605 13.7 5.291 8.457 5.291 8.457 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.137 7.176 calculate_rho_elec 110 8.6 0.225 0.225 7.136 7.176 sum_up_and_integrate 110 10.3 0.002 0.002 6.801 6.816 integrate_v_rspace 110 11.3 0.004 0.004 6.773 6.788 dbcsr_mm_accdrv_process 11614 15.7 3.811 4.279 5.815 6.135 init_scf_run 11 5.9 0.000 0.001 5.981 5.981 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.980 5.980 make_m2s 4110 13.4 0.043 0.044 5.444 5.959 qs_ot_get_p 110 10.4 0.001 0.001 5.834 5.917 make_images 4110 14.4 0.884 0.937 5.255 5.769 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.952 5.620 apply_single 110 13.6 0.000 0.000 4.952 5.620 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 5.178 5.582 cp_fm_cholesky_invert 11 10.9 5.381 5.386 5.381 5.386 pw_transfer 1331 11.6 0.075 0.075 4.940 4.946 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.822 4.829 ot_diis_step 99 11.5 0.015 0.016 4.782 4.782 fft_wrap_pw1pw2_140 451 13.1 1.278 1.283 4.201 4.210 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.189 4.195 grid_integrate_task_list 110 12.3 3.727 3.790 3.727 3.790 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.200 3.684 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.665 3.668 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.665 3.666 hybrid_alltoall_any 4261 16.3 0.262 0.560 2.864 3.607 density_rs2pw 110 9.6 0.004 0.004 3.549 3.601 make_images_data 4110 15.4 0.045 0.048 2.877 3.535 wfi_extrapolate 11 7.9 0.001 0.001 3.369 3.369 calculate_dm_sparse 110 9.5 0.001 0.001 3.205 3.237 multiply_cannon_sync_h2d 8220 15.4 3.128 3.152 3.128 3.152 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.134 3.136 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.063 3.064 cp_fm_diag_elpa_base 48 14.0 2.526 2.710 3.061 3.061 fft3d_ps 1111 14.6 1.295 1.299 2.877 2.886 grid_collocate_task_list 110 9.6 2.709 2.737 2.709 2.737 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.691 2.731 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.615 2.640 calculate_first_density_matrix 1 7.0 0.000 0.000 2.496 2.497 potential_pw2rs 110 12.3 0.021 0.021 2.261 2.267 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.181 2.241 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.105 2.181 mp_alltoall_d11v 2046 13.8 2.069 2.155 2.069 2.155 jit_kernel_multiply 10 15.3 1.808 2.147 1.808 2.147 cp_fm_cholesky_decompose 22 10.9 2.022 2.048 2.022 2.048 qs_create_task_list 11 7.9 0.000 0.001 1.885 1.929 generate_qs_task_list 11 8.9 0.731 0.785 1.885 1.929 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.873 1.880 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.768 1.809 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.267000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1241.181818, yerr=46.489135 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 633.491456E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57761. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 2788878. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.032 211.759 211.761 qs_mol_dyn_low 1 2.0 0.003 0.003 211.316 211.330 qs_forces 11 3.9 0.006 0.007 211.205 211.205 qs_energies 11 4.9 0.001 0.002 205.529 205.548 scf_env_do_scf 11 5.9 0.001 0.001 188.655 188.659 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 167.795 167.797 qs_scf_new_mos 117 7.6 0.001 0.001 127.720 127.972 qs_scf_loop_do_ot 117 8.6 0.001 0.001 127.720 127.971 dbcsr_multiply_generic 2507 12.6 0.192 0.197 127.305 127.873 velocity_verlet 10 3.0 0.001 0.001 126.506 126.507 ot_scf_mini 117 9.6 0.003 0.003 120.973 121.197 multiply_cannon 2507 13.6 0.241 0.247 102.276 104.562 multiply_cannon_loop 2507 14.6 2.394 2.453 100.050 102.166 ot_mini 117 10.6 0.001 0.001 67.556 67.814 multiply_cannon_multrec 60168 15.6 31.707 34.300 41.684 44.148 qs_ot_get_derivative 117 11.6 0.001 0.001 42.775 42.998 rebuild_ks_matrix 128 8.3 0.001 0.001 33.932 34.278 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.016 33.931 34.278 mp_waitall_1 267128 16.5 29.727 32.947 29.727 32.947 qs_ot_get_p 128 10.4 0.001 0.001 30.992 31.351 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.460 30.782 multiply_cannon_sync_h2d 60168 15.6 26.309 28.375 26.309 28.375 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.346 25.365 apply_single 128 13.6 0.001 0.001 24.346 25.365 ot_diis_step 117 11.6 0.008 0.008 24.510 24.511 qs_ot_p2m_diag 83 11.4 0.079 0.091 24.236 24.315 cp_dbcsr_syevd 83 12.4 0.005 0.005 21.484 21.485 init_scf_loop 11 6.9 0.000 0.001 20.782 20.783 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.207 20.369 cp_fm_diag_elpa 83 13.4 0.000 0.000 18.382 18.417 cp_fm_diag_elpa_base 83 14.4 18.306 18.351 18.377 18.413 multiply_cannon_metrocomm3 60168 15.6 0.117 0.124 16.518 18.263 prepare_preconditioner 11 7.9 0.000 0.000 16.165 16.204 make_preconditioner 11 8.9 0.000 0.000 16.165 16.204 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.381 15.554 make_m2s 5014 13.6 0.104 0.113 14.143 14.469 sum_up_and_integrate 128 10.3 0.002 0.004 14.350 14.367 integrate_v_rspace 128 11.3 0.004 0.004 14.291 14.310 make_images 5014 14.6 0.399 0.417 13.960 14.297 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.553 13.706 calculate_rho_elec 128 8.7 0.045 0.064 13.553 13.705 init_scf_run 11 5.9 0.000 0.001 12.705 12.705 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.705 12.705 mp_sum_l 7950 12.9 9.614 11.182 9.614 11.182 dbcsr_mm_accdrv_process 124484 16.2 4.699 4.883 9.547 10.251 wfi_extrapolate 11 7.9 0.001 0.001 9.239 9.239 cp_fm_cholesky_invert 11 10.9 9.095 9.103 9.095 9.103 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.721 8.908 calculate_dm_sparse 128 9.5 0.001 0.001 8.559 8.655 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.327 8.560 multiply_cannon_metrocomm1 60168 15.6 0.095 0.098 6.421 8.484 pw_transfer 1547 11.6 0.075 0.093 7.922 8.139 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.718 7.934 make_images_data 5014 15.6 0.066 0.071 6.875 7.774 density_rs2pw 128 9.7 0.006 0.007 7.211 7.748 grid_integrate_task_list 128 12.3 7.072 7.481 7.072 7.481 hybrid_alltoall_any 5200 16.5 0.296 2.294 6.016 7.293 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.936 6.945 fft_wrap_pw1pw2_140 523 13.2 1.120 1.169 6.643 6.848 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.050 6.152 fft3d_ps 1291 14.7 2.193 2.854 5.729 6.113 mp_alltoall_d11v 2415 14.1 4.413 5.617 4.413 5.617 grid_collocate_task_list 128 9.7 4.843 5.245 4.843 5.245 potential_pw2rs 128 12.3 0.009 0.010 4.805 4.853 cp_fm_cholesky_decompose 22 10.9 4.828 4.840 4.828 4.840 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=211.761000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.909091, yerr=6.402479 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/16/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 2.199914E+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 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 846.491648E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57903. MP_Allreduce 11138 958. MP_Sync 87 MP_Alltoall 1983 6316807. