=== 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: 6636911b7db0cba9d55e8e44d46088c49b56cb27 ################# 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 2.9.0, SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 13.09.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.12.0 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/01 job id: 49016007 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/02 job id: 49016008 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/03 job id: 49016009 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/04 job id: 49016011 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/05 job id: 49016014 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/06 job id: 49016017 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/07 job id: 49016018 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/08 job id: 49016021 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/09 job id: 49016023 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/10 job id: 49016025 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/11 job id: 49016026 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/12 job id: 49016027 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/13 job id: 49016028 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/14 job id: 49016029 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/15 job id: 49016030 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/16 job id: 49016031 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/17 job id: 49016032 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/18 job id: 49016034 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/19 job id: 49016035 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/20 job id: 49016036 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/21 job id: 49016037 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/22 job id: 49016038 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/23 job id: 49016039 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/24 job id: 49016040 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/25 job id: 49016041 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/26 job id: 49016042 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/27 job id: 49016043 --- 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 0.039 140.690 140.691 farming_run 1 2.0 139.987 140.002 140.636 140.643 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.496543E+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 0.011 0.029 116.911 116.911 qs_energies 1 2.0 0.000 0.000 116.614 116.620 mp2_main 1 3.0 0.000 0.000 113.739 113.745 mp2_gpw_main 1 4.0 0.037 0.044 112.402 112.408 mp2_ri_gpw_compute_in 1 5.0 0.184 0.210 92.140 92.274 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.551 54.684 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.166 40.935 45.703 get_2c_integrals 1 6.0 0.008 0.009 36.989 37.406 integrate_v_rspace 273 8.0 0.440 0.454 24.554 29.053 pw_transfer 6555 10.6 0.375 0.386 26.709 26.980 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.343 25.592 grid_integrate_task_list 273 9.0 20.460 25.364 20.460 25.364 fft_wrap_pw1pw2_100 2178 12.4 1.077 1.170 22.848 23.068 rpa_ri_compute_en 1 5.0 0.020 0.036 20.130 20.239 compute_2c_integrals 1 7.0 0.035 0.036 19.146 19.147 compute_2c_integrals_loop_lm 1 8.0 0.009 0.010 18.521 18.614 mp2_eri_2c_integrate_gpw 1 9.0 2.370 2.405 18.512 18.606 cp_fm_cholesky_decompose 12 8.2 17.939 18.392 17.939 18.392 cholesky_decomp 1 7.0 0.000 0.000 16.683 17.135 fft3d_s 5443 13.4 16.074 16.320 16.096 16.343 ao_to_mo_and_store_B_mult_1 272 7.0 10.774 15.308 10.774 15.308 calculate_wavefunction 272 8.0 5.419 5.492 12.312 13.057 rpa_num_int 1 6.0 0.002 0.032 10.815 10.832 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.716 10.753 calc_mat_Q 8 8.0 0.000 0.000 9.463 9.558 create_integ_mat 1 6.0 0.015 0.027 9.075 9.085 contract_S_to_Q 8 9.0 0.000 0.000 8.887 8.999 parallel_gemm_fm 14 9.1 0.000 0.000 8.520 8.654 parallel_gemm_fm_cosma 14 10.1 8.520 8.654 8.520 8.654 calc_potential_gpw 544 9.5 0.005 0.005 8.242 8.648 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.032 8.251 potential_pw2rs 545 10.0 0.107 0.108 7.429 8.081 collocate_single_gaussian 272 10.0 0.039 0.041 7.278 7.469 array2fm 1 7.0 0.000 0.000 6.777 7.330 pw_scatter_s 2720 13.7 4.343 4.418 4.343 4.418 pw_gather_s 2722 13.2 3.405 3.810 3.405 3.810 array2fm_buffer_send 1 8.0 2.983 3.146 2.983 3.146 scf_env_do_scf 1 3.0 0.000 0.000 2.536 2.536 init_scf_loop 1 4.0 0.000 0.000 2.353 2.354 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.408175, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2807.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 0.026 0.040 435.745 435.746 farming_run 1 2.0 434.114 434.118 435.664 435.666 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.248092E+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 29024774. 