=== 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: c08aeba1059963d86d9bf16a1dbde02da2d999ed ################# 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 1.12.0, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.8.2, 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: 28.08.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.12.0 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/01 job id: 48625758 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/02 job id: 48625759 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/03 job id: 48625760 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/04 job id: 48625761 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/05 job id: 48625762 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/06 job id: 48625765 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/07 job id: 48625767 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/08 job id: 48625768 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/09 job id: 48625769 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/10 job id: 48625771 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/11 job id: 48625772 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/12 job id: 48625773 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/13 job id: 48625774 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/14 job id: 48625775 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/15 job id: 48625776 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/16 job id: 48625777 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/17 job id: 48625778 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/18 job id: 48625779 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/19 job id: 48625780 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/20 job id: 48625781 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/21 job id: 48625782 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/22 job id: 48625783 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/23 job id: 48625784 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/24 job id: 48625785 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/25 job id: 48625786 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/26 job id: 48625787 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/27 job id: 48625788 --- 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 0.021 0.036 132.781 132.782 farming_run 1 2.0 132.358 132.361 132.751 132.754 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.504166E+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 0.006 0.025 114.556 114.557 qs_energies 1 2.0 0.000 0.000 114.365 114.368 mp2_main 1 3.0 0.000 0.000 112.260 112.263 mp2_gpw_main 1 4.0 0.028 0.034 111.236 111.239 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 92.301 92.672 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.625 54.996 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.165 41.015 46.067 get_2c_integrals 1 6.0 0.008 0.009 36.841 37.503 integrate_v_rspace 273 8.0 0.438 0.455 24.600 29.518 pw_transfer 6555 10.6 0.376 0.392 26.790 27.224 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 25.407 25.896 grid_integrate_task_list 273 9.0 20.518 25.896 20.518 25.896 fft_wrap_pw1pw2_100 2178 12.4 1.117 1.254 22.967 23.422 compute_2c_integrals 1 7.0 0.007 0.008 19.238 19.239 rpa_ri_compute_en 1 5.0 0.036 0.041 18.818 18.977 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.609 18.954 mp2_eri_2c_integrate_gpw 1 9.0 2.377 2.421 18.606 18.951 cp_fm_cholesky_decompose 12 8.2 17.550 18.253 17.550 18.253 cholesky_decomp 1 7.0 0.000 0.000 16.450 17.152 fft3d_s 5443 13.4 16.123 16.288 16.145 16.308 ao_to_mo_and_store_B_mult_1 272 7.0 10.776 15.319 10.776 15.319 calculate_wavefunction 272 8.0 5.415 5.565 12.324 13.035 rpa_num_int 1 6.0 0.000 0.002 10.648 10.649 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.625 10.637 calc_mat_Q 8 8.0 0.000 0.000 9.476 9.558 contract_S_to_Q 8 9.0 0.000 0.000 8.896 8.981 calc_potential_gpw 544 9.5 0.005 0.006 8.148 8.770 parallel_gemm_fm 14 9.1 0.000 0.000 8.486 8.580 parallel_gemm_fm_cosma 14 10.1 8.486 8.580 8.486 8.580 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.075 8.376 potential_pw2rs 545 10.0 0.107 0.109 7.556 8.182 create_integ_mat 1 6.0 0.022 0.028 7.687 7.688 collocate_single_gaussian 272 10.0 0.040 0.041 7.357 7.644 array2fm 1 7.0 0.000 0.000 6.651 7.123 pw_scatter_s 2720 13.7 4.312 4.481 4.312 4.481 pw_gather_s 2722 13.2 3.460 3.711 3.460 3.711 array2fm_buffer_send 1 8.0 2.949 3.151 2.949 3.151 pw_poisson_solve 545 10.5 1.119 1.186 2.208 2.414 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.235492, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2820.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 0.029 0.038 420.369 420.370 farming_run 1 2.0 419.622 419.629 420.324 420.327 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.235243E+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 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.033 217.140 217.141 qs_energies 1 2.0 0.000 0.000 216.923 216.930 scf_env_do_scf 1 3.0 0.000 0.000 115.190 115.190 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.290 114.297 rebuild_ks_matrix 4 6.0 0.000 0.000 114.289 114.296 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 114.289 114.295 hfx_ks_matrix 4 8.0 0.001 0.001 113.940 113.944 integrate_four_center 4 9.0 0.152 0.465 113.939 113.944 integrate_four_center_main 4 10.0 0.164 0.645 101.955 105.856 integrate_four_center_bin 265 11.0 101.791 105.416 101.791 105.416 mp2_main 1 3.0 0.000 0.000 101.442 101.449 mp2_gpw_main 1 4.0 0.047 0.069 100.562 100.572 init_scf_loop 1 4.0 0.000 0.000 96.974 96.974 mp2_ri_gpw_compute_in 1 5.0 0.064 0.065 73.982 75.030 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.728 54.776 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.159 41.289 46.277 integrate_v_rspace 95 8.0 0.398 0.563 27.777 32.605 pw_transfer 2240 10.6 0.145 0.188 29.455 29.870 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.476 28.891 ao_to_mo_and_store_B_mult_1 91 7.0 10.759 28.857 10.759 28.857 grid_integrate_task_list 95 9.0 23.193 28.224 23.193 28.224 mp2_ri_gpw_compute_en 1 5.0 0.054 0.063 26.417 27.996 fft_wrap_pw1pw2_100 730 12.4 1.254 1.479 26.228 26.663 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.842 1.926 24.731 24.744 get_2c_integrals 1 6.0 0.000 0.000 20.159 20.190 compute_2c_integrals 1 7.0 0.003 0.003 19.134 19.138 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.550 19.006 mp2_eri_2c_integrate_gpw 1 9.0 1.738 1.883 18.549 19.005 fft3d_s 1823 13.4 18.478 18.818 18.491 18.831 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.213 18.213 calculate_wavefunction 91 8.0 2.009 2.044 9.608 9.887 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.577 8.695 9.253 potential_pw2rs 186 10.0 0.033 0.035 8.397 9.002 local_gemm 172 8.0 8.138 8.690 8.138 8.690 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.114 8.568 mp2_ri_gpw_compute_en_comm 22 7.0 0.502 0.518 7.799 8.297 calc_potential_gpw 182 9.5 0.002 0.003 7.889 8.164 collocate_single_gaussian 91 10.0 0.017 0.019 7.715 8.013 mp_sync 37 10.5 3.945 7.217 3.945 7.217 integrate_four_center_load 4 10.0 0.000 0.000 6.785 6.790 hfx_load_balance 1 11.0 0.000 0.000 6.785 6.790 mp2_ri_gpw_compute_en_ener 172 7.0 6.350 6.425 6.350 6.425 mp_sendrecv_dm3 2068 8.0 5.824 6.323 5.824 6.323 pw_gather_s 912 13.2 4.488 4.986 4.488 4.986 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.561748, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1493.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 451.825664E+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 225721. 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 0.011 0.031 53.572 53.573 qs_mol_dyn_low 1 2.0 0.003 0.003 53.362 53.369 qs_forces 11 3.9 0.002 0.003 53.305 53.305 qs_energies 11 4.9 0.002 0.002 51.809 51.835 scf_env_do_scf 11 5.9 0.000 0.001 45.859 45.860 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.765 43.766 dbcsr_multiply_generic 2286 12.5 0.100 0.104 33.506 34.118 qs_scf_new_mos 108 7.5 0.000 0.000 33.810 34.096 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.810 34.095 ot_scf_mini 108 9.5 0.002 0.002 32.101 32.274 multiply_cannon 2286 13.5 0.190 0.207 26.251 28.197 multiply_cannon_loop 2286 14.5 1.829 1.941 25.601 27.573 velocity_verlet 10 3.0 0.001 0.002 26.948 26.948 ot_mini 108 10.5 0.001 0.001 19.471 19.720 qs_ot_get_derivative 108 11.5 0.001 0.001 16.451 16.621 mp_waitall_1 245248 16.5 8.530 14.617 8.530 14.617 multiply_cannon_metrocomm3 54864 15.5 0.073 0.080 6.150 13.126 multiply_cannon_multrec 54864 15.5 3.624 5.700 7.597 10.939 qs_ot_get_p 119 10.4 0.001 0.001 8.023 8.316 rebuild_ks_matrix 119 8.3 0.000 0.000 7.768 7.915 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.768 7.915 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.836 6.964 mp_sum_l 7287 12.8 5.351 6.961 5.351 6.961 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.011 6.488 multiply_cannon_sync_h2d 54864 15.5 5.114 6.456 5.114 6.456 dbcsr_mm_accdrv_process 76910 16.1 1.833 2.887 3.887 5.566 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.181 5.232 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.905 5.015 init_scf_run 11 5.9 0.000 0.001 4.651 4.651 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.651 4.651 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.393 4.393 sum_up_and_integrate 119 10.3 0.001 0.002 4.380 4.388 integrate_v_rspace 119 11.3 0.002 0.002 4.369 4.378 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.116 4.283 calculate_rho_elec 119 8.7 0.011 0.017 4.116 4.282 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.202 4.202 cp_fm_redistribute_end 50 14.0 2.143 4.174 2.149 4.177 cp_fm_diag_elpa_base 50 14.0 2.023 4.075 2.027 4.080 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.944 3.146 apply_single 119 13.6 0.000 0.000 2.943 3.146 calculate_dm_sparse 119 9.5 0.000 0.000 2.931 3.094 multiply_cannon_metrocomm1 54864 15.5 0.056 0.062 1.773 2.934 acc_transpose_blocks 54864 15.5 0.232 0.258 2.224 2.751 ot_diis_step 108 11.5 0.006 0.006 2.729 2.730 jit_kernel_multiply 13 15.8 1.993 2.595 1.993 2.595 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.417 2.493 calculate_first_density_matrix 1 7.0 0.000 0.000 2.374 2.384 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.273 2.275 wfi_extrapolate 11 7.9 0.001 0.001 2.203 2.203 density_rs2pw 119 9.7 0.004 0.004 2.087 2.177 grid_integrate_task_list 119 12.3 2.017 2.121 2.017 2.121 init_scf_loop 11 6.9 0.000 0.000 2.078 2.078 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.939 2.007 mp_sum_d 4135 12.0 1.362 1.986 1.362 1.986 potential_pw2rs 119 12.3 0.004 0.004 1.766 1.775 pw_transfer 1439 11.6 0.051 0.055 1.600 1.669 make_m2s 4572 13.5 0.053 0.056 1.582 1.629 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.525 1.597 make_images 4572 14.5 0.133 0.139 1.499 1.547 mp_alltoall_d11v 2130 13.8 1.338 1.501 1.338 1.501 acc_transpose_blocks_sync 164592 16.5 1.209 1.444 1.209 1.444 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.383 1.423 mp_waitany 12084 13.8 1.250 1.419 1.250 1.419 transfer_rs2pw 487 10.6 0.005 0.006 1.326 1.414 grid_collocate_task_list 119 9.7 1.349 1.412 1.349 1.412 fft3d_ps 1201 14.6 0.368 0.472 1.240 1.306 transfer_pw2rs 487 13.2 0.006 0.006 1.297 1.305 fft_wrap_pw1pw2_140 487 13.2 0.141 0.154 1.180 1.250 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.754 1.149 mp_alltoall_i22 627 13.8 0.798 1.097 0.798 1.097 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.573000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.818182, yerr=0.574960 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 487.800832E+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 1952382. 