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.076 193.412 193.415 qs_mol_dyn_low 1 2.0 0.003 0.003 192.940 192.960 qs_forces 11 3.9 0.004 0.005 192.835 192.836 qs_energies 11 4.9 0.018 0.057 186.051 186.068 scf_env_do_scf 11 5.9 0.001 0.002 169.486 169.496 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 136.193 136.197 velocity_verlet 10 3.0 0.003 0.015 121.852 121.854 dbcsr_multiply_generic 2527 12.6 0.200 0.209 98.361 99.667 qs_scf_new_mos 118 7.6 0.001 0.001 97.021 97.647 qs_scf_loop_do_ot 118 8.6 0.001 0.001 97.020 97.646 ot_scf_mini 118 9.6 0.004 0.005 92.156 92.828 multiply_cannon 2527 13.6 0.506 0.564 77.708 82.716 multiply_cannon_loop 2527 14.6 1.587 1.654 74.316 77.446 ot_mini 118 10.6 0.001 0.001 51.149 51.774 mp_waitall_1 216438 16.6 24.830 38.466 24.830 38.466 multiply_cannon_multrec 30324 15.6 20.851 24.823 31.487 35.688 rebuild_ks_matrix 129 8.3 0.001 0.001 32.653 33.300 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 32.652 33.300 init_scf_loop 11 6.9 0.001 0.003 33.196 33.197 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.403 29.990 multiply_cannon_metrocomm3 30324 15.6 0.097 0.103 16.222 29.846 qs_ot_get_derivative 118 11.6 0.001 0.002 28.701 29.359 prepare_preconditioner 11 7.9 0.000 0.000 28.869 28.948 make_preconditioner 11 8.9 0.000 0.001 28.869 28.948 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.536 28.100 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.504 23.761 apply_single 129 13.6 0.001 0.001 22.504 23.761 qs_ot_get_p 129 10.4 0.001 0.001 22.916 23.561 ot_diis_step 118 11.6 0.014 0.015 22.261 22.263 multiply_cannon_sync_h2d 30324 15.6 17.802 19.917 17.802 19.917 qs_ot_p2m_diag 83 11.4 0.189 0.217 17.899 17.938 cp_fm_cholesky_invert 11 10.9 16.908 16.921 16.908 16.921 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.668 16.669 make_m2s 5054 13.6 0.090 0.096 14.571 16.147 make_images 5054 14.6 1.184 1.394 14.356 15.932 sum_up_and_integrate 129 10.3 0.002 0.004 14.256 14.293 integrate_v_rspace 129 11.3 0.003 0.004 14.196 14.237 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.659 13.684 calculate_rho_elec 129 8.7 0.088 0.105 13.658 13.683 cp_fm_diag_elpa 83 13.4 0.000 0.000 13.369 13.403 cp_fm_diag_elpa_base 83 14.4 13.106 13.213 13.362 13.393 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.309 11.797 init_scf_run 11 5.9 0.000 0.001 11.563 11.564 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.562 11.564 multiply_cannon_metrocomm4 27797 15.6 0.108 0.121 3.974 11.135 mp_irecv_dv 70031 16.3 3.764 10.722 3.764 10.722 dbcsr_mm_accdrv_process 62734 16.2 5.472 6.239 10.087 10.646 make_images_data 5054 15.6 0.067 0.076 8.505 10.409 hybrid_alltoall_any 5240 16.5 0.351 1.537 7.314 9.763 pw_transfer 1559 11.6 0.085 0.096 8.806 8.863 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.581 8.639 wfi_extrapolate 11 7.9 0.001 0.001 8.444 8.445 density_rs2pw 129 9.7 0.006 0.007 7.302 7.840 fft_wrap_pw1pw2_140 527 13.2 1.219 1.244 7.534 7.614 grid_integrate_task_list 129 12.3 7.231 7.609 7.231 7.609 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.460 7.279 cp_fm_cholesky_decompose 22 10.9 7.038 7.109 7.038 7.109 calculate_dm_sparse 129 9.5 0.001 0.001 6.523 6.663 mp_sum_l 8010 12.9 4.302 6.388 4.302 6.388 fft3d_ps 1301 14.7 2.863 3.045 6.128 6.176 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.153 6.161 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.458 5.530 grid_collocate_task_list 129 9.7 5.042 5.436 5.042 5.436 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.224 5.424 mp_allgather_i34 2527 14.6 1.901 5.186 1.901 5.186 mp_alltoall_d11v 2423 14.1 4.312 4.982 4.312 4.982 potential_pw2rs 129 12.3 0.015 0.018 4.673 4.692 dbcsr_complete_redistribute 395 12.7 0.768 0.858 3.253 4.110 mp_sum_d 4499 12.1 2.720 4.094 2.720 4.094 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=193.415000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=805.909091, yerr=4.481366 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 2.951022E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 4017216 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.3 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 980.344832E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1052064 MPI messages size (bytes): total size 2.737208E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.601750E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 284544 37295751168 131072 < size <= 4194304 665856 1004519030784 4194304 < size <= 16777216 66080 937891119088 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4039 57685. MP_Allreduce 11174 996. MP_Sync 88 MP_Alltoall 1724 9394234. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.027 178.847 178.847 qs_mol_dyn_low 1 2.0 0.003 0.003 178.432 178.446 qs_forces 11 3.9 0.004 0.004 178.339 178.343 qs_energies 11 4.9 0.002 0.004 171.797 171.808 scf_env_do_scf 11 5.9 0.001 0.001 156.211 156.214 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 120.925 120.926 velocity_verlet 10 3.0 0.001 0.002 114.570 114.572 dbcsr_multiply_generic 2529 12.6 0.184 0.187 83.390 84.589 qs_scf_new_mos 118 7.6 0.001 0.001 83.358 83.699 qs_scf_loop_do_ot 118 8.6 0.001 0.001 83.358 83.698 ot_scf_mini 118 9.6 0.003 0.004 79.074 79.446 multiply_cannon 2529 13.6 0.509 0.528 63.373 68.272 multiply_cannon_loop 2529 14.6 1.139 1.204 60.490 63.082 ot_mini 118 10.6 0.001 0.001 43.670 44.042 init_scf_loop 11 6.9 0.000 0.000 35.183 35.184 mp_waitall_1 172006 16.6 25.856 35.103 25.856 35.103 prepare_preconditioner 11 7.9 0.000 0.000 31.196 31.246 make_preconditioner 11 8.9 0.000 0.000 31.196 31.246 rebuild_ks_matrix 129 8.3 0.001 0.001 30.471 30.938 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.018 30.471 30.938 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.803 30.235 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.498 27.917 multiply_cannon_multrec 20232 15.6 12.930 16.283 22.505 25.908 multiply_cannon_metrocomm3 20232 15.6 0.063 0.066 16.439 25.661 qs_ot_get_derivative 118 11.6 0.001 0.002 23.696 24.057 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.073 21.396 apply_single 129 13.6 0.001 0.001 20.073 21.396 qs_ot_get_p 129 10.4 0.001 0.001 20.542 20.971 ot_diis_step 118 11.6 0.018 0.018 19.866 19.866 qs_ot_p2m_diag 84 11.4 0.269 0.275 16.178 16.194 make_m2s 5058 13.6 0.080 0.085 14.913 15.859 make_images 5058 14.6 1.166 1.268 14.678 15.625 multiply_cannon_sync_h2d 20232 15.6 13.631 15.400 13.631 15.400 cp_dbcsr_syevd 84 12.4 0.005 0.006 15.094 15.095 cp_fm_cholesky_invert 11 10.9 14.531 14.540 14.531 14.540 sum_up_and_integrate 129 10.3 0.002 0.003 14.134 14.165 integrate_v_rspace 129 11.3 0.003 0.004 14.074 14.103 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.759 13.799 calculate_rho_elec 129 8.7 0.131 0.146 13.758 13.798 cp_fm_diag_elpa 84 13.4 0.000 0.000 11.774 11.792 cp_fm_diag_elpa_base 84 14.4 11.356 11.518 11.770 11.789 init_scf_run 11 5.9 0.000 0.001 10.559 10.559 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.559 10.559 make_images_data 5058 15.6 0.062 0.073 8.884 10.306 multiply_cannon_metrocomm4 17703 15.6 0.068 0.079 3.606 9.789 hybrid_alltoall_any 5245 16.5 0.455 2.064 7.677 9.734 mp_irecv_dv 50659 16.2 3.474 9.522 3.474 9.522 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.237 9.499 dbcsr_mm_accdrv_process 41846 16.2 5.741 6.305 9.043 9.282 pw_transfer 1559 11.6 0.086 0.104 8.809 8.925 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.583 8.704 grid_integrate_task_list 129 12.3 7.384 7.764 7.384 7.764 fft_wrap_pw1pw2_140 527 13.2 1.311 1.345 7.600 7.735 density_rs2pw 129 9.7 0.006 0.006 7.069 7.648 wfi_extrapolate 11 7.9 0.001 0.001 7.490 7.490 cp_fm_cholesky_decompose 22 10.9 7.414 7.453 7.414 7.453 cp_fm_upper_to_full 106 14.8 5.771 7.286 5.771 7.286 dbcsr_complete_redistribute 397 12.7 1.173 1.205 4.665 6.445 fft3d_ps 1301 14.7 2.785 3.019 5.939 6.021 calculate_dm_sparse 129 9.5 0.001 0.001 5.910 5.995 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.939 5.635 grid_collocate_task_list 129 9.7 5.189 5.544 5.189 5.544 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.526 5.531 mp_alltoall_d11v 2429 14.1 4.406 5.393 4.406 5.393 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.466 5.238 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.608 4.733 mp_sum_l 8016 12.9 3.251 4.640 3.251 4.640 potential_pw2rs 129 12.3 0.020 0.022 4.487 4.514 mp_allgather_i34 2529 14.6 1.312 4.391 1.312 4.391 transfer_fm_to_dbcsr 11 9.9 0.019 0.024 2.373 4.113 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.036 4.083 mp_alltoall_i22 720 14.1 1.937 3.851 1.937 3.851 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.724 3.724 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.847000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=926.818182, yerr=20.148621 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.154695E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57949. MP_Allreduce 11127 1081. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.030 188.690 188.691 qs_mol_dyn_low 1 2.0 0.003 0.003 187.984 187.997 qs_forces 11 3.9 0.004 0.004 187.849 187.858 qs_energies 11 4.9 0.001 0.002 180.748 180.756 scf_env_do_scf 11 5.9 0.001 0.001 163.948 163.961 velocity_verlet 10 3.0 0.018 0.032 124.732 124.735 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 117.856 117.857 dbcsr_multiply_generic 2507 12.6 0.188 0.194 80.932 82.012 qs_scf_new_mos 117 7.6 0.001 0.001 81.601 81.917 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.600 81.916 ot_scf_mini 117 9.6 0.003 0.003 77.028 77.378 multiply_cannon 2507 13.6 0.550 0.585 56.126 59.414 multiply_cannon_loop 2507 14.6 1.818 1.893 52.379 54.818 init_scf_loop 11 6.9 0.000 0.000 45.967 45.968 ot_mini 117 10.6 0.001 0.001 43.131 43.469 prepare_preconditioner 11 7.9 0.000 0.000 41.852 41.882 make_preconditioner 11 8.9 0.000 0.000 41.852 41.882 make_full_inverse_cholesky 11 9.9 0.011 0.023 35.487 40.475 multiply_cannon_multrec 30084 15.6 13.505 18.901 26.638 31.683 rebuild_ks_matrix 128 8.3 0.001 0.001 29.273 29.653 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 29.273 29.652 mp_waitall_1 147882 16.7 17.272 27.182 17.272 27.182 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.425 26.766 qs_ot_get_derivative 117 11.6 0.002 0.002 23.538 23.889 make_m2s 5014 13.6 0.096 0.101 20.292 21.340 make_images 5014 14.6 1.951 2.254 19.986 21.033 qs_ot_get_p 128 10.4 0.001 0.001 20.018 20.406 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.945 19.712 apply_single 128 13.6 0.001 0.001 18.945 19.712 ot_diis_step 117 11.6 0.018 0.019 19.469 19.471 cp_fm_upper_to_full 105 14.8 11.373 16.757 11.373 16.757 cp_fm_cholesky_invert 11 10.9 15.882 15.891 15.882 15.891 qs_ot_p2m_diag 83 11.4 0.343 0.392 15.701 15.751 multiply_cannon_metrocomm3 30084 15.6 0.050 0.053 6.413 15.393 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.291 14.293 sum_up_and_integrate 128 10.3 0.002 0.003 13.969 13.995 integrate_v_rspace 128 11.3 0.003 0.004 13.910 13.936 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.740 13.782 calculate_rho_elec 128 8.7 0.173 0.189 13.739 13.781 dbcsr_mm_accdrv_process 62264 16.2 8.667 9.483 12.705 13.166 dbcsr_complete_redistribute 395 12.7 1.498 1.618 9.119 12.894 make_images_data 5014 15.6 0.066 0.073 10.825 12.451 multiply_cannon_sync_h2d 30084 15.6 10.497 11.750 10.497 11.750 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.731 11.510 hybrid_alltoall_any 5200 16.5 0.528 2.190 9.602 11.289 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.037 11.054 cp_fm_diag_elpa_base 83 14.4 10.071 10.385 11.029 11.045 init_scf_run 11 5.9 0.000 0.001 10.885 10.886 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.885 10.886 