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 0.023 0.044 220.485 220.485 qs_energies 1 2.0 0.000 0.000 220.169 220.176 scf_env_do_scf 1 3.0 0.000 0.000 116.068 116.068 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.441 114.449 rebuild_ks_matrix 4 6.0 0.000 0.000 114.439 114.448 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.067 114.439 114.447 hfx_ks_matrix 4 8.0 0.001 0.001 113.910 113.914 integrate_four_center 4 9.0 0.152 0.470 113.909 113.914 integrate_four_center_main 4 10.0 0.099 0.366 101.925 105.176 integrate_four_center_bin 264 11.0 101.826 105.158 101.826 105.158 mp2_main 1 3.0 0.000 0.000 103.697 103.704 mp2_gpw_main 1 4.0 0.053 0.084 101.793 101.802 init_scf_loop 1 4.0 0.000 0.000 97.528 97.528 mp2_ri_gpw_compute_in 1 5.0 0.068 0.076 74.295 75.339 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.774 54.829 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.159 41.378 46.412 integrate_v_rspace 95 8.0 0.397 0.566 28.018 32.890 pw_transfer 2240 10.6 0.146 0.164 29.573 30.050 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.592 29.101 mp2_ri_gpw_compute_en 1 5.0 0.055 0.063 27.315 28.862 grid_integrate_task_list 95 9.0 23.265 28.344 23.265 28.344 ao_to_mo_and_store_B_mult_1 91 7.0 10.723 27.168 10.723 27.168 fft_wrap_pw1pw2_100 730 12.4 1.278 1.465 26.262 26.726 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.843 1.897 25.608 25.617 get_2c_integrals 1 6.0 0.000 0.000 20.241 20.461 compute_2c_integrals 1 7.0 0.011 0.027 19.211 19.229 compute_2c_integrals_loop_lm 1 8.0 0.004 0.008 18.594 18.930 fft3d_s 1823 13.4 18.474 18.916 18.489 18.929 mp2_eri_2c_integrate_gpw 1 9.0 1.722 1.835 18.590 18.922 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.534 18.535 calculate_wavefunction 91 8.0 2.020 2.052 9.623 9.859 mp2_ri_gpw_compute_en_expansio 172 7.0 0.554 0.595 8.781 9.534 mp2_ri_gpw_compute_en_comm 22 7.0 0.499 0.517 8.596 9.248 potential_pw2rs 186 10.0 0.033 0.035 8.591 9.197 local_gemm 172 8.0 8.226 8.959 8.226 8.959 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.145 8.596 calc_potential_gpw 182 9.5 0.002 0.002 7.894 8.204 collocate_single_gaussian 91 10.0 0.016 0.022 7.733 7.971 mp_sendrecv_dm3 2068 8.0 6.628 7.292 6.628 7.292 mp_sync 37 10.5 4.003 7.093 4.003 7.093 integrate_four_center_load 4 10.0 0.000 0.000 6.793 6.797 hfx_load_balance 1 11.0 0.000 0.000 6.793 6.797 mp2_ri_gpw_compute_en_ener 172 7.0 6.343 6.435 6.343 6.435 pw_gather_s 912 13.2 4.485 4.988 4.485 4.988 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.787969, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1507.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.124672E+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 275362. 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 0.029 0.066 64.891 64.893 qs_mol_dyn_low 1 2.0 0.003 0.006 64.482 64.526 qs_forces 11 3.9 0.004 0.017 64.375 64.385 qs_energies 11 4.9 0.004 0.011 62.448 62.468 scf_env_do_scf 11 5.9 0.001 0.002 55.312 55.313 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 51.603 51.604 qs_scf_new_mos 108 7.5 0.000 0.001 36.051 36.379 qs_scf_loop_do_ot 108 8.5 0.001 0.001 36.051 36.378 dbcsr_multiply_generic 2286 12.5 0.095 0.105 34.827 35.370 ot_scf_mini 108 9.5 0.002 0.003 34.285 34.448 velocity_verlet 10 3.0 0.001 0.003 32.105 32.108 multiply_cannon 2286 13.5 0.195 0.201 26.196 27.923 multiply_cannon_loop 2286 14.5 1.818 1.966 24.596 26.285 ot_mini 108 10.5 0.001 0.002 19.895 20.154 qs_ot_get_derivative 108 11.5 0.001 0.002 16.876 17.048 mp_waitall_1 245248 16.5 9.585 15.591 9.585 15.591 multiply_cannon_metrocomm3 54864 15.5 0.075 0.083 5.271 12.659 rebuild_ks_matrix 119 8.3 0.000 0.000 11.288 11.453 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.015 11.288 11.453 multiply_cannon_multrec 54864 15.5 3.712 5.965 7.625 11.089 qs_ks_update_qs_env 119 7.6 0.001 0.001 9.950 10.091 qs_ot_get_p 119 10.4 0.001 0.001 9.464 9.716 qs_rho_update_rho_low 119 7.7 0.001 0.002 7.226 7.344 calculate_rho_elec 119 8.7 0.011 0.017 7.225 7.343 sum_up_and_integrate 119 10.3 0.001 0.003 7.186 7.195 integrate_v_rspace 119 11.3 0.002 0.002 7.174 7.185 mp_sum_l 7287 12.8 4.099 6.328 4.099 6.328 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.669 6.105 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.892 5.941 multiply_cannon_sync_h2d 54864 15.5 5.073 5.779 5.073 5.779 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.567 5.665 dbcsr_mm_accdrv_process 76910 16.1 1.775 2.923 3.827 5.627 init_scf_run 11 5.9 0.000 0.001 5.508 5.509 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.508 5.508 density_rs2pw 119 9.7 0.004 0.004 5.000 5.109 cp_dbcsr_syevd 50 12.0 0.003 0.004 5.056 5.058 pw_transfer 1439 11.6 0.052 0.057 4.466 4.675 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 4.391 4.602 make_m2s 4572 13.5 0.054 0.058 4.231 4.488 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.454 4.459 make_images 4572 14.5 0.136 0.146 4.148 4.404 cp_fm_redistribute_end 50 14.0 2.271 4.388 2.282 4.397 potential_pw2rs 119 12.3 0.004 0.004 4.193 4.347 fft3d_ps 1201 14.6 0.372 0.487 4.113 4.320 cp_fm_diag_elpa_base 50 14.0 2.097 4.217 2.103 4.230 init_scf_loop 11 6.9 0.001 0.005 3.650 3.652 fft_wrap_pw1pw2_140 487 13.2 0.133 0.149 3.429 3.643 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.087 3.295 apply_single 119 13.6 0.000 0.000 3.087 3.295 calculate_dm_sparse 119 9.5 0.000 0.001 3.019 3.191 transfer_rs2pw 487 10.6 0.005 0.006 2.848 3.032 mp_alltoall_z22v 1201 16.6 2.828 3.027 2.828 3.027 mp_alltoall_d11v 2130 13.8 2.709 3.015 2.709 3.015 transfer_pw2rs 487 13.2 0.006 0.006 2.801 2.949 make_images_sizes 4572 15.5 0.004 0.005 2.372 2.894 mp_alltoall_i44 4572 16.5 2.367 2.890 2.367 2.890 ot_diis_step 108 11.5 0.006 0.010 2.812 2.813 acc_transpose_blocks 54864 15.5 0.224 0.259 2.177 2.779 multiply_cannon_metrocomm1 54864 15.5 0.057 0.062 1.681 2.778 wfi_extrapolate 11 7.9 0.001 0.002 2.748 2.748 calculate_first_density_matrix 1 7.0 0.000 0.002 2.635 2.639 jit_kernel_multiply 13 15.8 1.991 2.617 1.991 2.617 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.573 2.577 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.475 2.536 prepare_preconditioner 11 7.9 0.000 0.000 2.160 2.195 make_preconditioner 11 8.9 0.000 0.001 2.160 2.194 grid_integrate_task_list 119 12.3 1.972 2.103 1.972 2.103 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.045 2.097 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.033 2.078 yz_to_x 249 15.8 0.026 0.033 1.766 1.955 make_images_data 4572 15.5 0.042 0.047 1.157 1.879 mp_sum_d 4135 12.0 1.250 1.831 1.250 1.831 mp_waitany 12084 13.8 1.665 1.823 1.665 1.823 mp_allgather_i34 2286 14.5 1.125 1.748 1.125 1.748 hybrid_alltoall_any 4725 16.4 0.044 0.117 0.948 1.619 mp_sendrecv_dv 22610 12.7 1.384 1.519 1.384 1.519 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.490 1.512 acc_transpose_blocks_sync 164592 16.5 1.179 1.478 1.179 1.478 grid_collocate_task_list 119 9.7 1.357 1.425 1.357 1.425 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.393 1.418 dbcsr_complete_redistribute 329 12.2 0.053 0.085 1.307 1.366 arnoldi_extremal 119 11.4 0.001 0.004 1.145 1.311 arnoldi_normal_ev 119 12.4 0.002 0.006 1.143 1.307 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=64.893000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 487.497728E+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 540540. 