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 0.018 0.041 39.247 39.249 qs_mol_dyn_low 1 2.0 0.003 0.005 38.944 38.978 qs_forces 11 3.9 0.003 0.010 38.858 38.859 qs_energies 11 4.9 0.002 0.007 37.169 37.174 scf_env_do_scf 11 5.9 0.000 0.001 31.739 31.739 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.322 29.323 dbcsr_multiply_generic 2286 12.5 0.099 0.106 22.373 22.729 qs_scf_new_mos 108 7.5 0.001 0.001 20.652 20.877 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.652 20.876 ot_scf_mini 108 9.5 0.003 0.003 19.729 19.896 multiply_cannon 2286 13.5 0.211 0.219 17.205 18.628 velocity_verlet 10 3.0 0.004 0.018 18.359 18.360 multiply_cannon_loop 2286 14.5 1.199 1.264 16.044 17.511 ot_mini 108 10.5 0.001 0.001 12.196 12.427 mp_waitall_1 200699 16.5 5.595 10.942 5.595 10.942 qs_ot_get_derivative 108 11.5 0.001 0.001 9.820 9.989 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.079 9.668 multiply_cannon_multrec 27432 15.5 1.820 4.238 6.323 9.107 rebuild_ks_matrix 119 8.3 0.000 0.000 6.810 6.956 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.809 6.955 dbcsr_mm_accdrv_process 47894 16.0 3.704 6.026 4.422 6.531 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.979 6.110 qs_ot_get_p 119 10.4 0.001 0.002 4.622 4.844 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.691 4.603 init_scf_run 11 5.9 0.000 0.001 4.172 4.172 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.172 4.172 mp_sum_l 7287 12.8 2.086 4.074 2.086 4.074 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.075 3.735 apply_single 119 13.6 0.000 0.000 3.075 3.735 sum_up_and_integrate 119 10.3 0.001 0.002 3.722 3.730 integrate_v_rspace 119 11.3 0.002 0.005 3.708 3.716 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.636 3.668 calculate_rho_elec 119 8.7 0.021 0.025 3.636 3.668 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.972 2.990 make_m2s 4572 13.5 0.052 0.055 2.641 2.978 make_images 4572 14.5 0.207 0.244 2.553 2.892 calculate_first_density_matrix 1 7.0 0.001 0.005 2.659 2.660 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.573 2.574 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 2.332 2.420 init_scf_loop 11 6.9 0.001 0.004 2.389 2.391 ot_diis_step 108 11.5 0.011 0.014 2.328 2.329 multiply_cannon_sync_h2d 27432 15.5 1.705 2.318 1.705 2.318 calculate_dm_sparse 119 9.5 0.000 0.001 2.198 2.273 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.251 2.252 cp_fm_redistribute_end 50 14.0 1.142 2.227 1.145 2.230 cp_fm_diag_elpa_base 50 14.0 1.053 2.137 1.081 2.182 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.101 2.103 acc_transpose_blocks 27432 15.5 0.114 0.119 1.703 2.074 density_rs2pw 119 9.7 0.004 0.004 1.927 2.007 grid_integrate_task_list 119 12.3 1.841 1.968 1.841 1.968 jit_kernel_multiply 9 16.4 0.659 1.931 0.659 1.931 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.887 1.929 pw_transfer 1439 11.6 0.065 0.068 1.867 1.898 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.776 1.808 make_images_data 4572 15.5 0.048 0.054 1.245 1.598 prepare_preconditioner 11 7.9 0.000 0.000 1.491 1.517 make_preconditioner 11 8.9 0.001 0.004 1.490 1.517 potential_pw2rs 119 12.3 0.006 0.006 1.466 1.474 hybrid_alltoall_any 4725 16.4 0.054 0.116 1.105 1.471 wfi_extrapolate 11 7.9 0.001 0.001 1.455 1.455 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.390 1.450 fft3d_ps 1201 14.6 0.518 0.575 1.412 1.439 fft_wrap_pw1pw2_140 487 13.2 0.160 0.168 1.401 1.437 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.383 1.391 mp_alltoall_d11v 2130 13.8 1.225 1.364 1.225 1.364 grid_collocate_task_list 119 9.7 1.289 1.345 1.289 1.345 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.200 1.251 transfer_rs2pw 487 10.6 0.005 0.006 1.126 1.186 mp_sum_d 4135 12.0 0.565 1.064 0.565 1.064 acc_transpose_blocks_kernels 27432 16.5 0.189 0.279 0.743 0.993 mp_allgather_i34 2286 14.5 0.580 0.984 0.580 0.984 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 0.975 0.976 acc_transpose_blocks_sync 82296 16.5 0.820 0.943 0.820 0.943 transfer_pw2rs 487 13.2 0.005 0.006 0.909 0.914 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.889 0.903 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.249000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.181818, yerr=2.289032 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 521.449472E+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 0.016 0.046 31.701 31.701 qs_mol_dyn_low 1 2.0 0.003 0.003 31.445 31.453 qs_forces 11 3.9 0.002 0.002 31.381 31.382 qs_energies 11 4.9 0.002 0.005 29.843 29.845 scf_env_do_scf 11 5.9 0.001 0.002 25.233 25.233 scf_env_do_scf_inner_loop 108 6.5 0.009 0.063 22.847 22.847 dbcsr_multiply_generic 2286 12.5 0.093 0.098 16.739 16.828 qs_scf_new_mos 108 7.5 0.001 0.001 15.227 15.245 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.226 15.244 velocity_verlet 10 3.0 0.002 0.002 14.935 14.937 ot_scf_mini 108 9.5 0.002 0.003 14.472 14.483 multiply_cannon 2286 13.5 0.193 0.196 13.556 14.305 multiply_cannon_loop 2286 14.5 0.855 0.910 12.819 13.549 ot_mini 108 10.5 0.001 0.001 8.967 8.980 qs_ot_get_derivative 108 11.5 0.001 0.001 7.485 7.498 multiply_cannon_multrec 18288 15.5 1.888 2.944 6.928 7.303 rebuild_ks_matrix 119 8.3 0.000 0.000 5.929 5.944 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 5.928 5.944 dbcsr_mm_accdrv_process 38222 16.0 4.877 5.771 4.946 5.837 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.226 5.240 sum_up_and_integrate 119 10.3 0.001 0.001 3.508 3.514 integrate_v_rspace 119 11.3 0.002 0.003 3.495 3.503 init_scf_run 11 5.9 0.000 0.001 3.422 3.422 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.422 3.422 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.838 3.419 mp_waitall_1 158411 16.6 2.568 3.320 2.568 3.320 qs_ot_get_p 119 10.4 0.001 0.001 3.285 3.304 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.205 3.211 calculate_rho_elec 119 8.7 0.031 0.035 3.204 3.211 init_scf_loop 11 6.9 0.001 0.006 2.370 2.370 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.984 2.349 apply_single 119 13.6 0.000 0.000 1.984 2.349 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.440 2.232 calculate_first_density_matrix 1 7.0 0.000 0.002 2.213 2.214 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.201 2.209 grid_integrate_task_list 119 12.3 1.805 1.936 1.805 1.936 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.908 1.908 make_m2s 4572 13.5 0.044 0.045 1.755 1.885 density_rs2pw 119 9.7 0.004 0.004 1.774 1.862 calculate_dm_sparse 119 9.5 0.000 0.001 1.835 1.843 make_images 4572 14.5 0.193 0.206 1.671 1.799 pw_transfer 1439 11.6 0.065 0.068 1.789 1.796 acc_transpose_blocks 18288 15.5 0.080 0.081 1.617 1.710 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.697 1.705 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.681 1.688 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.664 1.677 cp_fm_diag_elpa_base 50 14.0 1.642 1.654 1.663 1.676 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.665 1.666 prepare_preconditioner 11 7.9 0.000 0.001 1.600 1.602 make_preconditioner 11 8.9 0.001 0.003 1.599 1.602 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.463 1.544 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.512 1.521 ot_diis_step 108 11.5 0.011 0.012 1.463 1.463 mp_sum_l 7287 12.8 1.050 1.417 1.050 1.417 fft_wrap_pw1pw2_140 487 13.2 0.211 0.217 1.373 1.383 potential_pw2rs 119 12.3 0.007 0.009 1.310 1.315 grid_collocate_task_list 119 9.7 1.233 1.313 1.233 1.313 fft3d_ps 1201 14.6 0.524 0.542 1.266 1.275 multiply_cannon_sync_h2d 18288 15.5 1.044 1.214 1.044 1.214 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.175 1.180 wfi_extrapolate 11 7.9 0.001 0.001 1.151 1.151 transfer_rs2pw 487 10.6 0.005 0.005 0.963 1.059 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.960 0.961 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.940 0.958 make_images_data 4572 15.5 0.047 0.051 0.750 0.901 hybrid_alltoall_any 4725 16.4 0.058 0.115 0.654 0.848 acc_transpose_blocks_sync 54864 16.5 0.729 0.821 0.729 0.821 transfer_pw2rs 487 13.2 0.004 0.004 0.799 0.801 acc_transpose_blocks_kernels 18288 16.5 0.218 0.225 0.789 0.796 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.773 0.775 mp_alltoall_d11v 2130 13.8 0.644 0.741 0.644 0.741 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.678 0.738 cp_fm_cholesky_invert 11 10.9 0.681 0.685 0.681 0.685 mp_alltoall_z22v 1201 16.6 0.592 0.657 0.592 0.657 mp_waitany 9880 13.7 0.534 0.644 0.534 0.644 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.701000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.454545, yerr=1.437399 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 556.961792E+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 0.027 0.072 36.712 36.713 qs_mol_dyn_low 1 2.0 0.003 0.004 36.420 36.428 qs_forces 11 3.9 0.002 0.003 36.362 36.363 qs_energies 11 4.9 0.003 0.006 34.641 34.648 scf_env_do_scf 11 5.9 0.001 0.002 29.609 29.610 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 26.179 26.180 dbcsr_multiply_generic 2286 12.5 0.098 0.103 19.903 20.003 velocity_verlet 10 3.0 0.002 0.002 18.556 18.558 qs_scf_new_mos 108 7.5 0.001 0.001 18.037 18.099 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.037 18.098 ot_scf_mini 108 9.5 0.003 0.003 17.004 17.060 multiply_cannon 2286 13.5 0.218 0.227 16.103 16.520 multiply_cannon_loop 2286 14.5 1.534 1.602 15.147 15.458 ot_mini 108 10.5 0.001 0.001 10.407 10.474 multiply_cannon_multrec 27432 15.5 2.459 3.133 9.007 9.265 qs_ot_get_derivative 108 11.5 0.001 0.001 8.542 8.600 dbcsr_mm_accdrv_process 47916 15.9 6.091 7.689 6.443 7.752 rebuild_ks_matrix 119 8.3 0.000 0.000 6.349 6.394 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.349 6.394 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.625 5.663 qs_ot_get_p 119 10.4 0.004 0.025 3.756 3.831 init_scf_run 11 5.9 0.000 0.001 3.653 3.654 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.653 3.654 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.088 3.536 sum_up_and_integrate 119 10.3 0.001 0.001 3.516 3.525 integrate_v_rspace 119 11.3 0.002 0.003 3.505 3.514 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.412 3.451 calculate_rho_elec 119 8.7 0.040 0.046 3.411 3.451 init_scf_loop 11 6.9 0.002 0.006 3.408 3.409 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.153 2.587 apply_single 119 13.6 0.000 0.000 2.152 2.587 prepare_preconditioner 11 7.9 0.000 0.000 2.561 2.569 make_preconditioner 11 8.9 0.001 0.003 2.561 2.569 acc_transpose_blocks 27432 15.5 0.119 0.123 2.409 2.518 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.172 2.498 make_m2s 4572 13.5 0.054 0.057 2.323 2.431 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 2.340 2.369 make_images 4572 14.5 0.276 0.337 2.215 2.323 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.297 2.306 calculate_first_density_matrix 1 7.0 0.001 0.004 2.262 2.265 mp_waitall_1 137007 16.6 1.797 2.232 1.797 2.232 calculate_dm_sparse 119 9.5 0.000 0.000 2.147 2.199 density_rs2pw 119 9.7 0.004 0.004 1.840 2.012 pw_transfer 1439 11.6 0.065 0.068 1.931 1.967 grid_integrate_task_list 119 12.3 1.828 1.910 1.828 1.910 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.890 1.890 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.839 1.879 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.827 1.828 ot_diis_step 108 11.5 0.012 0.012 1.822 1.822 mp_sum_l 7287 12.8 1.085 1.759 1.085 1.759 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.641 1.656 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.594 1.604 cp_fm_diag_elpa_base 50 14.0 1.558 1.576 1.592 1.601 fft_wrap_pw1pw2_140 487 13.2 0.245 0.259 1.507 1.549 acc_transpose_blocks_sync 82296 16.5 1.442 1.545 1.442 1.545 fft3d_ps 1201 14.6 0.553 0.602 1.358 1.381 grid_collocate_task_list 119 9.7 1.249 1.341 1.249 1.341 wfi_extrapolate 11 7.9 0.001 0.001 1.337 1.338 potential_pw2rs 119 12.3 0.009 0.010 1.321 1.325 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.289 1.303 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.242 1.252 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.754 1.218 transfer_rs2pw 487 10.6 0.004 0.005 0.971 1.159 cp_fm_upper_to_full 72 14.2 0.809 1.154 0.809 1.154 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.107 1.108 jit_kernel_multiply 6 16.1 0.286 1.068 0.286 1.068 dbcsr_complete_redistribute 329 12.2 0.126 0.158 0.779 1.052 mp_alltoall_d11v 2130 13.8 0.820 1.013 0.820 1.013 make_images_data 4572 15.5 0.047 0.052 0.863 0.990 hybrid_alltoall_any 4725 16.4 0.065 0.154 0.777 0.965 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.879 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.575 0.841 acc_transpose_blocks_kernels 27432 16.5 0.273 0.282 0.819 0.836 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.816 0.820 transfer_pw2rs 487 13.2 0.004 0.004 0.757 0.760 cp_fm_cholesky_invert 11 10.9 0.741 0.744 0.741 0.744 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.713000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.454545, yerr=3.985511 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/07/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 117.977176E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1384136 0.0% 0.0% 100.0% average stack size 0.0 0.0 106.2 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 622.497792E+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 0.021 0.037 28.220 28.221 qs_mol_dyn_low 1 2.0 0.003 0.003 28.035 28.043 qs_forces 11 3.9 0.002 0.002 27.802 27.802 qs_energies 11 4.9 0.001 0.001 26.101 26.103 scf_env_do_scf 11 5.9 0.000 0.001 21.487 21.487 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.044 19.045 velocity_verlet 10 3.0 0.002 0.002 14.649 14.663 dbcsr_multiply_generic 2286 12.5 0.090 0.094 12.856 12.928 qs_scf_new_mos 108 7.5 0.001 0.001 11.573 11.599 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.572 11.599 ot_scf_mini 108 9.5 0.002 0.002 10.869 10.895 multiply_cannon 2286 13.5 0.221 0.225 10.345 10.843 multiply_cannon_loop 2286 14.5 0.645 0.668 9.439 9.600 ot_mini 108 10.5 0.001 0.001 6.235 6.265 multiply_cannon_multrec 9144 15.5 1.667 1.931 5.914 6.141 rebuild_ks_matrix 119 8.3 0.000 0.000 5.675 5.700 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.674 5.699 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.046 5.066 qs_ot_get_derivative 108 11.5 0.001 0.001 4.960 4.986 dbcsr_mm_accdrv_process 12550 15.8 3.613 4.152 4.134 4.201 sum_up_and_integrate 119 10.3 0.001 0.001 3.376 3.381 integrate_v_rspace 119 11.3 0.003 0.003 3.366 3.372 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.247 3.251 calculate_rho_elec 119 8.7 0.060 0.061 3.246 3.251 init_scf_run 11 5.9 0.000 0.001 3.187 3.187 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.186 3.186 qs_ot_get_p 119 10.4 0.001 0.001 2.772 2.806 init_scf_loop 11 6.9 0.000 0.000 2.423 2.424 calculate_first_density_matrix 1 7.0 0.000 0.000 2.059 2.060 grid_integrate_task_list 119 12.3 1.861 1.935 1.861 1.935 pw_transfer 1439 11.6 0.065 0.068 1.917 