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.911 10.169 transfer_fm_to_dbcsr 11 9.9 0.001 0.007 6.342 9.999 mp_alltoall_i22 716 14.1 5.594 9.458 5.594 9.458 pw_transfer 1547 11.6 0.085 0.100 8.966 9.047 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.740 8.827 grid_integrate_task_list 128 12.3 7.530 7.915 7.530 7.915 fft_wrap_pw1pw2_140 523 13.2 1.443 1.493 7.772 7.869 wfi_extrapolate 11 7.9 0.001 0.001 7.594 7.594 cp_fm_cholesky_decompose 22 10.9 7.496 7.577 7.496 7.577 density_rs2pw 128 9.7 0.006 0.006 6.772 7.078 multiply_cannon_metrocomm4 25070 15.6 0.084 0.094 2.830 6.599 calculate_dm_sparse 128 9.5 0.001 0.001 6.498 6.597 mp_irecv_dv 76098 16.2 2.674 6.326 2.674 6.326 fft3d_ps 1291 14.7 2.849 2.932 5.920 5.983 grid_collocate_task_list 128 9.7 5.283 5.715 5.283 5.715 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.416 5.477 mp_alltoall_d11v 2415 14.1 4.973 5.418 4.973 5.418 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.519 4.618 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.334 4.421 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.415 4.417 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.245 4.321 potential_pw2rs 128 12.3 0.024 0.024 4.232 4.240 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.691000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1086.818182, yerr=19.399242 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 5.910926E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1979136 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.2 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 1.527497E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 243360 MPI messages size (bytes): total size 1.342108E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.514910E+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 115584 60599304192 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592540656 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9028 51. MP_Alltoall 9736 793528. MP_ISend 40516 2095952. MP_IRecv 40516 2095058. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 57652. MP_Allreduce 11179 1163. MP_Sync 88 MP_Alltoall 1724 18848050. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.049 168.142 168.143 qs_mol_dyn_low 1 2.0 0.003 0.003 167.729 167.744 qs_forces 11 3.9 0.004 0.004 167.643 167.644 qs_energies 11 4.9 0.005 0.006 160.269 160.274 scf_env_do_scf 11 5.9 0.001 0.001 143.204 143.220 velocity_verlet 10 3.0 0.001 0.002 110.993 110.997 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 107.901 107.903 dbcsr_multiply_generic 2535 12.6 0.183 0.195 72.114 72.402 qs_scf_new_mos 118 7.6 0.001 0.001 71.196 71.296 qs_scf_loop_do_ot 118 8.6 0.001 0.001 71.195 71.296 ot_scf_mini 118 9.6 0.003 0.004 66.726 66.790 multiply_cannon 2535 13.6 0.567 0.610 53.560 57.064 multiply_cannon_loop 2535 14.6 0.817 0.851 50.568 51.171 ot_mini 118 10.6 0.001 0.001 37.421 37.478 init_scf_loop 11 6.9 0.000 0.000 35.153 35.155 prepare_preconditioner 11 7.9 0.000 0.000 31.337 31.369 make_preconditioner 11 8.9 0.000 0.000 31.337 31.369 mp_waitall_1 127164 16.7 24.506 30.813 24.506 30.813 make_full_inverse_cholesky 11 9.9 0.012 0.025 29.353 29.603 rebuild_ks_matrix 129 8.3 0.001 0.001 29.016 29.069 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.015 29.068 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.487 26.535 multiply_cannon_multrec 10140 15.6 10.503 14.567 18.220 21.251 qs_ot_get_derivative 118 11.6 0.001 0.002 20.364 20.421 multiply_cannon_metrocomm3 10140 15.6 0.026 0.027 12.675 19.935 cp_fm_cholesky_invert 11 10.9 17.973 17.979 17.973 17.979 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.084 17.191 apply_single 129 13.6 0.001 0.001 17.083 17.190 ot_diis_step 118 11.6 0.020 0.021 16.988 16.988 qs_ot_get_p 129 10.4 0.001 0.001 16.528 16.605 make_m2s 5070 13.6 0.066 0.069 14.773 15.702 make_images 5070 14.6 2.190 2.631 14.463 15.386 sum_up_and_integrate 129 10.3 0.002 0.003 14.043 14.100 integrate_v_rspace 129 11.3 0.004 0.004 13.981 14.040 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.910 13.938 calculate_rho_elec 129 8.7 0.258 0.269 13.910 13.938 qs_ot_p2m_diag 84 11.4 0.502 0.507 12.829 12.843 cp_dbcsr_syevd 84 12.4 0.005 0.006 11.661 11.662 multiply_cannon_sync_h2d 10140 15.6 11.012 11.403 11.012 11.403 init_scf_run 11 5.9 0.000 0.001 10.281 10.281 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.281 10.281 hybrid_alltoall_any 5257 16.5 0.848 3.802 8.359 9.869 make_images_data 5070 15.6 0.056 0.066 8.520 9.811 pw_transfer 1559 11.6 0.086 0.100 9.202 9.242 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 8.977 9.024 cp_fm_diag_elpa 84 13.4 0.000 0.000 8.787 8.797 cp_fm_diag_elpa_base 84 14.4 8.549 8.630 8.783 8.795 grid_integrate_task_list 129 12.3 7.843 8.251 7.843 8.251 dbcsr_mm_accdrv_process 20958 16.1 3.480 4.633 7.362 8.075 fft_wrap_pw1pw2_140 527 13.2 1.822 1.851 7.983 8.065 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 7.984 8.027 cp_fm_cholesky_decompose 22 10.9 7.840 7.934 7.840 7.934 wfi_extrapolate 11 7.9 0.001 0.001 7.311 7.311 multiply_cannon_metrocomm1 10140 15.6 0.030 0.030 4.111 7.158 density_rs2pw 129 9.7 0.005 0.006 6.622 6.919 calculate_dm_sparse 129 9.5 0.001 0.001 6.162 6.231 grid_collocate_task_list 129 9.7 5.606 5.890 5.606 5.890 fft3d_ps 1301 14.7 2.790 2.879 5.675 5.713 dbcsr_complete_redistribute 397 12.7 2.107 2.176 5.080 5.428 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.224 5.226 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.115 5.141 mp_alltoall_d11v 2429 14.1 4.413 4.744 4.413 4.744 mp_allgather_i34 2535 14.6 1.162 4.616 1.162 4.616 potential_pw2rs 129 12.3 0.026 0.027 4.177 4.184 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.035 4.072 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.599 3.893 multiply_cannon_metrocomm4 7605 15.6 0.027 0.029 1.723 3.706 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.583 3.658 mp_irecv_dv 29150 15.9 1.681 3.632 1.681 3.632 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.576 3.608 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.234 3.544 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.414 3.496 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.439 3.448 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=168.143000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1436.636364, yerr=13.240933 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 11.606412E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 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 3.117785E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108490000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3981 59437. MP_Allreduce 11030 1505. MP_Sync 86 MP_Alltoall 1700 36954339. 