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 0.019 0.067 57.941 57.942 qs_mol_dyn_low 1 2.0 0.004 0.013 57.475 57.485 qs_forces 11 3.9 0.004 0.007 57.374 57.383 qs_energies 11 4.9 0.025 0.045 55.280 55.302 scf_env_do_scf 11 5.9 0.001 0.002 48.428 48.428 scf_env_do_scf_inner_loop 108 6.5 0.003 0.009 43.798 43.799 dbcsr_multiply_generic 2286 12.5 0.105 0.113 28.702 30.381 qs_scf_new_mos 108 7.5 0.001 0.001 29.705 29.977 qs_scf_loop_do_ot 108 8.5 0.001 0.001 29.704 29.977 ot_scf_mini 108 9.5 0.003 0.004 28.467 28.658 velocity_verlet 10 3.0 0.001 0.001 28.580 28.581 multiply_cannon 2286 13.5 0.213 0.226 19.408 24.200 multiply_cannon_loop 2286 14.5 1.210 1.279 17.412 22.513 ot_mini 108 10.5 0.001 0.001 17.196 17.497 mp_waitall_1 200699 16.5 10.023 16.040 10.023 16.040 qs_ot_get_derivative 108 11.5 0.001 0.001 14.110 14.301 multiply_cannon_metrocomm3 27432 15.5 0.072 0.076 5.384 12.149 rebuild_ks_matrix 119 8.3 0.000 0.000 9.965 10.460 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.016 9.964 10.460 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.841 9.279 multiply_cannon_multrec 27432 15.5 1.762 4.146 6.257 9.018 qs_ot_get_p 119 10.4 0.001 0.002 7.576 7.856 qs_rho_update_rho_low 119 7.7 0.001 0.001 6.758 6.856 calculate_rho_elec 119 8.7 0.021 0.024 6.758 6.855 mp_sum_l 7287 12.8 4.271 6.638 4.271 6.638 dbcsr_mm_accdrv_process 47894 16.0 3.659 6.043 4.417 6.563 sum_up_and_integrate 119 10.3 0.001 0.001 6.493 6.503 integrate_v_rspace 119 11.3 0.002 0.003 6.479 6.488 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.136 5.990 make_m2s 4572 13.5 0.052 0.055 4.591 5.238 init_scf_run 11 5.9 0.000 0.001 5.237 5.238 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.237 5.237 make_images 4572 14.5 0.208 0.246 4.504 5.147 density_rs2pw 119 9.7 0.004 0.004 4.903 4.986 qs_ot_p2m_diag 50 11.0 0.009 0.013 4.586 4.643 init_scf_loop 11 6.9 0.001 0.005 4.581 4.584 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.263 4.279 apply_single 119 13.6 0.000 0.000 3.263 4.279 potential_pw2rs 119 12.3 0.006 0.007 4.101 4.164 cp_dbcsr_syevd 50 12.0 0.004 0.004 4.037 4.039 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 3.833 3.930 transfer_rs2pw 487 10.6 0.005 0.005 3.512 3.680 cp_fm_diag_elpa 50 13.0 0.000 0.000 3.341 3.344 cp_fm_redistribute_end 50 14.0 1.687 3.277 1.702 3.280 prepare_preconditioner 11 7.9 0.000 0.001 3.226 3.254 make_preconditioner 11 8.9 0.000 0.001 3.225 3.254 pw_transfer 1439 11.6 0.065 0.069 3.091 3.200 make_images_data 4572 15.5 0.048 0.056 1.977 3.180 cp_fm_diag_elpa_base 50 14.0 1.527 3.092 1.562 3.151 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.066 3.148 transfer_pw2rs 487 13.2 0.005 0.005 3.090 3.146 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.999 3.110 ot_diis_step 108 11.5 0.011 0.011 2.878 2.879 calculate_first_density_matrix 1 7.0 0.000 0.002 2.751 2.753 fft3d_ps 1201 14.6 0.528 0.586 2.637 2.749 hybrid_alltoall_any 4725 16.4 0.055 0.119 1.680 2.738 calculate_dm_sparse 119 9.5 0.000 0.001 2.513 2.637 mp_sendrecv_dv 16779 12.7 2.472 2.530 2.472 2.530 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.468 2.472 wfi_extrapolate 11 7.9 0.001 0.001 2.388 2.389 make_images_sizes 4572 15.5 0.005 0.005 1.750 2.355 mp_alltoall_i44 4572 16.5 1.746 2.350 1.746 2.350 mp_alltoall_d11v 2130 13.8 2.046 2.307 2.046 2.307 mp_allgather_i34 2286 14.5 1.408 2.228 1.408 2.228 multiply_cannon_sync_h2d 27432 15.5 1.548 2.187 1.548 2.187 jit_kernel_multiply 10 16.5 0.699 2.175 0.699 2.175 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.966 2.111 acc_transpose_blocks 27432 15.5 0.115 0.121 1.711 2.070 fft_wrap_pw1pw2_140 487 13.2 0.158 0.168 1.883 1.997 grid_integrate_task_list 119 12.3 1.818 1.941 1.818 1.941 mp_sum_d 4135 12.0 1.246 1.893 1.246 1.893 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.745 1.821 transfer_rs2pw_50 119 11.7 0.182 0.193 1.544 1.737 mp_alltoall_z22v 1201 16.6 1.563 1.721 1.563 1.721 cp_fm_cholesky_invert 11 10.9 1.700 1.711 1.700 1.711 transfer_pw2rs_50 119 14.3 0.176 0.190 1.470 1.680 multiply_cannon_metrocomm1 27432 15.5 0.032 0.037 0.556 1.650 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.452 1.475 dbcsr_dot_sd 1205 11.9 0.071 0.088 0.883 1.375 dbcsr_complete_redistribute 329 12.2 0.089 0.127 1.261 1.368 grid_collocate_task_list 119 9.7 1.289 1.360 1.289 1.360 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.189 1.247 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=57.942000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.454545, yerr=1.437399 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 518.471680E+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 0.029 0.084 49.511 49.512 qs_mol_dyn_low 1 2.0 0.003 0.003 49.142 49.164 qs_forces 11 3.9 0.004 0.019 49.066 49.074 qs_energies 11 4.9 0.006 0.020 47.103 47.126 scf_env_do_scf 11 5.9 0.001 0.001 41.096 41.097 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 36.410 36.414 velocity_verlet 10 3.0 