1.925 make_m2s 4572 13.5 0.034 0.035 1.745 1.900 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.824 1.834 mp_waitall_1 115863 16.7 1.306 1.824 1.306 1.824 make_images 4572 14.5 0.271 0.302 1.657 1.811 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.806 1.808 density_rs2pw 119 9.7 0.003 0.004 1.657 1.768 calculate_dm_sparse 119 9.5 0.000 0.000 1.731 1.752 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.694 1.718 prepare_preconditioner 11 7.9 0.000 0.000 1.702 1.705 make_preconditioner 11 8.9 0.000 0.000 1.702 1.705 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.594 1.620 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.565 1.566 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.519 1.521 fft_wrap_pw1pw2_140 487 13.2 0.322 0.332 1.506 1.516 acc_transpose_blocks 9144 15.5 0.041 0.043 1.425 1.455 grid_collocate_task_list 119 9.7 1.294 1.395 1.294 1.395 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.355 1.362 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.324 1.339 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.295 1.303 cp_fm_diag_elpa_base 50 14.0 1.268 1.285 1.293 1.301 ot_diis_step 108 11.5 0.013 0.013 1.263 1.263 fft3d_ps 1201 14.6 0.554 0.566 1.238 1.246 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.210 1.242 apply_single 119 13.6 0.000 0.000 1.210 1.241 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.216 1.216 potential_pw2rs 119 12.3 0.010 0.011 1.190 1.193 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.124 1.130 wfi_extrapolate 11 7.9 0.001 0.001 1.078 1.078 jit_kernel_multiply 6 15.6 0.481 1.038 0.481 1.038 hybrid_alltoall_any 4725 16.4 0.065 0.174 0.751 0.965 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.931 make_images_data 4572 15.5 0.042 0.045 0.751 0.915 mp_alltoall_d11v 2130 13.8 0.757 0.895 0.757 0.895 transfer_rs2pw 487 10.6 0.004 0.005 0.757 0.853 cp_fm_cholesky_invert 11 10.9 0.847 0.850 0.847 0.850 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.808 0.815 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.751 0.755 acc_transpose_blocks_sync 27432 16.5 0.724 0.751 0.724 0.751 qs_env_update_s_mstruct 11 6.9 0.003 0.006 0.668 0.715 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.345 0.678 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.644 0.650 transfer_pw2rs 487 13.2 0.003 0.004 0.639 0.641 mp_allgather_i34 2286 14.5 0.224 0.640 0.224 0.640 mp_alltoall_z22v 1201 16.6 0.558 0.592 0.558 0.592 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.221000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=591.545455, yerr=4.075729 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 775.696384E+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 0.016 0.037 41.905 41.907 qs_mol_dyn_low 1 2.0 0.003 0.003 41.704 41.712 qs_forces 11 3.9 0.002 0.003 41.641 41.645 qs_energies 11 4.9 0.001 0.001 39.649 39.654 scf_env_do_scf 11 5.9 0.001 0.001 34.114 34.114 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.396 26.398 velocity_verlet 10 3.0 0.004 0.005 23.742 23.748 dbcsr_multiply_generic 2286 12.5 0.098 0.101 18.775 19.018 qs_scf_new_mos 108 7.5 0.001 0.001 17.075 17.168 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.075 17.167 ot_scf_mini 108 9.5 0.002 0.002 15.919 16.017 multiply_cannon 2286 13.5 0.293 0.296 14.907 15.865 multiply_cannon_loop 2286 14.5 0.844 0.863 13.629 14.619 ot_mini 108 10.5 0.001 0.001 9.669 9.786 multiply_cannon_multrec 9144 15.5 3.317 4.574 8.763 8.885 qs_ot_get_derivative 108 11.5 0.001 0.001 7.603 7.701 init_scf_loop 11 6.9 0.000 0.000 7.692 7.696 rebuild_ks_matrix 119 8.3 0.000 0.000 7.057 7.199 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.056 7.198 prepare_preconditioner 11 7.9 0.000 0.000 6.744 6.757 make_preconditioner 11 8.9 0.000 0.000 6.744 6.757 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.310 6.637 dbcsr_mm_accdrv_process 12550 15.8 4.754 6.475 5.303 6.536 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.355 6.484 cp_fm_upper_to_full 72 14.2 3.169 4.546 3.169 4.546 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.988 3.995 calculate_rho_elec 119 8.7 0.118 0.121 3.987 3.994 sum_up_and_integrate 119 10.3 0.001 0.001 3.752 3.758 integrate_v_rspace 119 11.3 0.003 0.004 3.742 3.748 init_scf_run 11 5.9 0.000 0.001 3.493 3.493 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.493 3.493 qs_ot_get_p 119 10.4 0.001 0.001 3.309 3.441 mp_waitall_1 94719 16.7 2.220 3.216 2.220 3.216 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.778 3.166 dbcsr_complete_redistribute 329 12.2 0.288 0.294 1.965 2.783 pw_transfer 1439 11.6 0.068 0.069 2.561 2.566 make_m2s 4572 13.5 0.037 0.037 2.353 2.528 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.207 2.512 apply_single 119 13.6 0.000 0.000 2.207 2.511 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.463 2.469 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.659 2.462 make_images 4572 14.5 0.356 0.391 2.234 2.409 calculate_dm_sparse 119 9.5 0.000 0.000 2.220 2.236 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.429 2.224 mp_alltoall_i22 627 13.8 1.398 2.221 1.398 2.221 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.279 2.177 grid_integrate_task_list 119 12.3 2.089 2.137 2.089 2.137 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.048 2.100 fft_wrap_pw1pw2_140 487 13.2 0.575 0.579 2.087 2.093 density_rs2pw 119 9.7 0.003 0.004 2.012 2.049 calculate_first_density_matrix 1 7.0 0.000 0.000 2.046 2.046 ot_diis_step 108 11.5 0.014 0.014 2.040 2.040 qs_ot_p2m_diag 50 11.0 0.044 0.045 1.961 1.962 acc_transpose_blocks 9144 15.5 0.044 0.044 1.773 1.806 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.793 1.794 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.764 1.768 mp_sum_l 7287 12.8 0.976 1.744 0.976 1.744 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.609 1.609 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.548 1.589 fft3d_ps 1201 14.6 0.595 0.606 1.560 1.562 grid_collocate_task_list 119 9.7 1.529 1.558 1.529 1.558 cp_fm_cholesky_invert 11 10.9 1.392 1.395 1.392 1.395 wfi_extrapolate 11 7.9 0.001 0.001 1.385 1.385 hybrid_alltoall_any 4725 16.4 0.090 0.151 1.075 1.344 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.308 1.328 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.326 1.326 cp_fm_diag_elpa_base 50 14.0 1.181 1.233 1.325 1.325 potential_pw2rs 119 12.3 0.014 0.015 1.248 1.249 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.225 1.232 make_images_data 4572 15.5 0.046 0.049 1.019 1.231 mp_alltoall_d11v 2130 13.8 1.149 1.186 1.149 1.186 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.096 1.114 acc_transpose_blocks_sync 27432 16.5 1.070 1.105 1.070 1.105 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.000 1.030 qs_create_task_list 11 7.9 0.000 0.000 0.928 0.940 generate_qs_task_list 11 8.9 0.367 0.386 0.928 0.940 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.911 0.924 mp_alltoall_z22v 1201 16.6 0.828 0.845 0.828 0.845 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.907000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=727.272727, yerr=16.074208 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 500.854784E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.039 80.765 80.767 qs_mol_dyn_low 1 2.0 0.003 0.007 80.386 80.396 qs_forces 11 3.9 0.005 0.015 80.316 80.318 qs_energies 11 4.9 0.002 0.006 77.471 77.496 scf_env_do_scf 11 5.9 0.000 0.001 68.606 68.609 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 63.177 63.177 dbcsr_multiply_generic 2055 12.4 0.106 0.109 50.180 50.447 qs_scf_new_mos 99 7.5 0.000 0.001 46.285 46.412 qs_scf_loop_do_ot 99 8.5 0.000 0.001 46.285 46.411 ot_scf_mini 99 9.5 0.002 0.002 43.937 44.027 multiply_cannon 2055 13.4 0.181 0.188 42.196 42.938 velocity_verlet 10 3.0 0.001 0.002 42.237 42.238 multiply_cannon_loop 2055 14.4 1.790 1.828 41.255 42.028 ot_mini 99 10.5 0.001 0.002 26.047 26.134 qs_ot_get_derivative 99 11.5 0.001 0.001 19.233 19.313 multiply_cannon_multrec 49320 15.4 11.456 12.168 17.405 18.088 rebuild_ks_matrix 110 8.3 0.000 0.001 14.417 14.525 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.416 14.524 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.641 12.748 mp_waitall_1 220248 16.4 10.335 11.151 10.335 11.151 multiply_cannon_sync_h2d 49320 15.4 9.672 10.357 9.672 10.357 qs_ot_get_p 110 10.4 0.001 0.001 9.519 9.635 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.318 7.850 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.215 7.726 apply_single 110 13.6 0.000 0.001 7.215 7.726 multiply_cannon_metrocomm3 49320 15.4 0.084 0.087 6.289 7.275 sum_up_and_integrate 110 10.3 0.002 0.003 6.955 6.970 integrate_v_rspace 110 11.3 0.003 0.004 6.930 6.950 init_scf_run 11 5.9 0.000 0.001 6.780 6.780 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.780 6.780 ot_diis_step 99 11.5 0.006 0.009 6.601 6.602 qs_ot_p2m_diag 48 11.0 0.014 0.030 6.481 6.514 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.238 6.376 calculate_rho_elec 110 8.6 0.021 0.025 6.238 6.375 dbcsr_mm_accdrv_process 87628 16.1 3.008 3.075 5.818 6.074 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.631 5.633 init_scf_loop 11 6.9 0.000 0.000 5.402 5.403 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.291 5.344 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.060 5.085 cp_fm_diag_elpa_base 48 14.0 5.047 5.072 5.059 5.083 mp_sum_l 6594 12.7 3.686 4.402 3.686 4.402 wfi_extrapolate 11 7.9 0.001 0.001 4.028 4.028 make_m2s 4110 13.4 0.060 0.065 3.848 3.945 make_images 4110 14.4 0.177 0.190 3.752 3.854 calculate_dm_sparse 110 9.5 0.001 0.001 3.715 3.812 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 3.563 3.570 density_rs2pw 110 9.6 0.004 0.005 3.298 3.492 grid_integrate_task_list 110 12.3 3.253 3.436 3.253 3.436 multiply_cannon_metrocomm1 49320 15.4 0.065 0.068 2.231 3.420 prepare_preconditioner 11 7.9 0.000 0.000 3.257 3.274 make_preconditioner 11 8.9 0.000 0.000 3.257 3.274 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.193 3.232 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.068 3.105 pw_transfer 1331 11.6 0.055 0.068 3.010 3.090 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.046 3.085 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.922 3.005 calculate_first_density_matrix 1 7.0 0.005 0.045 2.654 2.658 fft_wrap_pw1pw2_140 451 13.1 0.390 0.430 2.504 2.602 acc_transpose_blocks 49320 15.4 0.216 0.223 2.556 2.599 potential_pw2rs 110 12.3 0.006 0.006 2.563 2.580 jit_kernel_multiply 13 15.9 2.534 2.545 2.534 2.545 mp_alltoall_d11v 2046 13.8 2.014 2.463 2.014 2.463 fft3d_ps 1111 14.6 0.794 0.884 2.218 2.275 grid_collocate_task_list 110 9.6 2.157 2.257 2.157 2.257 mp_waitany 14300 13.8 1.844 2.118 1.844 2.118 transfer_rs2pw 451 10.6 0.005 0.006 1.903 2.087 mp_sum_d 3889 11.9 1.354 1.978 1.354 1.978 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.906 1.920 make_images_data 4110 15.4 0.043 0.046 1.743 1.881 cp_fm_cholesky_invert 11 10.9 1.806 1.810 1.806 1.810 hybrid_alltoall_any 4261 16.3 0.084 0.488 1.512 1.793 transfer_pw2rs 451 13.1 0.006 0.007 1.747 1.756 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.644 1.676 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=80.767000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.545455, yerr=2.535696 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 592.670720E+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 3025291. 