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 0.019 0.038 286.631 286.632 qs_mol_dyn_low 1 2.0 0.003 0.003 285.984 285.996 qs_forces 11 3.9 0.004 0.005 285.891 285.893 qs_energies 11 4.9 0.001 0.001 277.114 277.124 scf_env_do_scf 11 5.9 0.001 0.001 254.583 254.593 velocity_verlet 10 3.0 0.002 0.002 206.643 206.651 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 131.258 131.260 init_scf_loop 11 6.9 0.000 0.000 123.051 123.054 prepare_preconditioner 11 7.9 0.000 0.000 118.336 118.355 make_preconditioner 11 8.9 0.000 0.000 118.336 118.355 make_full_inverse_cholesky 11 9.9 0.036 0.039 94.646 115.490 qs_scf_new_mos 116 7.6 0.001 0.001 88.610 88.659 qs_scf_loop_do_ot 116 8.6 0.001 0.001 88.609 88.658 ot_scf_mini 116 9.6 0.004 0.004 83.777 83.800 dbcsr_multiply_generic 2485 12.5 0.239 0.253 81.330 81.680 cp_fm_upper_to_full 104 14.8 52.473 75.131 52.473 75.131 multiply_cannon 2485 13.5 0.685 0.730 58.976 60.091 multiply_cannon_loop 2485 14.5 1.041 1.085 55.069 56.130 ot_mini 116 10.6 0.001 0.001 43.730 43.764 dbcsr_complete_redistribute 393 12.7 3.989 4.027 29.562 42.699 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.088 39.225 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 23.647 36.691 mp_alltoall_i22 712 14.1 21.448 34.472 21.448 34.472 rebuild_ks_matrix 127 8.3 0.001 0.001 33.350 33.400 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 33.349 33.400 cp_fm_cholesky_invert 11 10.9 33.090 33.096 33.090 33.096 mp_waitall_1 102768 16.8 27.534 31.545 27.534 31.545 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.927 30.979 qs_ot_get_p 127 10.4 0.001 0.001 25.172 25.250 qs_ot_get_derivative 116 11.6 0.002 0.002 24.291 24.311 qs_ot_p2m_diag 82 11.4 0.868 0.874 21.209 21.236 multiply_cannon_metrocomm3 9940 15.5 0.025 0.026 18.942 20.015 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.441 19.443 ot_diis_step 116 11.6 0.021 0.022 19.403 19.403 make_m2s 4970 13.5 0.076 0.079 17.844 19.230 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.726 18.878 apply_single 127 13.6 0.001 0.001 18.725 18.878 make_images 4970 14.5 3.052 3.226 17.365 18.754 multiply_cannon_multrec 9940 15.5 10.234 12.031 18.147 18.213 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.304 16.305 cp_fm_diag_elpa_base 82 14.4 11.977 13.563 16.301 16.301 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.224 16.241 calculate_rho_elec 127 8.7 0.477 0.478 16.223 16.240 sum_up_and_integrate 127 10.3 0.002 0.002 15.797 15.887 integrate_v_rspace 127 11.3 0.005 0.005 15.735 15.826 multiply_cannon_sync_h2d 9940 15.5 14.192 14.243 14.192 14.243 init_scf_run 11 5.9 0.000 0.001 12.337 12.337 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.337 12.337 hybrid_alltoall_any 5155 16.4 1.301 3.053 10.070 12.059 make_images_data 4970 15.5 0.064 0.069 9.854 11.881 pw_transfer 1535 11.6 0.094 0.095 11.121 11.127 fft_wrap_pw1pw2 1281 12.7 0.011 0.012 10.885 10.890 fft_wrap_pw1pw2_140 519 13.2 3.010 3.055 9.716 9.725 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.524 9.539 dbcsr_mm_accdrv_process 20590 16.0 4.269 6.159 7.666 9.531 cp_fm_cholesky_decompose 22 10.9 8.856 8.964 8.856 8.964 wfi_extrapolate 11 7.9 0.001 0.001 8.816 8.816 grid_integrate_task_list 127 12.3 8.496 8.684 8.496 8.684 qs_energies_init_hamiltonians 11 5.9 0.004 0.006 8.053 8.055 density_rs2pw 127 9.7 0.005 0.005 7.349 7.400 calculate_dm_sparse 127 9.5 0.001 0.001 6.853 6.924 mp_alltoall_d11v 2401 14.1 6.723 6.850 6.723 6.850 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.561 6.609 grid_collocate_task_list 127 9.7 6.368 6.440 6.368 6.440 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.122 6.237 fft3d_ps 1281 14.7 2.827 2.830 6.180 6.207 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=286.632000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2841.000000, yerr=132.672529 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_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 1.260114E+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 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 0.024 0.049 85.401 85.402 qs_energies 1 2.0 0.000 0.000 84.775 84.782 ls_scf 1 3.0 0.000 0.000 83.849 83.858 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.826 73.030 multiply_cannon 111 7.7 0.017 0.020 56.005 57.125 multiply_cannon_loop 111 8.7 0.228 0.243 52.590 53.991 ls_scf_main 1 4.0 0.000 0.000 52.310 52.321 density_matrix_trs4 2 5.0 0.002 0.003 46.769 46.882 ls_scf_init_scf 1 4.0 0.000 0.000 28.500 28.503 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.341 27.391 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.233 25.252 mp_waitall_1 11031 10.9 22.405 25.083 22.405 25.083 multiply_cannon_multrec 2664 9.7 8.162 8.932 15.581 17.343 multiply_cannon_sync_h2d 2664 9.7 13.686 14.857 13.686 14.857 make_m2s 222 7.7 0.009 0.011 13.073 13.520 make_images 222 8.7 0.098 0.108 13.051 13.500 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.578 12.984 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.513 8.903 dbcsr_mm_accdrv_process 4760 10.4 0.594 0.711 7.035 8.194 make_images_data 222 9.7 0.004 0.005 7.649 8.177 hybrid_alltoall_any 227 10.6 0.215 1.839 6.552 7.898 dbcsr_mm_accdrv_process_sort 4760 11.4 6.219 7.134 6.219 7.134 calculate_norms 4752 9.8 5.531 6.189 5.531 6.189 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.026 5.196 mp_sum_l 887 5.1 3.225 5.174 3.225 5.174 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.418 3.863 make_images_sizes 222 9.7 0.000 0.000 0.779 3.676 mp_alltoall_i44 222 10.7 0.779 3.676 0.779 3.676 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.042 3.622 mp_irecv_dv 6231 10.9 2.026 3.614 2.026 3.614 arnoldi_extremal 4 6.8 0.000 0.000 3.254 3.283 arnoldi_normal_ev 4 7.8 0.001 0.003 3.254 3.283 build_subspace 16 8.4 0.009 0.012 