0.001 0.002 24.613 24.614 qs_scf_new_mos 108 7.5 0.001 0.001 23.727 23.755 qs_scf_loop_do_ot 108 8.5 0.001 0.001 23.727 23.754 dbcsr_multiply_generic 2286 12.5 0.102 0.107 22.755 23.718 ot_scf_mini 108 9.5 0.002 0.003 22.573 22.604 multiply_cannon 2286 13.5 0.197 0.206 15.720 17.888 multiply_cannon_loop 2286 14.5 0.870 0.909 14.181 16.779 ot_mini 108 10.5 0.001 0.001 13.174 13.247 qs_ot_get_derivative 108 11.5 0.001 0.001 10.948 10.971 mp_waitall_1 158411 16.6 6.855 10.409 6.855 10.409 rebuild_ks_matrix 119 8.3 0.000 0.000 8.996 9.267 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 8.995 9.266 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.995 8.237 multiply_cannon_multrec 18288 15.5 1.832 2.677 6.961 7.090 qs_ot_get_p 119 10.4 0.001 0.001 6.254 6.364 sum_up_and_integrate 119 10.3 0.001 0.001 6.185 6.190 integrate_v_rspace 119 11.3 0.003 0.003 6.171 6.178 qs_rho_update_rho_low 119 7.7 0.001 0.001 6.053 6.113 calculate_rho_elec 119 8.7 0.031 0.031 6.053 6.112 dbcsr_mm_accdrv_process 38222 16.0 4.952 5.792 5.040 5.857 multiply_cannon_metrocomm3 18288 15.5 0.048 0.049 2.441 5.813 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.044 4.675 init_scf_loop 11 6.9 0.001 0.003 4.640 4.644 init_scf_run 11 5.9 0.000 0.001 4.497 4.498 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.497 4.497 density_rs2pw 119 9.7 0.004 0.004 4.372 4.459 make_m2s 4572 13.5 0.044 0.046 3.696 4.247 make_images 4572 14.5 0.195 0.206 3.611 4.162 qs_ot_p2m_diag 50 11.0 0.012 0.013 3.847 3.868 potential_pw2rs 119 12.3 0.008 0.009 3.792 3.857 mp_sum_l 7287 12.8 2.954 3.776 2.954 3.776 prepare_preconditioner 11 7.9 0.000 0.000 3.489 3.497 make_preconditioner 11 8.9 0.000 0.001 3.489 3.496 cp_dbcsr_syevd 50 12.0 0.003 0.004 3.417 3.420 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.251 3.383 transfer_rs2pw 487 10.6 0.005 0.005 3.021 3.106 pw_transfer 1439 11.6 0.065 0.069 2.934 3.000 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.841 2.905 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.882 2.897 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.868 2.878 transfer_pw2rs 487 13.2 0.004 0.004 2.823 2.876 cp_fm_diag_elpa_base 50 14.0 2.827 2.853 2.864 2.874 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.304 2.559 apply_single 119 13.6 0.000 0.000 2.304 2.559 make_images_data 4572 15.5 0.048 0.054 1.731 2.545 fft3d_ps 1201 14.6 0.536 0.556 2.412 2.486 calculate_first_density_matrix 1 7.0 0.001 0.002 2.338 2.340 calculate_dm_sparse 119 9.5 0.000 0.000 2.277 2.304 mp_sendrecv_dv 11067 12.7 2.116 2.168 2.116 2.168 wfi_extrapolate 11 7.9 0.001 0.001 2.061 2.061 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.052 2.055 ot_diis_step 108 11.5 0.012 0.012 2.025 2.026 hybrid_alltoall_any 4725 16.4 0.059 0.119 1.420 2.023 grid_integrate_task_list 119 12.3 1.769 1.939 1.769 1.939 fft_wrap_pw1pw2_140 487 13.2 0.209 0.216 1.810 1.881 mp_alltoall_d11v 2130 13.8 1.591 1.821 1.591 1.821 make_images_sizes 4572 15.5 0.005 0.005 1.206 1.766 mp_alltoall_i44 4572 16.5 1.202 1.761 1.202 1.761 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.610 1.741 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 1.655 1.729 acc_transpose_blocks 18288 15.5 0.081 0.083 1.644 1.712 cp_fm_cholesky_invert 11 10.9 1.687 1.702 1.687 1.702 mp_alltoall_z22v 1201 16.6 1.603 1.691 1.603 1.691 mp_allgather_i34 2286 14.5 0.983 1.601 0.983 1.601 multiply_cannon_metrocomm1 18288 15.5 0.029 0.031 0.776 1.580 grid_collocate_task_list 119 9.7 1.238 1.324 1.238 1.324 transfer_pw2rs_50 119 14.3 0.142 0.149 1.222 1.321 transfer_rs2pw_50 119 11.7 0.154 0.159 1.186 1.303 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.266 1.296 mp_sum_d 4135 12.0 0.967 1.247 0.967 1.247 dbcsr_complete_redistribute 329 12.2 0.093 0.105 0.992 1.147 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.080 1.109 multiply_cannon_sync_h2d 18288 15.5 0.932 1.091 0.932 1.091 cp_fm_cholesky_decompose 22 10.9 1.071 1.074 1.071 1.074 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.053 1.060 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=49.512000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=494.181818, yerr=0.833196 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 555.597824E+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 0.039 0.043 54.237 54.237 qs_mol_dyn_low 1 2.0 0.003 0.003 53.962 53.972 qs_forces 11 3.9 0.002 0.003 53.870 53.875 qs_energies 11 4.9 0.002 0.002 51.812 51.823 scf_env_do_scf 11 5.9 0.000 0.001 45.544 45.546 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 39.218 39.219 velocity_verlet 10 3.0 0.002 0.002 28.123 28.126 qs_scf_new_mos 108 7.5 0.001 0.001 26.488 26.559 qs_scf_loop_do_ot 108 8.5 0.001 0.001 26.487 26.559 dbcsr_multiply_generic 2286 12.5 0.102 0.106 25.551 25.914 ot_scf_mini 108 9.5 0.002 0.003 25.119 25.177 multiply_cannon 2286 13.5 0.221 0.228 18.178 19.171 multiply_cannon_loop 2286 14.5 1.531 1.629 16.238 17.581 ot_mini 108 10.5 0.001 0.001 14.425 14.516 qs_ot_get_derivative 108 11.5 0.001 0.001 12.011 12.069 multiply_cannon_multrec 27432 15.5 2.504 3.222 9.102 9.370 rebuild_ks_matrix 119 8.3 0.000 0.000 9.201 9.368 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 9.200 9.368 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.215 8.370 mp_waitall_1 137007 16.6 5.815 8.041 5.815 8.041 dbcsr_mm_accdrv_process 47916 15.9 5.867 7.370 6.498 7.845 qs_ot_get_p 119 10.4 0.001 0.002 7.084 7.183 init_scf_loop 11 6.9 0.001 0.001 6.268 6.270 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.899 5.955 calculate_rho_elec 119 8.7 0.040 0.046 5.899 5.955 sum_up_and_integrate 119 10.3 0.001 0.001 5.918 5.928 integrate_v_rspace 119 11.3 0.003 0.003 5.906 5.916 make_m2s 4572 13.5 0.054 0.056 4.768 5.447 make_images 4572 14.5 0.275 0.338 