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 0.013 0.032 68.623 68.624 qs_mol_dyn_low 1 2.0 0.003 0.003 68.398 68.408 qs_forces 11 3.9 0.003 0.003 68.334 68.335 qs_energies 11 4.9 0.001 0.002 65.016 65.019 scf_env_do_scf 11 5.9 0.000 0.001 56.573 56.577 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.967 48.967 dbcsr_multiply_generic 2055 12.4 0.114 0.117 37.886 38.086 velocity_verlet 10 3.0 0.001 0.002 36.023 36.024 qs_scf_new_mos 99 7.5 0.001 0.001 33.165 33.307 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.164 33.306 multiply_cannon 2055 13.4 0.223 0.243 31.339 32.494 ot_scf_mini 99 9.5 0.003 0.003 31.502 31.629 multiply_cannon_loop 2055 14.4 1.168 1.194 30.073 30.880 ot_mini 99 10.5 0.001 0.001 18.388 18.530 multiply_cannon_multrec 24660 15.4 7.017 8.872 14.030 16.115 rebuild_ks_matrix 110 8.3 0.000 0.000 13.523 13.599 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.523 13.598 qs_ot_get_derivative 99 11.5 0.001 0.001 12.564 12.693 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.887 11.953 mp_waitall_1 176588 16.5 7.507 9.775 7.507 9.775 multiply_cannon_metrocomm3 24660 15.4 0.073 0.076 5.131 7.768 multiply_cannon_sync_h2d 24660 15.4 6.407 7.617 6.407 7.617 init_scf_loop 11 6.9 0.000 0.000 7.572 7.573 dbcsr_mm_accdrv_process 52282 16.1 5.468 6.212 6.844 7.173 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.522 7.138 apply_single 110 13.6 0.000 0.001 6.521 7.138 qs_ot_get_p 110 10.4 0.001 0.001 6.446 6.601 sum_up_and_integrate 110 10.3 0.002 0.002 6.361 6.373 integrate_v_rspace 110 11.3 0.002 0.003 6.333 6.345 init_scf_run 11 5.9 0.000 0.001 6.082 6.082 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.082 6.082 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.867 5.876 calculate_rho_elec 110 8.6 0.039 0.047 5.866 5.876 ot_diis_step 99 11.5 0.010 0.010 5.777 5.777 prepare_preconditioner 11 7.9 0.000 0.000 5.562 5.574 make_preconditioner 11 8.9 0.000 0.000 5.562 5.574 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.710 5.402 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.152 5.299 make_m2s 4110 13.4 0.057 0.060 4.194 4.657 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.590 4.611 make_images 4110 14.4 0.412 0.472 4.086 4.546 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.121 4.122 pw_transfer 1331 11.6 0.066 0.072 3.517 3.655 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.557 3.571 cp_fm_diag_elpa_base 48 14.0 3.510 3.525 3.554 3.568 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.410 3.551 wfi_extrapolate 11 7.9 0.001 0.001 3.513 3.513 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.268 3.346 grid_integrate_task_list 110 12.3 3.164 3.345 3.164 3.345 density_rs2pw 110 9.6 0.004 0.004 3.111 3.330 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.242 3.243 fft_wrap_pw1pw2_140 451 13.1 0.459 0.474 2.915 3.063 calculate_dm_sparse 110 9.5 0.001 0.001 2.965 2.994 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.900 2.933 hybrid_alltoall_any 4261 16.3 0.106 0.452 2.025 2.818 make_images_data 4110 15.4 0.049 0.052 2.291 2.779 fft3d_ps 1111 14.6 1.110 1.338 2.509 2.659 cp_fm_cholesky_invert 11 10.9 2.585 2.591 2.585 2.591 calculate_first_density_matrix 1 7.0 0.000 0.000 2.472 2.475 mp_sum_l 6594 12.7 1.724 2.347 1.724 2.347 grid_collocate_task_list 110 9.6 2.170 2.300 2.170 2.300 potential_pw2rs 110 12.3 0.009 0.009 2.233 2.245 jit_kernel_multiply 11 16.3 1.011 2.123 1.011 2.123 acc_transpose_blocks 24660 15.4 0.116 0.120 1.984 2.013 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.965 1.980 mp_alltoall_d11v 2046 13.8 1.754 1.938 1.754 1.938 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.855 1.856 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.792 1.798 transfer_rs2pw 451 10.6 0.006 0.007 1.423 1.711 mp_allgather_i34 2055 14.4 0.576 1.606 0.576 1.606 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.537 1.546 multiply_cannon_metrocomm4 22605 15.4 0.078 0.082 0.766 1.539 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.386 1.488 dbcsr_complete_redistribute 325 12.2 0.251 0.314 1.145 1.409 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.624000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=559.181818, yerr=7.517473 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 664.997888E+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 0.016 0.037 59.843 59.844 qs_mol_dyn_low 1 2.0 0.003 0.003 59.572 59.582 qs_forces 11 3.9 0.003 0.003 59.509 59.510 qs_energies 11 4.9 0.001 0.002 56.300 56.304 scf_env_do_scf 11 5.9 0.000 0.001 48.468 48.468 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.821 39.822 velocity_verlet 10 3.0 0.002 0.002 32.679 32.681 dbcsr_multiply_generic 2055 12.4 0.105 0.110 28.694 29.003 qs_scf_new_mos 99 7.5 0.001 0.001 25.218 25.348 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.217 25.347 ot_scf_mini 99 9.5 0.002 0.003 23.963 24.104 multiply_cannon 2055 13.4 0.212 0.226 22.415 23.586 multiply_cannon_loop 2055 14.4 0.822 0.853 21.344 22.527 ot_mini 99 10.5 0.001 0.001 13.747 13.891 rebuild_ks_matrix 110 8.3 0.000 0.000 12.171 12.355 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.170 12.354 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.704 10.870 multiply_cannon_multrec 16440 15.4 3.575 4.586 9.740 10.683 mp_waitall_1 139946 16.5 7.081 10.306 7.081 10.306 qs_ot_get_derivative 99 11.5 0.001 0.001 9.296 9.435 init_scf_loop 11 6.9 0.000 0.000 8.611 8.613 multiply_cannon_metrocomm3 16440 15.4 0.046 0.048 4.500 7.629 prepare_preconditioner 11 7.9 0.000 0.000 6.786 6.805 make_preconditioner 11 8.9 0.000 0.000 6.786 6.805 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.095 6.441 sum_up_and_integrate 110 10.3 0.001 0.002 6.283 6.352 integrate_v_rspace 110 11.3 0.003 0.003 6.256 6.326 dbcsr_mm_accdrv_process 34862 16.1 5.297 5.668 6.011 6.162 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.769 5.778 calculate_rho_elec 110 8.6 0.058 0.059 5.769 5.778 qs_ot_get_p 110 10.4 0.001 0.001 5.389 5.561 init_scf_run 11 5.9 0.000 0.001 5.432 5.433 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.432 5.432 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.906 5.423 apply_single 110 13.6 0.000 0.000 4.905 5.423 make_m2s 4110 13.4 0.049 0.051 4.191 4.478 ot_diis_step 99 11.5 0.011 0.011 4.418 4.418 make_images 4110 14.4 0.400 0.524 4.077 4.365 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.224 3.855 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.750 3.753 multiply_cannon_sync_h2d 16440 15.4 3.189 3.713 3.189 3.713 pw_transfer 1331 11.6 0.066 0.073 3.592 3.650 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.485 3.545 grid_integrate_task_list 110 12.3 3.184 3.392 3.184 3.392 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.377 3.378 density_rs2pw 110 9.6 0.004 0.005 3.011 3.249 fft_wrap_pw1pw2_140 451 13.1 0.578 0.586 3.012 3.072 wfi_extrapolate 11 7.9 0.001 0.001 2.976 2.976 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.848 2.860 cp_fm_diag_elpa_base 48 14.0 2.783 2.807 2.846 2.858 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.765 2.767 make_images_data 4110 15.4 0.045 0.050 2.375 2.730 hybrid_alltoall_any 4261 16.3 0.109 0.385 2.146 2.719 cp_fm_cholesky_invert 11 10.9 2.660 2.666 2.660 2.666 calculate_dm_sparse 110 9.5 0.001 0.001 2.527 2.561 multiply_cannon_metrocomm4 14385 15.4 0.048 0.052 0.912 2.523 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.431 2.510 fft3d_ps 1111 14.6 1.090 1.111 2.428 2.489 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.434 2.488 grid_collocate_task_list 110 9.6 2.214 2.428 2.214 2.428 mp_irecv_dv 48980 15.7 0.836 2.389 0.836 2.389 calculate_first_density_matrix 1 7.0 0.000 0.000 2.364 2.366 potential_pw2rs 110 12.3 0.011 0.012 2.144 2.204 mp_sum_l 6594 12.7 1.475 2.139 1.475 2.139 mp_alltoall_d11v 2046 13.8 1.830 2.071 1.830 2.071 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.959 1.960 dbcsr_complete_redistribute 325 12.2 0.330 0.366 1.459 1.897 cp_fm_upper_to_full 70 14.2 1.363 1.733 1.363 1.733 acc_transpose_blocks 16440 15.4 0.078 0.080 1.595 1.690 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.660 1.681 cp_fm_cholesky_decompose 22 10.9 1.592 1.617 1.592 1.617 transfer_rs2pw 451 10.6 0.005 0.006 1.306 1.555 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.380 1.487 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.462 1.476 mp_allgather_i34 2055 14.4 0.384 1.454 0.384 1.454 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.991 1.421 mp_waitany 17072 13.8 1.137 1.355 1.137 1.355 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.296 1.307 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.844000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.090909, yerr=8.240196 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 735.219712E+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 0.016 0.036 65.107 65.108 qs_mol_dyn_low 1 2.0 0.003 0.003 64.840 64.852 qs_forces 11 3.9 0.003 0.003 64.765 64.768 qs_energies 11 4.9 0.001 0.002 61.391 61.397 scf_env_do_scf 11 5.9 0.000 0.001 53.110 53.112 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.611 41.612 velocity_verlet 10 3.0 0.002 0.002 36.922 36.924 dbcsr_multiply_generic 2055 12.4 0.112 0.117 30.579 30.787 qs_scf_new_mos 99 7.5 0.001 0.001 27.188 27.305 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.188 27.304 ot_scf_mini 99 9.5 0.002 0.003 25.501 25.599 multiply_cannon 2055 13.4 0.236 0.257 23.608 24.743 multiply_cannon_loop 2055 14.4 1.399 1.477 22.207 22.901 ot_mini 99 10.5 0.001 0.001 14.696 14.826 multiply_cannon_multrec 24660 15.4 4.110 6.808 13.143 14.386 rebuild_ks_matrix 110 8.3 0.000 0.000 11.850 11.948 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.849 11.948 init_scf_loop 11 6.9 0.000 0.000 11.459 11.460 qs_ot_get_derivative 99 11.5 0.001 0.001 10.517 10.623 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.464 10.553 dbcsr_mm_accdrv_process 52304 16.0 7.752 9.325 8.875 9.869 prepare_preconditioner 11 7.9 0.000 0.000 9.743 9.761 make_preconditioner 11 8.9 0.000 0.000 9.743 9.761 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.997 9.433 sum_up_and_integrate 110 10.3 0.001 0.002 6.167 6.179 integrate_v_rspace 110 11.3 0.003 0.003 6.141 6.152 mp_waitall_1 121746 16.5 4.119 5.969 4.119 5.969 qs_ot_get_p 110 10.4 0.001 0.001 5.692 5.846 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.748 5.760 calculate_rho_elec 110 8.6 0.078 0.081 5.747 5.759 make_m2s 4110 13.4 0.059 0.063 5.352 5.717 make_images 4110 14.4 0.583 0.704 5.211 5.572 init_scf_run 11 5.9 0.000 0.001 5.565 5.565 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.565 5.565 cp_fm_upper_to_full 70 14.2 3.325 4.810 3.325 4.810 ot_diis_step 99 11.5 0.011 0.011 4.138 4.138 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.055 4.123 apply_single 110 13.6 0.000 0.000 4.055 4.123 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.855 3.872 dbcsr_complete_redistribute 325 12.2 0.425 0.465 2.608 3.739 pw_transfer 1331 11.6 0.065 0.073 3.598 3.632 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.482 3.529 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.492 3.529 grid_integrate_task_list 110 12.3 3.273 3.480 3.273 3.480 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.330 3.393 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.296 3.297 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.095 3.205 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.414 3.203 fft_wrap_pw1pw2_140 451 13.1 0.610 0.632 3.007 3.043 calculate_dm_sparse 110 9.5 0.001 0.001 2.996 3.037 hybrid_alltoall_any 4261 16.3 0.123 0.462 2.263 3.004 make_images_data 4110 15.4 0.049 0.052 2.647 2.994 density_rs2pw 110 9.6 0.004 0.004 2.880 2.984 wfi_extrapolate 11 7.9 0.001 0.001 2.980 2.980 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.902 2.904 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.736 2.833 mp_alltoall_i22 605 13.7 1.626 2.780 1.626 2.780 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.767 2.777 cp_fm_diag_elpa_base 48 14.0 2.622 2.682 2.765 2.776 cp_fm_cholesky_invert 11 10.9 2.660 2.668 2.660 2.668 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.544 2.579 acc_transpose_blocks 24660 15.4 0.114 0.119 2.409 2.559 multiply_cannon_sync_h2d 24660 15.4 2.367 2.512 2.367 2.512 calculate_first_density_matrix 1 7.0 0.000 0.000 2.490 2.493 fft3d_ps 1111 14.6 1.083 1.111 2.373 2.392 grid_collocate_task_list 110 9.6 2.267 2.383 2.267 2.383 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.226 2.228 mp_alltoall_d11v 2046 13.8 1.729 2.032 1.729 2.032 potential_pw2rs 110 12.3 0.013 0.014 1.993 1.997 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.791 1.828 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.613 1.710 mp_sum_l 6594 12.7 0.935 1.671 0.935 1.671 cp_fm_cholesky_decompose 22 10.9 1.606 1.667 1.606 1.667 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.627 1.636 mp_allgather_i34 2055 14.4 0.489 1.586 0.489 1.586 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.552 1.569 jit_kernel_multiply 8 15.9 0.776 1.556 0.776 1.556 acc_transpose_blocks_sync 73980 16.4 1.397 1.553 1.397 1.553 multiply_cannon_metrocomm4 20550 15.4 0.061 0.065 0.848 1.511 mp_irecv_dv 62702 16.1 0.743 1.430 0.743 1.430 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.108000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=696.090909, yerr=8.928550 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 852.557824E+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 0.015 0.035 54.601 54.602 qs_mol_dyn_low 1 2.0 0.003 0.003 54.368 54.378 qs_forces 11 3.9 0.003 0.003 54.307 54.308 qs_energies 11 4.9 0.001 0.002 50.671 50.674 scf_env_do_scf 11 5.9 0.000 0.001 42.528 42.528 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.907 34.908 velocity_verlet 10 3.0 0.002 0.002 30.874 30.877 dbcsr_multiply_generic 2055 12.4 0.104 0.108 23.156 23.288 qs_scf_new_mos 99 7.5 0.001 0.001 20.555 20.629 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.555 20.628 ot_scf_mini 99 9.5 0.002 0.002 19.316 19.363 multiply_cannon 2055 13.4 0.235 0.248 17.731 18.844 multiply_cannon_loop 2055 14.4 0.606 0.624 16.518 16.866 rebuild_ks_matrix 110 8.3 0.000 0.000 11.491 11.535 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.491 11.535 ot_mini 99 10.5 0.001 0.001 10.780 10.823 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.229 10.269 multiply_cannon_multrec 8220 15.4 3.198 4.644 7.588 8.816 init_scf_loop 11 6.9 0.000 0.000 7.575 7.577 mp_waitall_1 103326 16.6 5.901 7.575 5.901 7.575 qs_ot_get_derivative 99 11.5 0.001 0.001 7.026 7.072 sum_up_and_integrate 110 10.3 0.002 0.002 6.140 6.151 integrate_v_rspace 110 11.3 0.003 0.004 6.114 6.125 prepare_preconditioner 11 7.9 0.000 0.000 5.936 5.942 make_preconditioner 11 8.9 0.000 0.000 5.935 5.942 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.885 5.902 calculate_rho_elec 110 8.6 0.114 0.114 5.885 5.901 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.526 5.603 dbcsr_mm_accdrv_process 17442 15.9 3.064 4.115 4.250 5.165 init_scf_run 11 5.9 0.000 0.001 5.031 5.031 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.031 5.031 qs_ot_get_p 110 10.4 0.001 0.001 4.617 4.685 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 3.129 4.535 make_m2s 4110 13.4 0.038 0.039 4.107 4.303 make_images 4110 14.4 0.652 0.713 3.978 4.169 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.795 