3.159 3.161 ls_scf_post 1 4.0 0.000 0.000 3.039 3.046 ls_scf_store_result 1 5.0 0.000 0.000 2.854 2.901 dbcsr_special_finalize 555 9.7 0.005 0.006 2.294 2.843 dbcsr_merge_single_wm 555 10.7 0.454 0.585 2.285 2.834 make_images_pack 222 9.7 2.205 2.604 2.207 2.605 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.342 2.581 dbcsr_sort_data 658 11.4 2.091 2.569 2.091 2.569 dbcsr_matrix_vector_mult_local 304 10.0 2.071 2.475 2.073 2.477 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.312 2.413 buffer_matrices_ensure_size 222 8.7 1.745 2.098 1.745 2.098 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.799 1.800 rebuild_ks_matrix 3 7.3 0.000 0.000 1.790 1.791 qs_ks_build_kohn_sham_matrix 3 8.3 0.008 0.022 1.790 1.791 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.402000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_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 2.183397E+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 0.026 0.042 89.307 89.308 qs_energies 1 2.0 0.000 0.000 88.717 88.721 ls_scf 1 3.0 0.000 0.000 87.385 87.390 dbcsr_multiply_generic 111 6.7 0.015 0.016 73.619 73.939 multiply_cannon 111 7.7 0.027 0.043 52.453 56.456 ls_scf_main 1 4.0 0.000 0.000 54.036 54.041 multiply_cannon_loop 111 8.7 0.135 0.147 49.755 52.890 density_matrix_trs4 2 5.0 0.002 0.003 48.365 48.567 ls_scf_init_scf 1 4.0 0.000 0.000 29.636 29.637 mp_waitall_1 9105 10.9 20.305 29.035 20.305 29.035 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.469 28.540 multiply_cannon_multrec 1332 9.7 13.367 17.231 22.672 27.699 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.142 26.152 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.220 19.870 make_m2s 222 7.7 0.006 0.007 14.911 15.434 make_images 222 8.7 1.369 1.707 14.881 15.403 dbcsr_mm_accdrv_process 4041 10.4 0.356 0.538 8.903 10.489 dbcsr_mm_accdrv_process_sort 4041 11.4 8.420 9.951 8.420 9.951 make_images_data 222 9.7 0.004 0.005 8.597 9.376 hybrid_alltoall_any 227 10.6 0.541 2.573 8.025 9.115 mp_sum_l 887 5.1 4.783 7.983 4.783 7.983 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.218 7.822 mp_irecv_dv 3311 11.0 3.197 7.772 3.197 7.772 calculate_norms 2376 9.8 6.069 6.857 6.069 6.857 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.550 6.327 multiply_cannon_sync_h2d 1332 9.7 4.946 6.087 4.946 6.087 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.919 5.129 arnoldi_extremal 4 6.8 0.000 0.000 4.584 4.599 arnoldi_normal_ev 4 7.8 0.001 0.005 4.584 4.599 build_subspace 16 8.4 0.014 0.021 4.334 4.337 ls_scf_post 1 4.0 0.000 0.000 3.713 3.716 ls_scf_store_result 1 5.0 0.000 0.000 3.441 3.537 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.122 3.350 dbcsr_matrix_vector_mult_local 304 10.0 2.741 3.218 2.743 3.220 mp_allgather_i34 111 8.7 0.691 2.827 0.691 2.827 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.605 2.696 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.134 2.502 dbcsr_data_new 4174 10.1 2.114 2.423 2.114 2.423 make_images_pack 222 9.7 1.822 2.142 1.824 2.144 dbcsr_sort_data 436 11.2 1.850 2.063 1.850 2.063 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.849 1.851 rebuild_ks_matrix 3 7.3 0.000 0.000 1.836 1.838 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.836 1.838 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.308000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1761.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_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 2.933289E+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 0.023 0.042 92.899 92.900 qs_energies 1 2.0 0.000 0.000 92.219 92.222 ls_scf 1 3.0 0.000 0.000 90.811 90.815 dbcsr_multiply_generic 111 6.7 0.016 0.016 75.567 75.846 ls_scf_main 1 4.0 0.000 0.000 56.739 56.743 multiply_cannon 111 7.7 0.035 0.082 52.104 55.991 multiply_cannon_loop 111 8.7 0.116 0.131 49.334 53.925 density_matrix_trs4 2 5.0 0.002 0.003 50.841 51.094 mp_waitall_1 7281 11.0 23.542 33.606 23.542 33.606 ls_scf_init_scf 1 4.0 0.000 0.000 30.416 30.419 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.202 29.268 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.868 26.877 multiply_cannon_multrec 888 9.7 12.670 15.535 21.254 24.686 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.990 23.468 make_m2s 222 7.7 0.006 0.007 16.357 17.124 make_images 222 8.7 1.582 1.849 16.319 17.084 make_images_data 222 9.7 0.004 0.004 9.428 10.601 hybrid_alltoall_any 227 10.6 0.641 2.927 8.940 10.272 dbcsr_mm_accdrv_process 3754 10.4 0.329 0.487 8.105 9.344 mp_sum_l 887 5.1 5.401 9.129 5.401 9.129 dbcsr_mm_accdrv_process_sort 3754 11.4 7.672 8.857 7.672 8.857 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.266 7.346 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.472 7.338 mp_irecv_dv 2335 11.1 2.456 7.281 2.456 7.281 multiply_cannon_sync_h2d 888 9.7 6.055 7.176 6.055 7.176 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.770 6.903 arnoldi_extremal 4 6.8 0.000 0.000 5.045 5.056 arnoldi_normal_ev 4 7.8 0.001 0.005 5.045 5.056 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.745 4.968 calculate_norms 1584 9.8 4.331 4.765 4.331 4.765 build_subspace 16 8.4 0.014 0.020 4.750 4.755 mp_allgather_i34 111 8.7 0.886 3.770 0.886 3.770 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.428 3.763 ls_scf_post 1 4.0 0.000 0.000 3.656 3.661 dbcsr_matrix_vector_mult_local 304 10.0 3.022 3.598 3.024 3.600 ls_scf_store_result 1 5.0 0.000 0.000 3.405 3.477 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.845 2.951 dbcsr_data_new 4116 9.9 2.111 2.454 2.111 2.454 make_images_sizes 222 9.7 0.000 0.000 1.098 2.353 mp_alltoall_i44 222 10.7 1.097 2.352 1.097 2.352 dbcsr_sort_data 325 11.1 1.879 2.138 1.879 2.138 dbcsr_finalize 304 7.8 0.026 0.032 1.607 1.936 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.917 1.919 rebuild_ks_matrix 3 7.3 0.000 0.000 1.898 1.900 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.898 1.900 make_images_pack 222 9.7 1.623 1.884 1.626 1.886 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.900000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2245.