4.659 5.334 prepare_preconditioner 11 7.9 0.000 0.000 5.082 5.092 make_preconditioner 11 8.9 0.000 0.000 5.082 5.092 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.564 4.953 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.371 4.705 init_scf_run 11 5.9 0.000 0.001 4.649 4.649 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.648 4.649 qs_ot_p2m_diag 50 11.0 0.015 0.023 4.513 4.529 density_rs2pw 119 9.7 0.004 0.004 4.180 4.261 cp_dbcsr_syevd 50 12.0 0.004 0.005 3.938 3.939 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 3.499 3.525 potential_pw2rs 119 12.3 0.009 0.010 3.499 3.515 multiply_cannon_metrocomm3 27432 15.5 0.041 0.042 1.365 3.316 cp_fm_diag_elpa 50 13.0 0.000 0.000 3.278 3.286 cp_fm_diag_elpa_base 50 14.0 3.213 3.239 3.274 3.283 pw_transfer 1439 11.6 0.065 0.069 3.069 3.109 make_images_data 4572 15.5 0.048 0.051 2.205 3.076 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.976 3.019 transfer_rs2pw 487 10.6 0.004 0.005 2.752 2.839 mp_sum_l 7287 12.8 2.188 2.805 2.188 2.805 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.425 2.686 apply_single 119 13.6 0.000 0.000 2.425 2.686 hybrid_alltoall_any 4725 16.4 0.066 0.156 1.831 2.637 calculate_dm_sparse 119 9.5 0.000 0.000 2.565 2.626 acc_transpose_blocks 27432 15.5 0.120 0.124 2.397 2.580 fft3d_ps 1201 14.6 0.561 0.611 2.496 2.530 transfer_pw2rs 487 13.2 0.004 0.004 2.492 2.508 calculate_first_density_matrix 1 7.0 0.000 0.000 2.350 2.352 ot_diis_step 108 11.5 0.012 0.012 2.283 2.284 wfi_extrapolate 11 7.9 0.001 0.001 2.208 2.208 cp_fm_cholesky_invert 11 10.9 2.130 2.137 2.130 2.137 fft_wrap_pw1pw2_140 487 13.2 0.245 0.259 2.079 2.123 make_images_sizes 4572 15.5 0.005 0.005 1.435 2.091 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.083 2.089 mp_alltoall_i44 4572 16.5 1.431 2.086 1.431 2.086 mp_sendrecv_dv 8211 12.7 1.912 1.933 1.912 1.933 mp_alltoall_d11v 2130 13.8 1.749 1.900 1.749 1.900 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.804 1.880 grid_integrate_task_list 119 12.3 1.792 1.879 1.792 1.879 mp_alltoall_z22v 1201 16.6 1.815 1.843 1.815 1.843 dbcsr_complete_redistribute 329 12.2 0.118 0.144 1.447 1.769 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.724 1.758 mp_allgather_i34 2286 14.5 1.213 1.614 1.213 1.614 acc_transpose_blocks_sync 82296 16.5 1.414 1.587 1.414 1.587 cp_fm_upper_to_full 72 14.2 0.970 1.408 0.970 1.408 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.393 1.408 grid_collocate_task_list 119 9.7 1.251 1.345 1.251 1.345 cp_fm_cholesky_decompose 22 10.9 1.291 1.295 1.291 1.295 jit_kernel_multiply 7 16.1 0.565 1.275 0.565 1.275 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.906 1.232 multiply_cannon_metrocomm1 27432 15.5 0.035 0.036 0.592 1.195 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.162 1.167 transfer_rs2pw_50 119 11.7 0.139 0.146 1.113 1.165 transfer_pw2rs_50 119 14.3 0.128 0.134 1.039 1.104 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.084 1.097 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=54.237000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.090909, yerr=3.088234 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 619.012096E+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 0.041 0.059 45.721 45.731 qs_mol_dyn_low 1 2.0 0.003 0.003 45.444 45.468 qs_forces 11 3.9 0.002 0.002 45.366 45.369 qs_energies 11 4.9 0.002 0.004 43.249 43.258 scf_env_do_scf 11 5.9 0.000 0.001 37.447 37.447 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 32.184 32.184 velocity_verlet 10 3.0 0.012 0.013 24.377 24.381 qs_scf_new_mos 108 7.5 0.001 0.001 20.052 20.130 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.051 20.129 ot_scf_mini 108 9.5 0.002 0.002 19.010 19.057 dbcsr_multiply_generic 2286 12.5 0.096 0.098 18.708 19.046 multiply_cannon 2286 13.5 0.226 0.231 13.207 15.004 multiply_cannon_loop 2286 14.5 0.649 0.664 11.580 12.489 ot_mini 108 10.5 0.001 0.001 10.279 10.374 rebuild_ks_matrix 119 8.3 0.000 0.000 8.550 8.654 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 8.550 8.654 qs_ot_get_derivative 108 11.5 0.001 0.001 8.510 8.556 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.610 7.703 mp_waitall_1 115863 16.7 5.625 7.152 5.625 7.152 multiply_cannon_multrec 9144 15.5 1.807 1.977 6.141 6.337 qs_ot_get_p 119 10.4 0.001 0.001 6.010 6.106 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.730 5.741 calculate_rho_elec 119 8.7 0.060 0.061 5.729 5.740 sum_up_and_integrate 119 10.3 0.001 0.001 5.694 5.703 integrate_v_rspace 119 11.3 0.003 0.003 5.684 5.693 init_scf_loop 11 6.9 0.001 0.002 5.213 5.215 dbcsr_mm_accdrv_process 12550 15.8 3.622 4.255 4.223 4.299 prepare_preconditioner 11 7.9 0.000 0.000 4.200 4.217 make_preconditioner 11 8.9 0.000 0.000 4.200 4.217 make_m2s 4572 13.5 0.034 0.035 3.650 4.171 init_scf_run 11 5.9 0.000 0.001 4.111 4.112 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.111 4.111 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.011 4.083 make_images 4572 14.5 0.273 0.302 3.561 4.083 qs_ot_p2m_diag 50 11.0 0.022 0.023 3.948 3.952 density_rs2pw 119 9.7 0.003 0.004 3.811 3.858 cp_dbcsr_syevd 50 12.0 0.004 0.004 3.560 3.562 potential_pw2rs 119 12.3 0.010 0.011 3.119 3.144 pw_transfer 1439 11.6 0.066 0.068 3.035 3.075 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.985 3.011 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.979 2.984 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.941 2.982 cp_fm_diag_elpa_base 50 14.0 2.937 2.960 2.975 2.981 cp_fm_cholesky_invert 11 10.9 2.555 2.562 2.555 2.562 make_images_data 4572 15.5 0.042 0.045 1.760 2.521 transfer_rs2pw 487 10.6 0.004 0.004 2.437 2.505 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.390 2.411 fft3d_ps 1201 14.6 0.567 0.579 2.355 2.391 hybrid_alltoall_any 4725 