3.857 apply_single 110 13.6 0.000 0.000 3.794 3.856 pw_transfer 1331 11.6 0.065 0.070 3.787 3.794 ot_diis_step 99 11.5 0.012 0.012 3.732 3.732 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.680 3.691 grid_integrate_task_list 110 12.3 3.383 3.565 3.383 3.565 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.274 3.277 fft_wrap_pw1pw2_140 451 13.1 0.779 0.794 3.199 3.216 density_rs2pw 110 9.6 0.004 0.004 2.864 2.978 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.939 2.940 cp_fm_cholesky_invert 11 10.9 2.846 2.850 2.846 2.850 wfi_extrapolate 11 7.9 0.001 0.001 2.680 2.680 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.658 2.659 hybrid_alltoall_any 4261 16.3 0.200 0.852 2.207 2.587 grid_collocate_task_list 110 9.6 2.369 2.542 2.369 2.542 calculate_dm_sparse 110 9.5 0.001 0.001 2.467 2.503 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.494 2.496 multiply_cannon_sync_h2d 8220 15.4 2.350 2.458 2.350 2.458 make_images_data 4110 15.4 0.040 0.046 2.171 2.435 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.379 2.386 cp_fm_diag_elpa_base 48 14.0 2.324 2.346 2.377 2.385 fft3d_ps 1111 14.6 1.133 1.151 2.343 2.360 calculate_first_density_matrix 1 7.0 0.000 0.000 2.257 2.258 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.106 2.132 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.993 2.028 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.789 2.014 mp_alltoall_d11v 2046 13.8 1.705 1.972 1.705 1.972 potential_pw2rs 110 12.3 0.016 0.016 1.948 1.951 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.897 1.925 cp_fm_cholesky_decompose 22 10.9 1.655 1.677 1.655 1.677 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.485 1.604 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.593 1.597 jit_kernel_multiply 9 15.6 0.869 1.524 0.869 1.524 dbcsr_complete_redistribute 325 12.2 0.587 0.603 1.434 1.519 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.433 1.443 mp_allgather_i34 2055 14.4 0.384 1.406 0.384 1.406 acc_transpose_blocks 8220 15.4 0.039 0.040 1.362 1.379 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.483 1.351 mp_irecv_dv 24056 15.7 0.456 1.309 0.456 1.309 qs_create_task_list 11 7.9 0.001 0.001 1.208 1.307 generate_qs_task_list 11 8.9 0.372 0.440 1.207 1.307 transfer_rs2pw 451 10.6 0.005 0.006 1.085 1.231 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.195 1.230 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.787 1.222 mp_waitany 9240 13.8 1.045 1.148 1.045 1.148 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.602000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=806.909091, yerr=11.873994 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 1.392919E+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 0.018 0.041 86.275 86.276 qs_mol_dyn_low 1 2.0 0.003 0.003 85.974 85.984 qs_forces 11 3.9 0.003 0.004 85.910 85.911 qs_energies 11 4.9 0.001 0.002 81.784 81.786 scf_env_do_scf 11 5.9 0.000 0.001 71.885 71.885 velocity_verlet 10 3.0 0.002 0.002 55.269 55.275 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.890 43.891 dbcsr_multiply_generic 2055 12.4 0.118 0.124 29.598 29.763 init_scf_loop 11 6.9 0.000 0.000 27.925 27.928 qs_scf_new_mos 99 7.5 0.001 0.001 26.819 26.905 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.818 26.904 prepare_preconditioner 11 7.9 0.000 0.000 25.949 25.962 make_preconditioner 11 8.9 0.000 0.000 25.949 25.962 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.223 25.398 ot_scf_mini 99 9.5 0.002 0.002 25.036 25.107 multiply_cannon 2055 13.4 0.336 0.360 22.447 23.047 multiply_cannon_loop 2055 14.4 0.829 0.842 20.650 21.113 cp_fm_upper_to_full 70 14.2 12.593 18.010 12.593 18.010 ot_mini 99 10.5 0.001 0.001 14.278 14.351 rebuild_ks_matrix 110 8.3 0.000 0.000 13.459 13.553 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.458 13.552 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.209 12.295 dbcsr_complete_redistribute 325 12.2 1.028 1.063 7.300 10.546 multiply_cannon_multrec 8220 15.4 4.057 4.185 9.716 9.874 qs_ot_get_derivative 99 11.5 0.001 0.001 9.534 9.608 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.267 9.510 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.711 8.920 mp_alltoall_i22 605 13.7 5.340 8.618 5.340 8.618 mp_waitall_1 84994 16.7 7.526 8.526 7.526 8.526 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.970 7.003 calculate_rho_elec 110 8.6 0.225 0.225 6.970 7.002 sum_up_and_integrate 110 10.3 0.002 0.002 6.651 6.664 integrate_v_rspace 110 11.3 0.004 0.004 6.623 6.636 make_m2s 4110 13.4 0.043 0.044 5.329 5.843 dbcsr_mm_accdrv_process 11614 15.7 3.908 4.092 5.512 5.706 init_scf_run 11 5.9 0.000 0.001 5.657 5.657 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.656 5.656 make_images 4110 14.4 0.889 0.941 5.141 5.655 qs_ot_get_p 110 10.4 0.001 0.001 5.484 5.570 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.857 5.368 apply_single 110 13.6 0.000 0.000 4.857 5.368 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.899 5.263 cp_fm_cholesky_invert 11 10.9 5.130 5.134 5.130 5.134 pw_transfer 1331 11.6 0.074 0.074 4.826 4.832 ot_diis_step 99 11.5 0.015 0.016 4.728 4.728 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.709 4.716 fft_wrap_pw1pw2_140 451 13.1 1.277 1.282 4.124 4.131 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.900 3.907 grid_integrate_task_list 110 12.3 3.678 3.726 3.678 3.726 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.642 3.643 hybrid_alltoall_any 4261 16.3 0.263 0.566 2.823 3.545 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.004 3.481 density_rs2pw 110 9.6 0.004 0.004 3.441 3.459 make_images_data 4110 15.4 0.045 0.048 2.804 3.428 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.386 3.386 wfi_extrapolate 11 7.9 0.001 0.001 3.299 3.299 calculate_dm_sparse 110 9.5 0.001 0.001 3.149 3.172 multiply_cannon_sync_h2d 8220 15.4 3.119 3.152 3.119 3.152 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.889 2.894 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.810 2.811 cp_fm_diag_elpa_base 48 14.0 2.275 2.469 2.809 2.809 fft3d_ps 1111 14.6 1.297 1.305 2.766 2.774 grid_collocate_task_list 110 9.6 2.667 2.689 2.667 2.689 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.642 2.684 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.375 2.406 calculate_first_density_matrix 1 7.0 0.000 0.000 2.243 2.244 qs_env_update_s_mstruct 11 6.9 0.003 0.005 2.159 2.215 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.093 2.208 potential_pw2rs 110 12.3 0.022 0.022 2.190 2.195 mp_alltoall_d11v 2046 13.8 2.024 2.076 2.024 2.076 cp_fm_cholesky_decompose 22 10.9 1.932 1.951 1.932 1.951 qs_create_task_list 11 7.9 0.000 0.000 1.877 1.921 generate_qs_task_list 11 8.9 0.729 0.783 1.877 1.921 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.850 1.859 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.795 1.847 acc_transpose_blocks 8220 15.4 0.041 0.041 1.726 1.744 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.276000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1237.181818, yerr=57.432677 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 635.125760E+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 4020 57505. MP_Allreduce 11129 795. MP_Sync 87 MP_Alltoall 2226 2529087. 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 0.014 0.041 206.294 206.297 qs_mol_dyn_low 1 2.0 0.003 0.003 205.856 205.871 qs_forces 11 3.9 0.005 0.005 205.772 205.773 qs_energies 11 4.9 0.001 0.002 200.147 200.160 scf_env_do_scf 11 5.9 0.001 0.001 183.757 183.761 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 163.046 163.048 dbcsr_multiply_generic 2507 12.6 0.179 0.183 125.277 126.054 qs_scf_new_mos 117 7.6 0.001 0.001 123.801 124.065 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.801 124.065 velocity_verlet 10 3.0 0.001 0.002 123.628 123.629 ot_scf_mini 117 9.6 0.003 0.003 117.062 117.333 multiply_cannon 2507 13.6 0.239 0.249 101.442 103.642 multiply_cannon_loop 2507 14.6 2.405 2.461 99.260 101.030 ot_mini 117 10.6 0.001 0.001 65.851 66.098 multiply_cannon_multrec 60168 15.6 31.829 33.918 41.558 43.670 qs_ot_get_derivative 117 11.6 0.001 0.001 41.100 41.328 rebuild_ks_matrix 128 8.3 0.001 0.001 33.501 33.856 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.501 33.855 mp_waitall_1 267128 16.5 28.673 31.542 28.673 31.542 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.075 30.403 qs_ot_get_p 128 10.4 0.001 0.001 28.983 29.244 multiply_cannon_sync_h2d 60168 15.6 26.345 28.607 26.345 28.607 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.240 24.970 apply_single 128 13.6 0.001 0.001 24.240 24.969 ot_diis_step 117 11.6 0.008 0.008 24.512 24.513 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.359 22.448 init_scf_loop 11 6.9 0.000 0.000 20.639 20.640 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.710 19.888 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.562 19.562 multiply_cannon_metrocomm3 60168 15.6 0.118 0.121 15.919 18.677 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.624 16.658 cp_fm_diag_elpa_base 83 14.4 16.557 16.591 16.621 16.655 prepare_preconditioner 11 7.9 0.000 0.000 16.069 16.118 make_preconditioner 11 8.9 0.000 0.000 16.069 16.118 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.290 15.497 make_m2s 5014 13.6 0.104 0.113 13.697 14.057 sum_up_and_integrate 128 10.3 0.002 0.004 13.966 13.985 integrate_v_rspace 128 11.3 0.003 0.004 13.908 13.928 make_images 5014 14.6 0.397 0.418 13.514 13.889 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.029 13.167 calculate_rho_elec 128 8.7 0.045 0.064 13.028 13.166 init_scf_run 11 5.9 0.000 0.002 12.254 12.255 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.254 12.254 mp_sum_l 7950 12.9 8.883 9.923 8.883 9.923 dbcsr_mm_accdrv_process 124484 16.2 4.687 4.860 9.298 9.847 cp_fm_cholesky_invert 11 10.9 9.287 9.295 9.287 9.295 wfi_extrapolate 11 7.9 0.001 0.001 9.049 9.049 calculate_dm_sparse 128 9.5 0.001 0.001 8.571 8.659 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.137 8.223 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.989 8.123 make_images_data 5014 15.6 0.067 0.073 6.726 7.740 pw_transfer 1547 11.6 0.074 0.089 7.518 7.692 multiply_cannon_metrocomm1 60168 15.6 0.095 0.099 6.308 7.530 grid_integrate_task_list 128 12.3 7.092 7.519 7.092 7.519 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.316 7.480 density_rs2pw 128 9.7 0.006 0.007 6.679 7.315 hybrid_alltoall_any 5200 16.5 0.295 2.266 5.895 7.215 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.625 6.635 fft_wrap_pw1pw2_140 523 13.2 1.125 1.177 6.451 6.616 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.764 5.856 fft3d_ps 1291 14.7 2.193 2.854 5.321 5.656 mp_alltoall_d11v 2415 14.1 4.263 5.487 4.263 5.487 grid_collocate_task_list 128 9.7 4.840 5.361 4.840 5.361 cp_fm_cholesky_decompose 22 10.9 4.570 4.587 4.570 4.587 potential_pw2rs 128 12.3 0.009 0.010 4.469 4.490 mp_sum_d 4479 12.1 3.485 4.478 3.485 4.478 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.297000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.636364, yerr=6.984043 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/16/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 2.183246E+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 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 839.229440E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 58100. MP_Allreduce 11103 959. MP_Sync 87 MP_Alltoall 1969 5014482. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.034 189.350 189.352 qs_mol_dyn_low 1 2.0 0.003 0.003 188.972 188.988 qs_forces 11 3.9 0.004 0.005 188.891 188.892 qs_energies 11 4.9 0.001 0.002 182.170 182.179 scf_env_do_scf 11 5.9 0.001 0.001 165.819 165.829 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 132.919 132.922 velocity_verlet 10 3.0 0.001 0.002 119.063 119.064 dbcsr_multiply_generic 2507 12.6 0.188 0.195 97.099 98.500 qs_scf_new_mos 117 7.6 0.001 0.001 94.538 95.016 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.537 95.016 ot_scf_mini 117 9.6 0.004 0.004 89.737 90.272 multiply_cannon 2507 13.6 0.501 0.554 76.891 80.942 multiply_cannon_loop 2507 14.6 1.576 1.647 73.981 76.874 ot_mini 117 10.6 0.001 0.001 50.351 50.929 mp_waitall_1 214728 16.6 24.179 38.206 24.179 38.206 multiply_cannon_multrec 30084 15.6 20.865 25.579 31.533 36.639 rebuild_ks_matrix 128 8.3 0.001 0.001 32.072 32.813 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.072 32.813 init_scf_loop 11 6.9 0.000 0.000 32.810 32.811 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.848 29.519 qs_ot_get_derivative 117 11.6 0.001 0.001 28.314 28.841 multiply_cannon_metrocomm3 30084 15.6 0.098 0.105 15.901 28.831 prepare_preconditioner 11 7.9 0.000 0.000 28.542 28.605 make_preconditioner 11 8.9 0.000 0.000 28.542 28.605 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.207 27.758 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.102 23.284 apply_single 128 13.6 0.001 0.001 22.102 23.284 qs_ot_get_p 128 10.4 0.001 0.001 21.584 22.269 ot_diis_step 117 11.6 0.014 0.015 21.856 21.858 multiply_cannon_sync_h2d 30084 15.6 17.832 20.019 17.832 20.019 qs_ot_p2m_diag 83 11.4 0.189 0.217 16.643 16.678 cp_fm_cholesky_invert 11 10.9 16.634 16.646 16.634 16.646 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.424 15.425 make_m2s 5014 13.6 0.090 0.096 14.104 15.281 make_images 5014 14.6 1.180 1.376 13.893 15.070 sum_up_and_integrate 128 10.3 0.002 0.004 13.984 14.016 integrate_v_rspace 128 11.3 0.003 0.004 13.923 13.958 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.360 13.390 calculate_rho_elec 128 8.7 0.087 0.104 13.359 13.390 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.228 12.257 cp_fm_diag_elpa_base 83 14.4 11.973 12.068 12.223 12.247 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.169 11.553 init_scf_run 11 5.9 0.000 0.001 11.530 11.532 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.530 11.532 multiply_cannon_metrocomm4 27577 15.6 0.105 0.117 3.897 10.972 dbcsr_mm_accdrv_process 62242 16.2 5.490 6.246 10.117 10.798 mp_irecv_dv 69486 16.3 3.692 10.566 3.692 10.566 make_images_data 5014 15.6 0.067 0.076 8.264 9.703 hybrid_alltoall_any 5200 16.5 0.352 1.545 7.059 9.429 pw_transfer 1547 11.6 0.085 0.096 8.573 8.629 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.350 8.412 wfi_extrapolate 11 7.9 0.001 0.001 8.313 8.313 density_rs2pw 128 9.7 0.006 0.007 7.061 7.626 fft_wrap_pw1pw2_140 523 13.2 1.214 1.238 7.391 7.473 grid_integrate_task_list 128 12.3 7.186 7.460 7.186 7.460 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.430 7.233 cp_fm_cholesky_decompose 22 10.9 6.937 7.009 6.937 7.009 calculate_dm_sparse 128 9.5 0.001 0.001 6.453 6.599 mp_sum_l 7950 12.9 4.306 6.427 4.306 6.427 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.176 6.185 fft3d_ps 1291 14.7 2.841 3.008 5.908 5.950 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.289 5.480 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.369 5.444 grid_collocate_task_list 128 9.7 5.006 5.410 5.006 5.410 mp_alltoall_d11v 2415 14.1 4.194 4.742 4.194 4.742 potential_pw2rs 128 12.3 0.016 0.019 4.478 4.497 mp_allgather_i34 2507 14.6 1.419 4.412 1.419 4.412 mp_sum_d 4474 12.1 2.687 4.223 2.687 4.223 dbcsr_complete_redistribute 395 12.7 0.775 0.873 3.185 4.028 dbcsr_dot_sd 1318 12.0 0.753 0.833 2.505 3.805 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.352000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=800.000000, yerr=3.162278 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.950886E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 4015680 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.9 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 941.383680E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1051232 MPI messages size (bytes): total size 2.737039E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.603648E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 284160 37245419520 131072 < size <= 4194304 665408 1004401590272 4194304 < size <= 16777216 66080 937889764224 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57897. MP_Allreduce 11137 997. MP_Sync 87 MP_Alltoall 1724 9394172. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.033 173.793 173.795 qs_mol_dyn_low 1 2.0 0.003 0.003 173.407 173.423 qs_forces 11 3.9 0.004 0.005 173.314 173.315 qs_energies 11 4.9 0.001 0.002 166.787 166.800 scf_env_do_scf 11 5.9 0.001 0.001 151.343 151.345 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 116.974 116.975 velocity_verlet 10 3.0 0.002 0.002 112.008 112.010 dbcsr_multiply_generic 2527 12.6 0.180 0.184 82.281 83.487 qs_scf_new_mos 118 7.6 0.001 0.001 79.813 80.122 qs_scf_loop_do_ot 118 8.6 0.001 0.001 79.812 80.121 ot_scf_mini 118 9.6 0.003 0.004 75.519 75.911 multiply_cannon 2527 13.6 0.514 0.531 62.657 67.509 multiply_cannon_loop 2527 14.6 1.138 1.211 59.588 62.385 ot_mini 118 10.6 0.001 0.001 42.541 42.916 init_scf_loop 11 6.9 0.000 0.000 34.272 34.273 mp_waitall_1 171878 16.6 24.571 33.767 24.571 33.767 rebuild_ks_matrix 129 8.3 0.001 0.001 30.309 30.767 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.017 30.309 30.766 prepare_preconditioner 11 7.9 0.000 0.000 30.294 30.343 make_preconditioner 11 8.9 0.000 0.000 30.294 30.343 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.959 29.357 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.342 27.758 multiply_cannon_multrec 20216 15.6 13.282 16.149 22.642 25.536 multiply_cannon_metrocomm3 20216 15.6 0.063 0.067 15.271 24.848 qs_ot_get_derivative 118 11.6 0.002 0.002 22.893 23.276 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.657 20.622 apply_single 129 13.6 0.001 0.001 19.656 20.622 ot_diis_step 118 11.6 0.018 0.020 19.533 19.533 qs_ot_get_p 129 10.4 0.001 0.001 18.170 18.563 make_m2s 5054 13.6 0.080 0.086 14.702 15.995 make_images 5054 14.6 1.161 1.246 14.467 15.755 multiply_cannon_sync_h2d 20216 15.6 13.799 15.093 13.799 15.093 cp_fm_cholesky_invert 11 10.9 14.087 14.095 14.087 14.095 qs_ot_p2m_diag 83 11.4 0.265 0.272 13.923 13.931 sum_up_and_integrate 129 10.3 0.002 0.003 13.882 13.909 integrate_v_rspace 129 11.3 0.003 0.003 13.822 13.850 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.489 13.531 calculate_rho_elec 129 8.7 0.131 0.146 13.489 13.530 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.884 12.884 make_images_data 5054 15.6 0.063 0.073 8.809 10.522 init_scf_run 11 5.9 0.000 0.001 10.413 10.414 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.413 10.413 hybrid_alltoall_any 5240 16.5 0.452 2.037 7.765 9.983 cp_fm_diag_elpa 83 13.4 0.000 0.001 9.764 9.787 cp_fm_diag_elpa_base 83 14.4 9.364 9.504 9.761 9.784 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.976 9.241 multiply_cannon_metrocomm4 17689 15.6 0.066 0.075 3.489 9.102 dbcsr_mm_accdrv_process 41830 16.2 5.617 5.909 8.818 8.972 mp_irecv_dv 50625 16.2 3.359 8.850 3.359 8.850 pw_transfer 1559 11.6 0.086 0.103 8.644 8.758 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.418 8.538 grid_integrate_task_list 129 12.3 7.366 7.773 7.366 7.773 fft_wrap_pw1pw2_140 527 13.2 1.313 1.354 7.505 7.636 wfi_extrapolate 11 7.9 0.001 0.001 7.433 7.433 density_rs2pw 129 9.7 0.006 0.006 6.836 7.266 cp_fm_cholesky_decompose 22 10.9 7.179 7.196 7.179 7.196 cp_fm_upper_to_full 105 14.8 5.616 7.148 5.616 7.148 dbcsr_complete_redistribute 395 12.7 1.171 1.205 4.496 6.210 calculate_dm_sparse 129 9.5 0.001 0.001 5.886 5.999 fft3d_ps 1301 14.7 2.779 3.008 5.775 5.852 grid_collocate_task_list 129 9.7 5.191 5.675 5.191 5.675 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.465 5.471 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.684 5.318 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.355 5.069 mp_allgather_i34 2527 14.6 1.557 4.796 1.557 4.796 mp_alltoall_d11v 2423 14.1 4.169 4.736 4.169 4.736 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.587 4.726 mp_sum_l 8010 12.9 3.076 4.590 3.076 4.590 potential_pw2rs 129 12.3 0.021 0.023 4.278 4.287 transfer_fm_to_dbcsr 11 9.9 0.019 0.024 2.316 4.025 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.916 3.945 mp_alltoall_i22 718 14.1 1.941 3.850 1.941 3.850 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.749 3.750 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.470 3.511 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=173.795000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=890.636364, yerr=8.926699 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.172857E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58132. MP_Allreduce 11095 1083. MP_Sync 86 MP_Alltoall 1700 12496371. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.071 184.719 184.724 qs_mol_dyn_low 1 2.0 0.003 0.003 184.183 184.196 qs_forces 11 3.9 0.005 0.015 184.083 184.087 qs_energies 11 4.9 0.003 0.011 177.021 177.031 scf_env_do_scf 11 5.9 0.001 0.001 160.320 160.333 velocity_verlet 10 3.0 0.002 0.002 122.137 122.140 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 114.168 114.170 dbcsr_multiply_generic 2485 12.5 0.185 0.192 79.123 79.760 qs_scf_new_mos 116 7.6 0.001 0.001 78.558 78.891 qs_scf_loop_do_ot 116 8.6 0.001 0.001 78.557 78.890 ot_scf_mini 116 9.6 0.003 0.003 74.043 74.364 multiply_cannon 2485 13.5 0.549 0.584 55.338 57.822 multiply_cannon_loop 2485 14.5 1.824 1.888 51.672 53.156 init_scf_loop 11 6.9 0.002 0.016 46.029 46.030 ot_mini 116 10.6 0.001 0.001 42.129 42.441 prepare_preconditioner 11 7.9 0.000 0.001 41.980 42.014 make_preconditioner 11 8.9 0.001 0.003 41.980 42.014 make_full_inverse_cholesky 11 9.9 0.009 0.022 35.716 40.751 multiply_cannon_multrec 29820 15.5 13.548 18.644 26.342 31.047 rebuild_ks_matrix 127 8.3 0.001 0.001 28.824 29.089 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.020 28.824 29.089 mp_waitall_1 146592 16.7 16.500 26.494 16.500 26.494 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.006 26.248 qs_ot_get_derivative 116 11.6 0.001 0.002 22.870 23.194 make_m2s 4970 13.5 0.095 0.099 19.555 20.509 make_images 4970 14.5 1.922 2.160 19.248 20.205 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.728 19.352 apply_single 127 13.6 0.001 0.001 18.728 19.352 ot_diis_step 116 11.6 0.017 0.018 19.130 19.131 qs_ot_get_p 127 10.4 0.001 0.001 18.308 18.689 cp_fm_upper_to_full 104 14.8 11.347 16.780 11.347 16.780 cp_fm_cholesky_invert 11 10.9 16.192 16.201 16.192 16.201 multiply_cannon_metrocomm3 29820 15.5 0.050 0.052 6.319 14.696 qs_ot_p2m_diag 82 11.4 0.339 0.385 14.290 14.341 sum_up_and_integrate 127 10.3 0.002 0.005 13.773 13.799 integrate_v_rspace 127 11.3 0.004 0.008 13.713 13.741 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.484 13.523 calculate_rho_elec 127 8.7 0.171 0.187 13.484 13.522 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.908 12.910 dbcsr_mm_accdrv_process 61748 16.2 8.356 9.215 12.369 12.861 dbcsr_complete_redistribute 393 12.7 1.505 1.629 8.833 12.612 make_images_data 4970 15.5 0.066 0.072 10.355 11.937 multiply_cannon_sync_h2d 29820 15.5 10.396 11.309 10.396 11.309 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.457 11.250 hybrid_alltoall_any 5155 16.4 0.521 2.204 9.398 10.906 init_scf_run 11 5.9 0.000 0.001 10.659 10.661 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.659 10.661 transfer_fm_to_dbcsr 11 9.9 0.002 0.014 6.241 9.933 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.881 9.887 cp_fm_diag_elpa_base 82 14.4 8.934 9.243 9.875 9.884 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.561 9.789 mp_alltoall_i22 712 14.1 5.547 9.210 5.547 9.210 pw_transfer 1535 11.6 0.085 0.100 8.779 8.866 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.555 8.649 grid_integrate_task_list 127 12.3 7.495 7.880 7.495 7.880 fft_wrap_pw1pw2_140 519 13.2 1.434 1.471 7.639 7.741 multiply_cannon_metrocomm4 24850 15.5 0.083 0.094 2.780 7.718 cp_fm_cholesky_decompose 22 10.9 7.439 7.566 7.439 7.566 wfi_extrapolate 11 7.9 0.001 0.001 7.520 7.520 mp_irecv_dv 75445 16.2 2.626 7.429 2.626 7.429 density_rs2pw 127 9.7 0.006 0.006 6.613 6.980 calculate_dm_sparse 127 9.5 0.001 0.001 6.208 6.276 fft3d_ps 1281 14.7 2.830 2.917 5.755 5.807 grid_collocate_task_list 127 9.7 5.252 5.519 5.252 5.519 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.398 5.432 mp_alltoall_d11v 2401 14.1 4.600 4.908 4.600 4.908 qs_energies_init_hamiltonians 11 5.9 0.004 0.022 4.531 4.534 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.433 4.529 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.330 4.427 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.188 4.244 potential_pw2rs 127 12.3 0.023 0.023 4.168 4.176 dbcsr_special_finalize 12425 15.5 0.066 0.069 3.493 3.710 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=184.724000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1099.454545, yerr=19.592966 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/19/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 5.910120E+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 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.546600E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526748E+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 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592466352 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57624. MP_Allreduce 11184 1163. MP_Sync 88 MP_Alltoall 1724 18848034. 