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_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 3.340161E+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 0.057 0.075 96.423 96.424 qs_energies 1 2.0 0.000 0.000 95.828 95.834 ls_scf 1 3.0 0.000 0.000 94.165 94.170 dbcsr_multiply_generic 111 6.7 0.017 0.017 77.966 78.192 ls_scf_main 1 4.0 0.000 0.000 58.508 58.509 multiply_cannon 111 7.7 0.041 0.104 51.230 55.790 density_matrix_trs4 2 5.0 0.002 0.003 52.410 52.534 multiply_cannon_loop 111 8.7 0.152 0.171 46.186 49.675 ls_scf_init_scf 1 4.0 0.000 0.000 32.429 32.431 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.205 31.280 mp_waitall_1 6369 11.0 22.154 29.678 22.154 29.678 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.750 28.762 multiply_cannon_multrec 1332 9.7 14.176 17.617 22.094 25.097 make_m2s 222 7.7 0.006 0.008 21.057 22.472 make_images 222 8.7 3.148 3.624 21.007 22.424 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 8.851 16.887 make_images_data 222 9.7 0.004 0.004 11.712 13.298 hybrid_alltoall_any 227 10.6 0.797 3.764 11.038 12.837 dbcsr_mm_accdrv_process 3641 10.4 0.278 0.476 7.549 9.066 dbcsr_mm_accdrv_process_sort 3641 11.4 7.110 8.569 7.110 8.569 mp_sum_l 887 5.1 4.081 7.519 4.081 7.519 multiply_cannon_sync_h2d 1332 9.7 5.441 6.080 5.441 6.080 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.093 6.045 mp_irecv_dv 3229 10.9 2.068 5.967 2.068 5.967 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.154 5.942 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.420 5.741 arnoldi_extremal 4 6.8 0.000 0.000 5.192 5.204 arnoldi_normal_ev 4 7.8 0.001 0.005 5.192 5.204 build_subspace 16 8.4 0.014 0.021 4.859 4.867 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.511 4.681 mp_allgather_i34 111 8.7 2.162 4.564 2.162 4.564 calculate_norms 2376 9.8 4.190 4.497 4.190 4.497 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.585 3.878 dbcsr_matrix_vector_mult_local 304 10.0 3.196 3.693 3.198 3.695 dbcsr_sort_data 658 11.4 3.115 3.463 3.115 3.463 ls_scf_post 1 4.0 0.000 0.000 3.228 3.233 dbcsr_special_finalize 555 9.7 0.006 0.008 2.855 3.176 dbcsr_merge_single_wm 555 10.7 0.539 0.654 2.846 3.168 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.008 3.042 ls_scf_store_result 1 5.0 0.000 0.000 2.970 3.032 dbcsr_data_release 10477 10.7 1.590 2.473 1.590 2.473 dbcsr_finalize 304 7.8 0.049 0.061 1.813 1.987 buffer_matrices_ensure_size 222 8.7 1.587 1.937 1.587 1.937 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.424000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2699.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_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 4.757000E+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 0.035 0.051 92.422 92.422 qs_energies 1 2.0 0.000 0.000 91.700 91.703 ls_scf 1 3.0 0.000 0.000 89.764 89.767 dbcsr_multiply_generic 111 6.7 0.017 0.019 71.304 71.579 ls_scf_main 1 4.0 0.000 0.000 56.992 56.992 multiply_cannon 111 7.7 0.088 0.135 52.766 55.615 multiply_cannon_loop 111 8.7 0.088 0.093 50.151 51.830 density_matrix_trs4 2 5.0 0.002 0.003 50.013 50.090 ls_scf_init_scf 1 4.0 0.000 0.000 29.325 29.327 mp_waitall_1 5436 11.0 24.882 28.758 24.882 28.758 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.094 28.125 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.008 26.019 multiply_cannon_multrec 444 9.7 13.656 16.323 20.746 22.252 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.915 15.855 make_m2s 222 7.7 0.005 0.005 13.725 14.708 make_images 222 8.7 2.039 2.475 13.657 14.639 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.280 14.531 hybrid_alltoall_any 227 10.6 0.802 3.822 8.274 9.956 make_images_data 222 9.7 0.003 0.004 8.506 9.849 multiply_cannon_sync_h2d 444 9.7 6.731 8.148 6.731 8.148 dbcsr_mm_accdrv_process 3003 10.4 0.341 0.410 6.783 7.956 dbcsr_mm_accdrv_process_sort 3003 11.4 6.419 7.550 6.419 7.550 arnoldi_extremal 4 6.8 0.000 0.000 5.846 5.859 arnoldi_normal_ev 4 7.8 0.002 0.005 5.846 5.859 build_subspace 16 8.4 0.015 0.019 5.450 5.460 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.413 4.626 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.183 4.393 mp_sum_l 887 5.1 2.809 4.358 2.809 4.358 dbcsr_matrix_vector_mult_local 304 10.0 3.725 4.203 3.727 4.205 mp_allgather_i34 111 8.7 1.199 3.848 1.199 3.848 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.532 3.730 calculate_norms 792 9.8 3.609 3.716 3.609 3.716 mp_irecv_dv 1241 11.2 1.517 3.672 1.517 3.672 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.534 3.614 ls_scf_post 1 4.0 0.000 0.000 3.447 3.450 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.994 3.431 make_images_sizes 222 9.7 0.000 0.000 0.836 3.342 mp_alltoall_i44 222 10.7 0.835 3.341 0.835 3.341 ls_scf_store_result 1 5.0 0.000 0.000 3.199 3.241 dbcsr_finalize 304 7.8 0.062 0.077 2.196 2.299 dbcsr_data_new 4608 9.7 1.771 2.197 1.771 2.197 dbcsr_merge_all 275 8.9 0.479 0.532 2.054 2.136 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.003 2.004 rebuild_ks_matrix 3 7.3 0.000 0.000 1.972 1.972 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.972 1.972 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 1.920 1.921 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.422000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3703.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_performance_tests/26/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 30.358840E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 339931 0.0% 0.0% 100.0% average stack size 0.0 0.0 28341.7 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 8.872993E+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 0.071 0.094 105.507 105.508 qs_energies 1 2.0 0.000 0.000 104.082 104.085 ls_scf 1 3.0 0.000 0.000 101.175 101.177 dbcsr_multiply_generic 111 6.7 0.024 0.026 74.830 74.986 ls_scf_main 1 4.0 0.000 0.000 63.587 63.588 density_matrix_trs4 2 5.0 0.002 0.003 54.643 54.695 multiply_cannon 111 7.7 0.144 0.231 48.833 50.948 multiply_cannon_loop 111 8.7 0.097 0.099 45.836 46.281 ls_scf_init_scf 1 4.0 0.000 0.000 33.818 33.819 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.318 32.344 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.630 29.638 mp_waitall_1 4527 11.1 21.651 25.310 21.651 25.310 make_m2s 222 7.7 0.005 0.005 22.265 23.347 make_images 222 8.7 3.578 3.898 22.158 23.237 multiply_cannon_multrec 444 9.7 17.834 18.473 22.503 