16.4 0.066 0.178 1.581 2.306 calculate_dm_sparse 119 9.5 0.000 0.000 2.134 2.194 transfer_pw2rs 487 13.2 0.003 0.004 2.128 2.159 grid_integrate_task_list 119 12.3 2.043 2.151 2.043 2.151 fft_wrap_pw1pw2_140 487 13.2 0.322 0.332 2.097 2.139 mp_sum_l 7287 12.8 1.411 2.125 1.411 2.125 calculate_first_density_matrix 1 7.0 0.000 0.000 2.121 2.124 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 0.957 2.092 multiply_cannon_metrocomm1 9144 15.5 0.024 0.024 1.548 1.929 wfi_extrapolate 11 7.9 0.001 0.001 1.913 1.913 mp_allgather_i34 2286 14.5 0.933 1.808 0.933 1.808 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.775 1.778 make_images_sizes 4572 15.5 0.005 0.005 1.107 1.722 mp_alltoall_i44 4572 16.5 1.101 1.717 1.101 1.717 mp_alltoall_z22v 1201 16.6 1.657 1.707 1.657 1.707 mp_alltoall_d11v 2130 13.8 1.559 1.692 1.559 1.692 ot_diis_step 108 11.5 0.013 0.013 1.683 1.683 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.535 1.633 apply_single 119 13.6 0.000 0.000 1.535 1.632 mp_sendrecv_dv 5355 12.7 1.589 1.616 1.589 1.616 grid_collocate_task_list 119 9.7 1.427 1.491 1.427 1.491 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.444 1.479 acc_transpose_blocks 9144 15.5 0.042 0.042 1.445 1.464 jit_kernel_multiply 8 15.6 0.561 1.404 0.561 1.404 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.351 1.389 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.292 1.295 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.245 1.261 cp_fm_cholesky_decompose 22 10.9 1.113 1.119 1.113 1.119 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.074 1.083 dbcsr_complete_redistribute 329 12.2 0.159 0.165 1.000 1.060 transfer_rs2pw_50 119 11.7 0.126 0.129 0.917 0.998 arnoldi_extremal 119 11.4 0.002 0.002 0.948 0.992 arnoldi_normal_ev 119 12.4 0.002 0.002 0.946 0.991 transfer_pw2rs_50 119 14.3 0.111 0.120 0.851 0.942 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.877 0.927 yz_to_x 606 15.1 0.053 0.056 0.877 0.915 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=45.731000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=587.000000, yerr=6.660603 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 759.414784E+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 0.018 0.065 57.277 57.282 qs_mol_dyn_low 1 2.0 0.003 0.003 56.919 56.934 qs_forces 11 3.9 0.002 0.002 56.841 56.842 qs_energies 11 4.9 0.002 0.003 54.527 54.531 scf_env_do_scf 11 5.9 0.001 0.001 47.565 47.566 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 37.137 37.140 velocity_verlet 10 3.0 0.002 0.002 32.254 32.260 dbcsr_multiply_generic 2286 12.5 0.103 0.106 24.555 24.800 qs_scf_new_mos 108 7.5 0.001 0.001 24.051 24.093 qs_scf_loop_do_ot 108 8.5 0.001 0.001 24.050 24.093 ot_scf_mini 108 9.5 0.002 0.002 22.605 22.667 multiply_cannon 2286 13.5 0.305 0.317 17.536 18.295 multiply_cannon_loop 2286 14.5 0.879 0.910 15.533 16.055 ot_mini 108 10.5 0.001 0.001 13.333 13.364 qs_ot_get_derivative 108 11.5 0.001 0.001 10.896 10.958 init_scf_loop 11 6.9 0.008 0.015 10.381 10.386 rebuild_ks_matrix 119 8.3 0.000 0.000 9.554 9.600 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 9.553 9.600 prepare_preconditioner 11 7.9 0.000 0.000 9.169 9.175 make_preconditioner 11 8.9 0.000 0.000 9.169 9.175 multiply_cannon_multrec 9144 15.5 3.424 4.878 8.998 9.097 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.601 8.999 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.590 8.627 mp_waitall_1 94719 16.7 6.160 7.678 6.160 7.678 dbcsr_mm_accdrv_process 12550 15.8 4.783 6.618 5.436 6.716 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.883 5.908 calculate_rho_elec 119 8.7 0.118 0.121 5.883 5.908 qs_ot_get_p 119 10.4 0.001 0.002 5.497 5.579 make_m2s 4572 13.5 0.037 0.038 4.797 5.535 sum_up_and_integrate 119 10.3 0.001 0.001 5.481 5.486 integrate_v_rspace 119 11.3 0.003 0.003 5.470 5.476 make_images 4572 14.5 0.357 0.388 4.679 5.417 cp_fm_upper_to_full 72 14.2 3.239 4.713 3.239 4.713 init_scf_run 11 5.9 0.000 0.001 4.706 4.706 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.705 4.705 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.970 4.387 pw_transfer 1439 11.6 0.068 0.069 3.964 3.997 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.866 3.900 make_images_data 4572 15.5 0.046 0.049 2.604 3.812 density_rs2pw 119 9.7 0.003 0.004 3.634 3.654 hybrid_alltoall_any 4725 16.4 0.091 0.151 2.342 3.554 qs_ot_p2m_diag 50 11.0 0.043 0.044 3.385 3.395 dbcsr_complete_redistribute 329 12.2 0.297 0.302 2.450 3.357 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 2.725 3.348 fft_wrap_pw1pw2_140 487 13.2 0.576 0.579 3.096 3.130 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.961 3.003 fft3d_ps 1201 14.6 0.603 0.616 2.962 2.993 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.587 2.954 apply_single 119 13.6 0.000 0.000 2.587 2.954 cp_fm_cholesky_invert 11 10.9 2.904 2.914 2.904 2.914 cp_dbcsr_syevd 50 12.0 0.004 0.004 2.888 2.890 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.958 2.847 calculate_dm_sparse 119 9.5 0.000 0.001 2.706 2.745 mp_alltoall_i22 627 13.8 1.674 2.668 1.674 2.668 potential_pw2rs 119 12.3 0.014 0.014 2.647 2.665 calculate_first_density_matrix 1 7.0 0.000 0.000 2.548 2.553 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.563 2.427 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.398 2.398 cp_fm_diag_elpa_base 50 14.0 2.228 2.302 2.395 2.396 ot_diis_step 108 11.5 0.014 0.015 2.370 2.371 mp_alltoall_z22v 1201 16.6 2.217 2.250 2.217 2.250 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.244 2.247 grid_integrate_task_list 119 12.3 2.173 2.191 2.173 2.191 wfi_extrapolate 11 7.9 0.001 0.001 2.067 2.067 mp_sum_l 7287 12.8 1.703 2.065 1.703 2.065 mp_allgather_i34 2286 14.5 0.947 2.061 0.947 2.061 mp_alltoall_d11v 2130 13.8 1.972 2.009 1.972 