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 0.017 0.035 169.892 169.893 qs_mol_dyn_low 1 2.0 0.003 0.003 169.502 169.519 qs_forces 11 3.9 0.004 0.005 169.405 169.409 qs_energies 11 4.9 0.002 0.002 162.035 162.044 scf_env_do_scf 11 5.9 0.001 0.001 144.740 144.751 velocity_verlet 10 3.0 0.002 0.002 112.073 112.076 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 109.269 109.271 dbcsr_multiply_generic 2529 12.6 0.181 0.192 72.953 73.397 qs_scf_new_mos 118 7.6 0.001 0.001 72.377 72.499 qs_scf_loop_do_ot 118 8.6 0.001 0.001 72.377 72.499 ot_scf_mini 118 9.6 0.003 0.004 67.913 68.040 multiply_cannon 2529 13.6 0.562 0.585 54.009 57.275 multiply_cannon_loop 2529 14.6 0.819 0.854 51.033 52.039 ot_mini 118 10.6 0.001 0.001 37.988 38.114 init_scf_loop 11 6.9 0.000 0.000 35.314 35.316 mp_waitall_1 126876 16.7 25.445 32.325 25.445 32.325 prepare_preconditioner 11 7.9 0.000 0.000 31.463 31.485 make_preconditioner 11 8.9 0.000 0.000 31.463 31.485 make_full_inverse_cholesky 11 9.9 0.017 0.028 29.372 29.634 rebuild_ks_matrix 129 8.3 0.001 0.001 29.142 29.277 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.141 29.277 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.597 26.718 multiply_cannon_multrec 10116 15.6 10.375 15.185 18.003 22.602 multiply_cannon_metrocomm3 10116 15.6 0.026 0.028 13.226 21.176 qs_ot_get_derivative 118 11.6 0.001 0.002 20.647 20.769 cp_fm_cholesky_invert 11 10.9 17.981 17.987 17.981 17.987 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.356 17.500 apply_single 129 13.6 0.001 0.001 17.356 17.500 ot_diis_step 118 11.6 0.020 0.021 17.272 17.273 qs_ot_get_p 129 10.4 0.001 0.001 17.005 17.137 make_m2s 5058 13.6 0.066 0.070 15.070 15.922 make_images 5058 14.6 2.197 2.626 14.762 15.618 sum_up_and_integrate 129 10.3 0.002 0.002 14.098 14.149 integrate_v_rspace 129 11.3 0.004 0.004 14.038 14.087 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.034 14.060 calculate_rho_elec 129 8.7 0.258 0.269 14.033 14.060 qs_ot_p2m_diag 84 11.4 0.502 0.507 13.237 13.252 cp_dbcsr_syevd 84 12.4 0.005 0.006 12.051 12.052 multiply_cannon_sync_h2d 10116 15.6 10.777 11.320 10.777 11.320 init_scf_run 11 5.9 0.000 0.004 10.442 10.443 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.442 10.442 make_images_data 5058 15.6 0.056 0.065 8.731 9.979 hybrid_alltoall_any 5245 16.5 0.848 3.790 8.513 9.903 pw_transfer 1559 11.6 0.086 0.094 9.268 9.300 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.148 9.161 cp_fm_diag_elpa_base 84 14.4 8.910 8.982 9.145 9.157 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.043 9.080 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.165 8.264 grid_integrate_task_list 129 12.3 7.822 8.109 7.822 8.109 fft_wrap_pw1pw2_140 527 13.2 1.818 1.850 8.025 8.066 dbcsr_mm_accdrv_process 20934 16.1 3.372 4.494 7.276 7.989 cp_fm_cholesky_decompose 22 10.9 7.826 7.950 7.826 7.950 multiply_cannon_metrocomm1 10116 15.6 0.030 0.031 4.278 7.544 wfi_extrapolate 11 7.9 0.001 0.001 7.320 7.320 density_rs2pw 129 9.7 0.006 0.006 6.677 7.000 calculate_dm_sparse 129 9.5 0.001 0.001 6.169 6.275 grid_collocate_task_list 129 9.7 5.615 5.916 5.615 5.916 fft3d_ps 1301 14.7 2.786 2.865 5.745 5.772 dbcsr_complete_redistribute 397 12.7 2.127 2.178 5.196 5.566 multiply_cannon_metrocomm4 7587 15.6 0.026 0.030 1.884 5.518 mp_irecv_dv 29102 15.9 1.842 5.419 1.842 5.419 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.235 5.236 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.187 5.199 mp_alltoall_d11v 2429 14.1 4.597 4.893 4.597 4.893 mp_allgather_i34 2529 14.6 1.156 4.719 1.156 4.719 potential_pw2rs 129 12.3 0.027 0.028 4.201 4.205 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.100 4.163 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.595 3.863 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.618 3.711 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.668 3.706 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.374 3.672 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.427 3.490 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.455 3.469 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=169.893000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1452.181818, yerr=29.362370 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.171639E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3955 59814. MP_Allreduce 10965 1506. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.047 287.748 287.749 qs_mol_dyn_low 1 2.0 0.003 0.003 287.219 287.235 qs_forces 11 3.9 0.004 0.005 287.134 287.138 qs_energies 11 4.9 0.005 0.006 278.345 278.348 scf_env_do_scf 11 5.9 0.001 0.001 255.549 255.558 velocity_verlet 10 3.0 0.002 0.002 207.710 207.717 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 132.747 132.750 init_scf_loop 11 6.9 0.000 0.000 122.517 122.520 prepare_preconditioner 11 7.9 0.000 0.000 117.694 117.729 make_preconditioner 11 8.9 0.000 0.000 117.694 117.729 make_full_inverse_cholesky 11 9.9 0.038 0.039 94.646 114.810 qs_scf_new_mos 116 7.6 0.001 0.001 89.928 90.077 qs_scf_loop_do_ot 116 8.6 0.001 0.001 89.927 90.076 ot_scf_mini 116 9.6 0.004 0.004 85.018 85.123 dbcsr_multiply_generic 2485 12.5 0.209 0.220 82.786 83.616 cp_fm_upper_to_full 104 14.8 52.366 74.335 52.366 74.335 multiply_cannon 2485 13.5 0.672 0.707 59.344 60.486 multiply_cannon_loop 2485 14.5 1.038 1.058 55.081 57.172 ot_mini 116 10.6 0.001 0.001 44.774 44.907 dbcsr_complete_redistribute 393 12.7 4.023 4.081 29.130 41.322 copy_fm_to_dbcsr 208 11.6 0.001 0.002 25.548 37.693 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 23.005 35.000 rebuild_ks_matrix 127 8.3 0.001 0.001 33.638 33.809 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 33.638 33.809 mp_waitall_1 102768 16.8 28.407 33.722 28.407 33.722 cp_fm_cholesky_invert 11 10.9 33.237 33.244 33.237 33.244 mp_alltoall_i22 712 14.1 20.790 33.071 20.790 33.071 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.191 31.359 qs_ot_get_p 127 10.4 0.001 0.001 25.111 25.231 qs_ot_get_derivative 116 11.6 0.002 0.002 24.625 24.735 multiply_cannon_metrocomm3 9940 15.5 0.025 0.027 19.103 21.213 qs_ot_p2m_diag 82 11.4 0.869 0.874 21.011 21.038 make_m2s 4970 13.5 0.075 0.079 18.598 20.313 ot_diis_step 116 11.6 0.022 0.022 20.102 20.102 make_images 4970 14.5 3.060 3.253 18.122 19.836 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.300 19.661 apply_single 127 13.6 0.001 0.001 19.300 19.661 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.241 19.242 multiply_cannon_multrec 9940 15.5 10.248 12.027 18.156 18.227 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.203 16.230 calculate_rho_elec 127 8.7 0.477 0.478 16.202 16.229 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.107 16.108 cp_fm_diag_elpa_base 82 14.4 11.762 13.350 16.104 16.104 sum_up_and_integrate 127 10.3 0.002 0.002 15.802 15.893 integrate_v_rspace 127 11.3 0.004 0.005 15.740 15.831 multiply_cannon_sync_h2d 9940 15.5 14.198 14.235 14.198 14.235 hybrid_alltoall_any 5155 16.4 1.306 3.041 10.860 13.380 make_images_data 4970 15.5 0.064 0.069 10.577 12.976 init_scf_run 11 5.9 0.000 0.001 12.412 12.413 scf_env_initial_rho_setup 11 6.9 0.000 0.000 12.412 12.413 pw_transfer 1535 11.6 0.094 0.094 11.161 11.168 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 10.925 10.928 fft_wrap_pw1pw2_140 519 13.2 3.009 3.055 9.759 9.766 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.632 9.706 dbcsr_mm_accdrv_process 20590 16.0 4.203 6.066 7.660 9.515 wfi_extrapolate 11 7.9 0.001 0.001 9.081 9.081 cp_fm_cholesky_decompose 22 10.9 8.813 8.829 8.813 8.829 grid_integrate_task_list 127 12.3 8.496 8.692 8.496 8.692 qs_energies_init_hamiltonians 11 5.9 0.023 0.030 8.166 8.167 density_rs2pw 127 9.7 0.006 0.006 7.324 7.371 calculate_dm_sparse 127 9.5 0.001 0.001 7.010 7.065 mp_alltoall_d11v 2401 14.1 6.817 6.965 6.817 6.965 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.505 6.565 grid_collocate_task_list 127 9.7 6.366 6.448 6.366 6.448 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.273 6.357 fft3d_ps 1281 14.7 2.830 2.839 6.220 6.277 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=287.749000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2877.909091, yerr=154.913732 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 1.259721E+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 44631284. 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 0.030 0.053 84.834 84.835 qs_energies 1 2.0 0.000 0.000 84.240 84.245 ls_scf 1 3.0 0.000 0.000 83.347 83.352 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.369 72.527 multiply_cannon 111 7.7 0.017 0.020 55.887 56.976 multiply_cannon_loop 111 8.7 0.222 0.243 52.475 53.655 ls_scf_main 1 4.0 0.000 0.000 52.150 52.150 density_matrix_trs4 2 5.0 0.002 0.003 46.657 46.724 ls_scf_init_scf 1 4.0 0.000 0.001 28.146 28.147 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.022 27.073 mp_waitall_1 11031 10.9 22.359 25.542 22.359 25.542 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.945 24.959 multiply_cannon_multrec 2664 9.7 8.153 8.863 15.655 17.340 multiply_cannon_sync_h2d 2664 9.7 13.495 14.716 13.495 14.716 make_m2s 222 7.7 0.008 0.011 12.981 13.458 make_images 222 8.7 0.098 0.109 12.959 13.438 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.661 13.014 make_images_data 222 9.7 0.004 0.005 7.576 8.170 dbcsr_mm_accdrv_process 4760 10.4 0.582 0.687 7.119 8.138 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.457 8.054 hybrid_alltoall_any 227 10.6 0.215 1.834 6.480 7.696 dbcsr_mm_accdrv_process_sort 4760 11.4 6.335 7.256 6.335 7.256 calculate_norms 4752 9.8 5.515 6.187 5.515 6.187 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.995 5.137 mp_sum_l 887 5.1 2.972 4.539 2.972 4.539 make_images_sizes 222 9.7 0.000 0.000 0.782 3.665 mp_alltoall_i44 222 10.7 0.782 3.665 0.782 3.665 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.048 3.621 mp_irecv_dv 6231 10.9 2.030 3.588 2.030 3.588 arnoldi_extremal 4 6.8 0.000 0.000 3.255 3.277 arnoldi_normal_ev 4 7.8 0.001 0.002 3.255 3.277 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.211 3.272 build_subspace 16 8.4 0.009 0.012 3.168 3.170 ls_scf_post 1 4.0 0.000 0.000 3.051 3.056 ls_scf_store_result 1 5.0 0.000 0.000 2.858 2.894 dbcsr_special_finalize 555 9.7 0.005 0.006 2.270 2.733 dbcsr_merge_single_wm 555 10.7 0.462 0.578 2.261 2.724 make_images_pack 222 9.7 2.207 2.631 2.209 2.633 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.340 2.594 dbcsr_matrix_vector_mult_local 304 10.0 2.076 2.490 2.078 2.492 dbcsr_sort_data 658 11.4 2.058 2.455 2.058 2.455 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.288 2.356 buffer_matrices_ensure_size 222 8.7 1.773 2.096 1.773 2.096 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.761 1.762 rebuild_ks_matrix 3 7.3 0.000 0.000 1.751 1.753 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.751 1.753 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.835000, 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/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 2.172158E+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 89.271 89.272 qs_energies 1 2.0 0.000 0.000 88.830 88.837 ls_scf 1 3.0 0.000 0.000 87.507 87.513 dbcsr_multiply_generic 111 6.7 0.015 0.016 73.730 74.021 multiply_cannon 111 7.7 0.028 0.039 52.548 56.131 ls_scf_main 1 4.0 0.000 0.000 54.252 54.253 multiply_cannon_loop 111 8.7 0.135 0.146 49.836 52.820 density_matrix_trs4 2 5.0 0.002 0.003 48.561 48.804 mp_waitall_1 9105 10.9 20.536 30.160 20.536 30.160 ls_scf_init_scf 1 4.0 0.000 0.001 29.592 29.593 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.418 28.499 multiply_cannon_multrec 1332 9.7 13.318 17.302 22.716 27.829 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.045 26.057 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.332 20.862 make_m2s 222 7.7 0.006 0.008 15.057 15.603 make_images 222 8.7 1.374 1.703 15.027 15.573 dbcsr_mm_accdrv_process 4041 10.4 0.307 0.510 8.999 10.575 dbcsr_mm_accdrv_process_sort 4041 11.4 8.519 10.055 8.519 10.055 make_images_data 222 9.7 0.004 0.005 8.702 9.549 hybrid_alltoall_any 227 10.6 0.540 2.553 8.041 9.411 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.207 7.919 mp_irecv_dv 3311 11.0 3.187 7.867 3.187 7.867 mp_sum_l 887 5.1 4.658 7.753 4.658 7.753 calculate_norms 2376 9.8 6.061 6.865 6.061 6.865 multiply_cannon_sync_h2d 1332 9.7 4.883 6.381 4.883 6.381 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.484 5.942 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.961 5.216 arnoldi_extremal 4 6.8 0.000 0.000 4.609 4.621 arnoldi_normal_ev 4 7.8 0.001 0.005 4.608 4.621 build_subspace 16 8.4 0.014 0.021 4.358 4.361 ls_scf_post 1 4.0 0.000 0.000 3.663 3.669 ls_scf_store_result 1 5.0 0.000 0.000 3.370 3.488 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.136 3.372 dbcsr_matrix_vector_mult_local 304 10.0 2.753 3.234 2.755 3.236 mp_allgather_i34 111 8.7 0.703 2.880 0.703 2.880 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.596 2.692 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.148 2.506 dbcsr_data_new 4174 10.1 2.115 2.407 2.115 2.407 make_images_pack 222 9.7 1.832 2.137 1.834 2.139 dbcsr_sort_data 436 11.2 1.841 2.114 1.841 2.114 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.852 1.854 rebuild_ks_matrix 3 7.3 0.000 0.000 1.839 1.841 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.004 1.839 1.841 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.272000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1778.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 2.901664E+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 0.029 0.048 92.915 92.916 qs_energies 1 2.0 0.000 0.000 92.391 92.394 ls_scf 1 3.0 0.000 0.000 90.978 90.981 dbcsr_multiply_generic 111 6.7 0.016 0.017 75.678 75.976 ls_scf_main 1 4.0 0.000 0.000 56.637 56.641 multiply_cannon 111 7.7 0.035 0.093 52.335 56.014 multiply_cannon_loop 111 8.7 0.116 0.130 49.594 53.918 density_matrix_trs4 2 5.0 0.002 0.003 50.743 50.920 mp_waitall_1 7281 11.0 23.808 33.492 23.808 33.492 ls_scf_init_scf 1 4.0 0.000 0.001 30.651 30.654 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.443 29.544 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.097 27.112 multiply_cannon_multrec 888 9.7 12.623 15.311 21.294 24.594 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 11.142 23.157 make_m2s 222 7.7 0.006 0.006 16.391 17.129 make_images 222 8.7 1.591 1.861 16.354 17.090 make_images_data 222 9.7 0.004 0.004 9.462 10.590 hybrid_alltoall_any 227 10.6 0.641 2.948 8.991 10.264 dbcsr_mm_accdrv_process 3754 10.4 0.311 0.490 8.209 9.472 dbcsr_mm_accdrv_process_sort 3754 11.4 7.774 8.982 7.774 8.982 mp_sum_l 887 5.1 5.262 8.868 5.262 8.868 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.849 7.823 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.111 7.151 multiply_cannon_sync_h2d 888 9.7 5.968 7.105 5.968 7.105 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.499 6.991 mp_irecv_dv 2335 11.1 2.483 6.938 2.483 6.938 arnoldi_extremal 4 6.8 0.000 0.000 5.159 5.179 arnoldi_normal_ev 4 7.8 0.001 0.004 5.159 5.179 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.770 4.988 build_subspace 16 8.4 0.014 0.020 4.859 4.866 calculate_norms 1584 9.8 4.380 4.843 4.380 4.843 mp_allgather_i34 111 8.7 0.866 3.805 0.866 3.805 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.434 3.773 ls_scf_post 1 4.0 0.000 0.000 3.690 3.695 dbcsr_matrix_vector_mult_local 304 10.0 3.036 3.620 3.038 3.622 ls_scf_store_result 1 5.0 0.000 0.000 3.417 3.513 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.824 2.914 dbcsr_data_new 4116 9.9 2.093 2.457 2.093 2.457 make_images_sizes 222 9.7 0.000 0.000 1.082 2.370 mp_alltoall_i44 222 10.7 1.082 2.369 1.082 2.369 dbcsr_sort_data 325 11.1 1.886 2.129 1.886 2.129 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.884 1.886 rebuild_ks_matrix 3 7.3 0.000 0.000 1.866 1.868 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.866 1.868 make_images_pack 222 9.7 1.625 1.864 1.628 1.867 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.916000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2227.