23.235 hybrid_alltoall_any 227 10.6 1.655 3.615 12.516 15.280 make_images_data 222 9.7 0.003 0.004 12.687 14.817 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.374 10.890 multiply_cannon_sync_h2d 444 9.7 8.792 8.839 8.792 8.839 arnoldi_extremal 4 6.8 0.000 0.000 7.464 7.474 arnoldi_normal_ev 4 7.8 0.003 0.009 7.463 7.474 build_subspace 16 8.4 0.026 0.036 6.911 6.924 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.544 5.696 dbcsr_matrix_vector_mult_local 304 10.0 5.107 5.418 5.109 5.421 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.082 5.176 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.848 5.099 dbcsr_mm_accdrv_process 1814 10.4 0.289 0.409 4.490 4.628 dbcsr_mm_accdrv_process_sort 1814 11.4 4.123 4.257 4.123 4.257 ls_scf_post 1 4.0 0.000 0.000 3.769 3.772 make_images_sizes 222 9.7 0.000 0.000 1.449 3.586 mp_alltoall_i44 222 10.7 1.449 3.586 1.449 3.586 ls_scf_store_result 1 5.0 0.000 0.000 3.480 3.515 mp_allgather_i34 111 8.7 1.114 3.482 1.114 3.482 calculate_norms 792 9.8 3.230 3.267 3.230 3.267 dbcsr_finalize 304 7.8 0.082 0.089 3.071 3.182 dbcsr_merge_all 275 8.9 0.882 0.919 2.857 2.962 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.875 2.875 dbcsr_complete_redistribute 5 7.6 1.437 1.477 2.738 2.860 matrix_ls_to_qs 2 6.0 0.000 0.000 2.404 2.532 dbcsr_sort_data 325 11.1 2.440 2.512 2.440 2.512 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.383 2.385 dbcsr_data_new 6591 9.6 1.846 2.335 1.846 2.335 dbcsr_new_transposed 4 7.5 0.240 0.252 2.299 2.327 rebuild_ks_matrix 3 7.3 0.000 0.000 2.317 2.319 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.317 2.319 dbcsr_frobenius_norm 74 6.6 2.058 2.136 2.200 2.233 dbcsr_add_d 103 6.2 0.000 0.000 2.124 2.205 dbcsr_add_anytype 103 7.2 0.859 0.890 2.124 2.205 dbcsr_data_release 12724 10.6 1.972 2.191 1.972 2.191 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.508000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6991.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c_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 593.068032E+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 72984768. 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 0.017 0.048 236.702 236.704 qs_mol_dyn_low 1 2.0 0.004 0.045 235.730 235.754 qs_forces 5 3.8 0.004 0.004 235.584 235.586 qs_energies 5 4.8 0.001 0.002 232.557 232.580 scf_env_do_scf 5 5.8 0.000 0.001 217.715 217.719 scf_env_do_scf_inner_loop 105 6.6 0.007 0.027 190.699 190.701 qs_scf_new_mos 105 7.6 0.001 0.001 149.882 150.044 qs_scf_loop_do_ot 105 8.6 0.001 0.001 149.881 150.043 ot_scf_mini 105 9.6 0.003 0.003 139.939 140.094 dbcsr_multiply_generic 1445 12.2 0.127 0.136 134.305 134.672 multiply_cannon 1445 13.2 0.275 0.285 114.964 116.853 multiply_cannon_loop 1445 14.2 2.847 2.993 113.269 114.626 velocity_verlet 4 3.0 0.004 0.016 107.113 107.114 ot_mini 105 10.6 0.001 0.001 60.730 60.843 qs_ot_get_p 112 10.4 0.001 0.001 49.234 49.517 multiply_cannon_multrec 69360 15.2 29.658 34.779 39.878 45.364 mp_waitall_1 488190 16.1 35.233 41.853 35.233 41.853 qs_ot_get_derivative 55 11.6 0.001 0.001 39.006 39.123 qs_ot_p2m_diag 40 11.0 0.020 0.030 38.064 38.154 cp_dbcsr_syevd 40 12.0 0.002 0.002 34.823 34.823 multiply_cannon_sync_h2d 69360 15.2 29.115 33.034 29.115 33.034 multiply_cannon_metrocomm3 69360 15.2 0.199 0.211 25.933 32.991 cp_fm_syevd 40 13.0 0.000 0.000 29.483 29.616 rebuild_ks_matrix 110 8.4 0.000 0.000 29.072 29.257 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.030 29.072 29.256 init_scf_loop 7 6.6 0.000 0.000 26.981 26.982 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.689 26.854 cp_fm_redistribute_end 40 14.0 12.307 24.564 12.313 24.565 cp_fm_syevd_base 40 14.0 12.244 24.502 12.244 24.502 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.977 23.202 apply_single 62 13.6 0.000 0.000 22.977 23.202 prepare_preconditioner 7 7.6 0.000 0.000 22.131 22.161 make_preconditioner 7 8.6 0.000 0.000 22.131 22.161 ot_new_cg_direction 55 11.6 0.001 0.001 20.968 20.968 qs_rho_update_rho_low 110 7.6 0.001 0.001 17.551 17.935 calculate_rho_elec 110 8.6 0.030 0.032 17.550 17.934 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.790 15.005 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.774 14.833 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.897 13.997 mp_sum_l 4764 12.2 12.348 13.242 12.348 13.242 init_scf_run 5 5.8 0.000 0.001 12.251 12.252 scf_env_initial_rho_setup 5 6.8 0.001 0.003 12.251 12.252 density_rs2pw 110 9.6 0.005 0.007 11.770 12.219 pw_transfer 1645 12.4 0.079 0.100 11.890 12.082 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.751 11.945 calculate_dm_sparse 110 9.5 0.000 0.001 11.623 11.829 dbcsr_mm_accdrv_process 154766 15.8 6.242 6.427 10.088 11.009 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.604 10.663 qs_vxc_create 110 10.4 0.002 0.003 10.439 10.474 fft_wrap_pw1pw2_240 915 15.0 1.148 1.222 10.276 10.458 cp_fm_cholesky_invert 7 10.6 10.398 10.406 10.398 10.406 check_diag 80 13.5 8.687 8.955 9.806 9.964 transfer_rs2pw 445 10.6 0.007 0.008 8.303 8.809 fft3d_pb 915 16.0 2.384 2.570 8.482 8.688 sum_up_and_integrate 60 10.3 0.001 0.002 8.614 8.632 integrate_v_rspace 60 11.3 0.002 0.002 8.597 8.616 acc_transpose_blocks 69360 15.2 0.356 0.371 7.628 8.149 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 7.722 7.737 xc_rho_set_and_dset_create 110 12.4 0.077 0.097 7.369 7.612 multiply_cannon_metrocomm1 69360 15.2 0.097 0.102 4.759 7.595 calculate_first_density_matrix 1 7.0 0.000 0.001 7.389 7.399 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.321 7.378 make_full_single_inverse 7 9.6 0.001 0.001 7.076 7.106 make_m2s 2890 13.2 0.078 0.087 6.528 7.087 xc_vxc_pw_create 60 11.3 0.039 0.050 6.969 7.004 make_images 2890 14.2 0.241 0.262 6.421 6.979 xc_pw_derive 510 13.4 0.005 0.007 6.227 6.288 mp_alltoall_z22v 2340 17.7 5.508 5.873 5.508 5.873 acc_transpose_blocks_kernels 69360 16.2 0.853 0.895 4.870 5.278 mp_waitany 7680 13.5 4.355 4.996 4.355 4.996 potential_pw2rs 60 12.3 0.003 0.003 4.798 4.859 multiply_cannon_metrocomm4 67915 15.2 0.183 0.201 2.026 4.781 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=236.704000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 15d06616f0e7b0bb6ad4842b5fcf90aaaf2b983c Summary: empty Status: OK