2.009 acc_transpose_blocks 9144 15.5 0.044 0.045 1.893 1.942 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.821 1.881 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.849 1.850 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.777 1.786 make_images_sizes 4572 15.5 0.005 0.005 1.083 1.770 mp_alltoall_i44 4572 16.5 1.078 1.765 1.078 1.765 transfer_rs2pw 487 10.6 0.005 0.005 1.678 1.705 grid_collocate_task_list 119 9.7 1.587 1.600 1.587 1.600 transfer_pw2rs 487 13.2 0.003 0.004 1.533 1.554 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.464 1.476 cp_fm_cholesky_decompose 22 10.9 1.321 1.327 1.321 1.327 yz_to_x 606 15.1 0.057 0.059 1.275 1.300 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.182 1.191 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.175 1.186 acc_transpose_blocks_sync 27432 16.5 1.124 1.172 1.124 1.172 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=57.282000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=711.181818, yerr=17.765227 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 499.683328E+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 752263. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.030 101.722 101.722 qs_mol_dyn_low 1 2.0 0.003 0.003 101.416 101.427 qs_forces 11 3.9 0.008 0.012 101.334 101.338 qs_energies 11 4.9 0.002 0.002 97.999 98.012 scf_env_do_scf 11 5.9 0.000 0.001 87.602 87.605 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 78.415 78.416 qs_scf_new_mos 99 7.5 0.001 0.001 55.786 55.918 qs_scf_loop_do_ot 99 8.5 0.001 0.001 55.785 55.918 dbcsr_multiply_generic 2055 12.4 0.107 0.109 54.624 55.004 ot_scf_mini 99 9.5 0.002 0.002 53.202 53.341 velocity_verlet 10 3.0 0.001 0.002 52.112 52.114 multiply_cannon 2055 13.4 0.190 0.196 43.354 44.688 multiply_cannon_loop 2055 14.4 1.800 1.869 41.587 42.968 ot_mini 99 10.5 0.001 0.001 29.128 29.227 qs_ot_get_derivative 99 11.5 0.001 0.001 22.189 22.308 rebuild_ks_matrix 110 8.3 0.000 0.000 18.063 18.210 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 18.063 18.209 multiply_cannon_multrec 49320 15.4 11.274 11.996 17.335 17.903 qs_ks_update_qs_env 110 7.6 0.001 0.001 15.933 16.061 qs_ot_get_p 110 10.4 0.001 0.001 15.031 15.191 mp_waitall_1 220248 16.4 13.653 15.114 13.653 15.114 qs_ot_p2m_diag 48 11.0 0.013 0.019 10.644 10.702 multiply_cannon_sync_h2d 49320 15.4 9.354 10.054 9.354 10.054 sum_up_and_integrate 110 10.3 0.002 0.003 9.886 9.905 integrate_v_rspace 110 11.3 0.003 0.003 9.861 9.885 cp_dbcsr_syevd 48 12.0 0.003 0.004 9.601 9.604 qs_rho_update_rho_low 110 7.6 0.001 0.001 9.305 9.418 calculate_rho_elec 110 8.6 0.021 0.025 9.305 9.418 init_scf_loop 11 6.9 0.000 0.001 9.137 9.138 cp_fm_diag_elpa 48 13.0 0.000 0.000 8.517 8.539 cp_fm_diag_elpa_base 48 14.0 8.477 8.505 8.513 8.535 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.974 8.492 init_scf_run 11 5.9 0.000 0.001 7.989 7.990 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.989 7.990 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.347 7.938 apply_single 110 13.6 0.000 0.001 7.347 7.938 multiply_cannon_metrocomm3 49320 15.4 0.085 0.089 6.143 7.376 make_m2s 4110 13.4 0.061 0.064 6.544 6.746 ot_diis_step 99 11.5 0.005 0.006 6.722 6.723 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.659 6.723 make_images 4110 14.4 0.179 0.193 6.448 6.652 prepare_preconditioner 11 7.9 0.000 0.000 6.579 6.594 make_preconditioner 11 8.9 0.000 0.000 6.579 6.594 density_rs2pw 110 9.6 0.004 0.005 6.246 6.393 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.309 6.373 dbcsr_mm_accdrv_process 87628 16.1 2.985 3.172 5.932 6.260 pw_transfer 1331 11.6 0.056 0.070 5.667 5.905 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 5.577 5.821 multiply_cannon_metrocomm1 49320 15.4 0.065 0.069 3.043 5.439 potential_pw2rs 110 12.3 0.006 0.006 5.254 5.406 mp_sum_l 6594 12.7 4.278 5.242 4.278 5.242 fft3d_ps 1111 14.6 0.785 0.882 4.882 5.104 wfi_extrapolate 11 7.9 0.001 0.001 5.033 5.033 fft_wrap_pw1pw2_140 451 13.1 0.381 0.421 4.308 4.547 calculate_dm_sparse 110 9.5 0.001 0.001 4.134 4.238 transfer_rs2pw 451 10.6 0.005 0.006 3.618 3.866 mp_alltoall_z22v 1111 16.6 3.473 3.796 3.473 3.796 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.773 3.778 cp_fm_cholesky_invert 11 10.9 3.757 3.766 3.757 3.766 transfer_pw2rs 451 13.1 0.006 0.006 3.511 3.665 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.553 3.612 grid_integrate_task_list 110 12.3 3.210 3.437 3.210 3.437 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.294 3.353 mp_alltoall_d11v 2046 13.8 2.912 3.310 2.912 3.310 make_images_data 4110 15.4 0.046 0.049 2.808 3.255 make_images_sizes 4110 15.4 0.004 0.004 2.207 2.982 mp_alltoall_i44 4110 16.4 2.203 2.978 2.203 2.978 hybrid_alltoall_any 4261 16.3 0.086 0.480 2.338 2.837 calculate_first_density_matrix 1 7.0 0.000 0.000 2.814 2.818 jit_kernel_multiply 13 15.9 2.673 2.728 2.673 2.728 acc_transpose_blocks 49320 15.4 0.211 0.223 2.577 2.702 mp_waitany 14300 13.8 2.229 2.464 2.229 2.464 grid_collocate_task_list 110 9.6 2.162 2.301 2.162 2.301 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.265 2.286 yz_to_x 341 15.4 0.060 0.069 1.840 2.134 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=101.722000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.272727, yerr=1.958242 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 592.805888E+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 3001529. 