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 3.412300E+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 0.049 0.093 97.897 97.898 qs_energies 1 2.0 0.000 0.000 97.140 97.157 ls_scf 1 3.0 0.008 0.061 95.262 95.280 dbcsr_multiply_generic 111 6.7 0.017 0.022 78.844 79.079 ls_scf_main 1 4.0 0.001 0.016 59.114 59.115 multiply_cannon 111 7.7 0.046 0.144 51.842 56.365 density_matrix_trs4 2 5.0 0.002 0.005 53.006 53.130 multiply_cannon_loop 111 8.7 0.154 0.168 46.801 50.195 ls_scf_init_scf 1 4.0 0.067 0.156 32.867 32.877 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.591 31.709 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.939 28.952 mp_waitall_1 6369 11.0 22.643 28.355 22.643 28.355 multiply_cannon_multrec 1332 9.7 14.194 17.449 22.224 25.093 make_m2s 222 7.7 0.007 0.008 21.193 22.643 make_images 222 8.7 3.152 3.621 21.143 22.594 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 8.944 17.192 make_images_data 222 9.7 0.004 0.005 11.788 13.461 hybrid_alltoall_any 227 10.6 0.798 3.793 11.187 12.894 dbcsr_mm_accdrv_process 3641 10.4 0.300 0.475 7.656 9.202 dbcsr_mm_accdrv_process_sort 3641 11.4 7.218 8.718 7.218 8.718 mp_sum_l 887 5.1 4.213 8.255 4.213 8.255 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.288 6.635 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.135 6.156 mp_irecv_dv 3229 10.9 2.109 6.071 2.109 6.071 multiply_cannon_sync_h2d 1332 9.7 5.464 5.987 5.464 5.987 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.745 5.886 arnoldi_extremal 4 6.8 0.000 0.000 5.195 5.210 arnoldi_normal_ev 4 7.8 0.001 0.004 5.195 5.210 build_subspace 16 8.4 0.014 0.021 4.866 4.876 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.535 4.706 calculate_norms 2376 9.8 4.185 4.488 4.185 4.488 mp_allgather_i34 111 8.7 2.155 4.438 2.155 4.438 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.587 3.897 dbcsr_matrix_vector_mult_local 304 10.0 3.212 3.716 3.214 3.718 dbcsr_sort_data 658 11.4 3.099 3.383 3.099 3.383 ls_scf_post 1 4.0 0.002 0.016 3.273 3.291 dbcsr_special_finalize 555 9.7 0.006 0.007 2.841 3.266 dbcsr_merge_single_wm 555 10.7 0.542 0.665 2.833 3.258 ls_scf_store_result 1 5.0 0.000 0.000 2.998 3.076 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.011 3.060 dbcsr_data_release 10477 10.7 1.597 2.384 1.597 2.384 dbcsr_finalize 304 7.8 0.049 0.061 1.816 1.976 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.898000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2707.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 4.763279E+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 0.042 0.057 92.628 92.628 qs_energies 1 2.0 0.000 0.000 91.817 91.822 ls_scf 1 3.0 0.000 0.000 89.875 89.880 dbcsr_multiply_generic 111 6.7 0.017 0.019 71.417 71.639 ls_scf_main 1 4.0 0.000 0.000 57.152 57.153 multiply_cannon 111 7.7 0.087 0.162 52.861 55.695 multiply_cannon_loop 111 8.7 0.087 0.093 50.268 52.017 density_matrix_trs4 2 5.0 0.002 0.003 50.207 50.271 ls_scf_init_scf 1 4.0 0.000 0.001 29.309 29.310 mp_waitall_1 5436 11.0 24.772 28.957 24.772 28.957 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.072 28.102 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.967 25.980 multiply_cannon_multrec 444 9.7 13.764 16.267 20.940 22.380 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.921 15.463 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.193 14.951 make_m2s 222 7.7 0.005 0.005 13.706 14.681 make_images 222 8.7 2.041 2.480 13.639 14.611 hybrid_alltoall_any 227 10.6 0.800 3.821 8.243 9.986 make_images_data 222 9.7 0.003 0.004 8.474 9.766 dbcsr_mm_accdrv_process 3003 10.4 0.361 0.400 6.870 8.003 multiply_cannon_sync_h2d 444 9.7 6.741 7.843 6.741 7.843 dbcsr_mm_accdrv_process_sort 3003 11.4 6.509 7.608 6.509 7.608 arnoldi_extremal 4 6.8 0.000 0.000 5.928 5.943 arnoldi_normal_ev 4 7.8 0.001 0.004 5.928 5.942 build_subspace 16 8.4 0.015 0.019 5.534 5.542 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.415 4.590 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.205 4.419 dbcsr_matrix_vector_mult_local 304 10.0 3.730 4.216 3.733 4.218 mp_sum_l 887 5.1 2.847 4.173 2.847 4.173 mp_allgather_i34 111 8.7 1.171 3.790 1.171 3.790 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.525 3.772 calculate_norms 792 9.8 3.616 3.755 3.616 3.755 mp_irecv_dv 1241 11.2 1.512 3.726 1.512 3.726 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.520 3.590 ls_scf_post 1 4.0 0.000 0.000 3.414 3.419 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.040 3.370 make_images_sizes 222 9.7 0.000 0.000 0.843 3.340 mp_alltoall_i44 222 10.7 0.843 3.339 0.843 3.339 ls_scf_store_result 1 5.0 0.000 0.000 3.199 3.238 dbcsr_data_new 4608 9.7 1.803 2.339 1.803 2.339 dbcsr_finalize 304 7.8 0.062 0.077 2.202 2.295 dbcsr_merge_all 275 8.9 0.477 0.525 2.059 2.133 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.006 2.007 rebuild_ks_matrix 3 7.3 0.000 0.000 1.974 1.974 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.004 1.974 1.974 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 1.925 1.926 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.628000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3728.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 8.798548E+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 0.078 0.098 105.413 105.414 qs_energies 1 2.0 0.000 0.000 104.097 104.103 ls_scf 1 3.0 0.000 0.000 101.177 101.182 dbcsr_multiply_generic 111 6.7 0.023 0.028 74.849 74.980 ls_scf_main 1 4.0 0.000 0.000 63.840 63.840 density_matrix_trs4 2 5.0 0.002 0.003 54.923 54.979 multiply_cannon 111 7.7 0.187 0.273 48.277 50.141 multiply_cannon_loop 111 8.7 0.098 0.098 45.307 46.112 ls_scf_init_scf 1 4.0 0.001 0.001 33.544 33.545 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.045 32.071 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.348 29.354 mp_waitall_1 4527 11.1 21.849 25.594 21.849 25.594 make_m2s 222 7.7 0.005 0.005 22.803 23.885 make_images 222 8.7 3.576 3.882 22.696 23.777 multiply_cannon_multrec 444 9.7 17.795 18.376 22.515 23.137 hybrid_alltoall_any 227 10.6 1.657 3.621 12.884 15.952 make_images_data 222 9.7 0.003 0.004 13.183 15.334 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 10.053 10.731 multiply_cannon_sync_h2d 444 9.7 8.792 8.859 8.792 8.859 arnoldi_extremal 4 6.8 0.000 0.000 7.489 7.498 arnoldi_normal_ev 4 7.8 0.002 0.008 7.488 7.498 build_subspace 16 8.4 0.026 0.035 6.940 6.950 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.555 5.713 dbcsr_matrix_vector_mult_local 304 10.0 5.122 5.424 5.124 5.427 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.056 5.148 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.849 5.113 dbcsr_mm_accdrv_process 1814 10.4 0.290 0.352 4.527 4.653 dbcsr_mm_accdrv_process_sort 1814 11.4 4.192 4.325 4.192 4.325 ls_scf_post 1 4.0 0.000 0.000 3.792 3.798 make_images_sizes 222 9.7 0.000 0.000 1.488 3.724 mp_alltoall_i44 222 10.7 1.488 3.723 1.488 3.723 ls_scf_store_result 1 5.0 0.000 0.000 3.511 3.548 mp_allgather_i34 111 8.7 1.089 3.466 1.089 3.466 calculate_norms 792 9.8 3.231 3.267 3.231 3.267 dbcsr_finalize 304 7.8 0.082 0.089 3.077 3.177 dbcsr_merge_all 275 8.9 0.886 0.922 2.862 2.957 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.890 2.890 dbcsr_complete_redistribute 5 7.6 1.441 1.483 2.746 2.856 dbcsr_sort_data 325 11.1 2.441 2.503 2.441 2.503 matrix_ls_to_qs 2 6.0 0.000 0.000 2.379 2.488 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.386 2.388 rebuild_ks_matrix 3 7.3 0.000 0.000 2.320 2.322 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.320 2.322 dbcsr_new_transposed 4 7.5 0.242 0.253 2.274 2.288 dbcsr_data_new 6591 9.6 1.823 2.275 1.823 2.275 dbcsr_frobenius_norm 74 6.6 2.054 2.130 2.215 2.251 dbcsr_add_d 103 6.2 0.000 0.000 2.125 2.202 dbcsr_add_anytype 103 7.2 0.859 0.889 2.125 2.202 dbcsr_data_release 12724 10.6 1.976 2.169 1.976 2.169 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.414000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7052.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/c08aeba1059963d86d9bf16a1dbde02da2d999ed_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 592.801792E+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 824446437. 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 0.028 0.094 241.988 241.989 qs_mol_dyn_low 1 2.0 0.003 0.025 240.986 241.007 qs_forces 5 3.8 0.006 0.050 240.858 240.874 qs_energies 5 4.8 0.002 0.028 237.749 237.781 scf_env_do_scf 5 5.8 0.000 0.001 224.405 224.408 scf_env_do_scf_inner_loop 105 6.6 0.003 0.009 195.939 195.940 qs_scf_new_mos 105 7.6 0.000 0.001 152.879 153.057 qs_scf_loop_do_ot 105 8.6 0.001 0.001 152.879 153.056 ot_scf_mini 105 9.6 0.003 0.004 142.883 143.108 dbcsr_multiply_generic 1445 12.2 0.126 0.134 133.384 133.843 multiply_cannon 1445 13.2 0.274 0.288 113.649 115.754 multiply_cannon_loop 1445 14.2 2.842 2.996 111.629 113.242 velocity_verlet 4 3.0 0.002 0.028 110.510 110.511 ot_mini 105 10.6 0.001 0.015 61.527 61.636 qs_ot_get_p 112 10.4 0.001 0.001 51.167 51.483 multiply_cannon_multrec 69360 15.2 29.626 34.434 39.380 44.567 mp_waitall_1 488190 16.1 36.290 43.390 36.290 43.390 qs_ot_get_derivative 55 11.6 0.001 0.001 39.710 39.823 qs_ot_p2m_diag 40 11.0 0.020 0.030 39.711 39.814 cp_dbcsr_syevd 40 12.0 0.002 0.002 36.358 36.359 multiply_cannon_metrocomm3 69360 15.2 0.204 0.218 26.247 34.238 multiply_cannon_sync_h2d 69360 15.2 29.095 33.654 29.095 33.654 cp_fm_syevd 40 13.0 0.000 0.001 30.876 31.012 rebuild_ks_matrix 110 8.4 0.000 0.000 29.905 30.084 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.015 29.904 30.083 init_scf_loop 7 6.6 0.000 0.006 28.426 28.428 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.448 27.610 cp_fm_redistribute_end 40 14.0 12.929 25.792 12.936 25.796 cp_fm_syevd_base 40 14.0 12.844 25.706 12.844 25.706 prepare_preconditioner 7 7.6 0.000 0.000 23.421 23.452 make_preconditioner 7 8.6 0.000 0.009 23.421 23.452 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.098 23.381 apply_single 62 13.6 0.000 0.000 23.098 23.381 ot_new_cg_direction 55 11.6 0.001 0.003 21.097 21.098 qs_rho_update_rho_low 110 7.6 0.001 0.002 19.230 19.612 calculate_rho_elec 110 8.6 0.030 0.032 19.229 19.612 make_full_inverse_cholesky 7 9.6 0.000 0.000 16.024 16.085 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.954 15.203 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.481 14.580 density_rs2pw 110 9.6 0.005 0.007 13.235 13.728 mp_sum_l 4764 12.2 12.362 13.162 12.362 13.162 pw_transfer 1645 12.4 0.080 0.106 11.852 12.143 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 11.712 12.006 calculate_dm_sparse 110 9.5 0.000 0.001 11.567 11.757 cp_fm_cholesky_invert 7 10.6 11.063 11.076 11.063 11.076 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.587 10.674 qs_vxc_create 110 10.4 0.002 0.005 10.498 10.543 init_scf_run 5 5.8 0.000 0.001 10.542 10.543 scf_env_initial_rho_setup 5 6.8 0.000 0.006 10.542 10.542 fft_wrap_pw1pw2_240 915 15.0 1.173 1.266 10.215 10.420 dbcsr_mm_accdrv_process 154766 15.8 6.162 6.348 9.621 10.411 transfer_rs2pw 445 10.6 0.007 0.008 9.873 10.384 check_diag 80 13.5 8.685 8.937 10.087 10.224 sum_up_and_integrate 60 10.3 0.001 0.003 9.376 9.387 integrate_v_rspace 60 11.3 0.001 0.003 9.359 9.371 fft3d_pb 915 16.0 2.405 2.743 8.361 8.722 multiply_cannon_metrocomm1 69360 15.2 0.096 0.102 4.748 7.885 xc_rho_set_and_dset_create 110 12.4 0.077 0.098 7.314 7.567 make_m2s 2890 13.2 0.078 0.086 6.891 7.466 make_images 2890 14.2 0.239 0.259 6.785 7.359 make_full_single_inverse 7 9.6 0.001 0.002 7.098 7.128 xc_vxc_pw_create 60 11.3 0.039 0.050 7.037 7.083 acc_transpose_blocks 69360 15.2 0.356 0.372 6.222 6.554 xc_pw_derive 510 13.4 0.006 0.007 6.241 6.313 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 5.966 5.981 mp_alltoall_z22v 2340 17.7 5.367 5.708 5.367 5.708 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.001 5.468 5.548 potential_pw2rs 60 12.3 0.003 0.003 5.509 5.547 calculate_first_density_matrix 1 7.0 0.000 0.005 5.483 5.501 mp_waitany 7680 13.5 4.565 5.005 4.565 5.005 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=241.989000, 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: c08aeba1059963d86d9bf16a1dbde02da2d999ed Summary: empty Status: OK