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 0.023 0.064 92.125 92.125 qs_mol_dyn_low 1 2.0 0.003 0.003 91.726 91.774 qs_forces 11 3.9 0.003 0.003 91.406 91.408 qs_energies 11 4.9 0.004 0.022 87.710 87.740 scf_env_do_scf 11 5.9 0.001 0.002 77.853 77.856 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 65.540 65.543 velocity_verlet 10 3.0 0.001 0.002 47.850 47.858 dbcsr_multiply_generic 2055 12.4 0.120 0.125 45.141 46.830 qs_scf_new_mos 99 7.5 0.001 0.001 45.308 45.533 qs_scf_loop_do_ot 99 8.5 0.001 0.001 45.307 45.533 ot_scf_mini 99 9.5 0.003 0.004 43.169 43.404 multiply_cannon 2055 13.4 0.228 0.249 33.645 38.085 multiply_cannon_loop 2055 14.4 1.180 1.216 31.418 36.024 ot_mini 99 10.5 0.001 0.001 24.256 24.399 mp_waitall_1 176588 16.5 11.995 20.296 11.995 20.296 qs_ot_get_derivative 99 11.5 0.001 0.001 17.788 18.049 rebuild_ks_matrix 110 8.3 0.000 0.001 16.153 16.300 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 16.153 16.299 multiply_cannon_multrec 24660 15.4 6.650 9.030 13.693 15.973 multiply_cannon_metrocomm3 24660 15.4 0.073 0.077 6.844 14.939 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.290 14.415 init_scf_loop 11 6.9 0.001 0.004 12.250 12.254 qs_ot_get_p 110 10.4 0.001 0.002 11.282 11.446 prepare_preconditioner 11 7.9 0.000 0.000 9.861 9.878 make_preconditioner 11 8.9 0.000 0.001 9.861 9.878 make_full_inverse_cholesky 11 9.9 0.000 0.000 9.310 9.557 sum_up_and_integrate 110 10.3 0.002 0.002 8.655 8.669 integrate_v_rspace 110 11.3 0.003 0.003 8.628 8.639 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.311 8.401 calculate_rho_elec 110 8.6 0.039 0.047 8.310 8.401 qs_ot_p2m_diag 48 11.0 0.029 0.044 8.015 8.085 cp_dbcsr_syevd 48 12.0 0.003 0.004 7.299 7.301 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.622 7.239 apply_single 110 13.6 0.000 0.001 6.622 7.239 dbcsr_mm_accdrv_process 52282 16.1 5.494 6.387 6.882 7.214 multiply_cannon_sync_h2d 24660 15.4 5.948 7.211 5.948 7.211 init_scf_run 11 5.9 0.000 0.001 7.190 7.191 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.190 7.190 make_m2s 4110 13.4 0.057 0.059 6.380 7.037 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.254 6.971 make_images 4110 14.4 0.412 0.462 6.272 6.925 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.276 6.289 cp_fm_diag_elpa_base 48 14.0 6.200 6.237 6.272 6.285 ot_diis_step 99 11.5 0.010 0.010 6.152 6.153 mp_sum_l 6594 12.7 4.500 5.770 4.500 5.770 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.405 5.531 density_rs2pw 110 9.6 0.004 0.005 5.206 5.412 cp_fm_cholesky_invert 11 10.9 5.160 5.178 5.160 5.178 pw_transfer 1331 11.6 0.067 0.075 4.779 4.996 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.672 4.893 potential_pw2rs 110 12.3 0.009 0.009 4.355 4.435 wfi_extrapolate 11 7.9 0.001 0.001 4.390 4.390 make_images_data 4110 15.4 0.049 0.053 3.379 4.161 fft3d_ps 1111 14.6 1.110 1.338 3.776 4.005 hybrid_alltoall_any 4261 16.3 0.108 0.457 2.874 3.883 calculate_dm_sparse 110 9.5 0.001 0.001 3.525 3.586 fft_wrap_pw1pw2_140 451 13.1 0.453 0.470 3.355 3.584 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.461 3.465 grid_integrate_task_list 110 12.3 3.119 3.392 3.119 3.392 mp_alltoall_d11v 2046 13.8 2.802 3.352 2.802 3.352 transfer_rs2pw 451 10.6 0.006 0.007 2.955 3.099 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.973 3.035 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.778 2.920 transfer_pw2rs 451 13.1 0.005 0.006 2.833 2.904 mp_allgather_i34 2055 14.4 1.526 2.886 1.526 2.886 cp_fm_cholesky_decompose 22 10.9 2.813 2.819 2.813 2.819 calculate_first_density_matrix 1 7.0 0.000 0.002 2.674 2.678 mp_alltoall_z22v 1111 16.6 2.295 2.449 2.295 2.449 multiply_cannon_metrocomm4 22605 15.4 0.076 0.083 0.852 2.363 grid_collocate_task_list 110 9.6 2.168 2.324 2.168 2.324 mp_sum_d 3889 11.9 1.777 2.283 1.777 2.283 make_images_sizes 4110 15.4 0.005 0.005 1.528 2.171 mp_alltoall_i44 4110 16.4 1.524 2.166 1.524 2.166 mp_irecv_dv 57340 16.2 0.728 2.136 0.728 2.136 dbcsr_complete_redistribute 325 12.2 0.230 0.292 1.707 2.055 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.026 2.052 acc_transpose_blocks 24660 15.4 0.117 0.120 2.016 2.049 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.933 1.936 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=92.125000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=559.272727, yerr=7.020631 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 668.745728E+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 0.037 0.102 83.465 83.480 qs_mol_dyn_low 1 2.0 0.003 0.003 82.994 83.018 qs_forces 11 3.9 0.003 0.004 82.918 82.919 qs_energies 11 4.9 0.004 0.019 79.433 79.460 scf_env_do_scf 11 5.9 0.001 0.002 69.855 69.856 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 56.313 56.315 velocity_verlet 10 3.0 0.001 0.002 44.311 44.313 dbcsr_multiply_generic 2055 12.4 0.112 0.118 36.810 38.474 qs_scf_new_mos 99 7.5 0.001 0.001 37.749 37.923 qs_scf_loop_do_ot 99 8.5 0.001 0.001 37.748 37.922 ot_scf_mini 99 9.5 0.002 0.003 36.056 36.240 multiply_cannon 2055 13.4 0.216 0.228 25.679 29.629 multiply_cannon_loop 2055 14.4 0.825 0.869 23.670 28.165 ot_mini 99 10.5 0.001 0.001 19.771 20.018 mp_waitall_1 139946 16.5 11.563 18.496 11.563 18.496 rebuild_ks_matrix 110 8.3 0.000 0.000 14.444 14.979 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 14.443 14.978 qs_ot_get_derivative 99 11.5 0.001 0.001 14.406 14.589 init_scf_loop 11 6.9 0.001 0.006 13.481 13.490 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.796 13.289 multiply_cannon_metrocomm3 16440 15.4 0.047 0.048 6.189 12.812 prepare_preconditioner 11 7.9 0.000 0.000 11.373 11.402 make_preconditioner 11 8.9 0.000 0.001 11.373 11.402 multiply_cannon_multrec 16440 15.4 3.424 4.863 9.789 11.279 make_full_inverse_cholesky 11 9.9 0.000 0.000 10.533 10.987 qs_ot_get_p 110 10.4 0.001 0.003 10.379 10.643 sum_up_and_integrate 110 10.3 0.002 0.002 8.073 8.087 integrate_v_rspace 110 11.3 0.003 0.003 8.046 8.061 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.958 8.034 calculate_rho_elec 110 8.6 0.059 0.059 7.958 8.034 qs_ot_p2m_diag 48 11.0 0.042 0.044 7.370 7.416 make_m2s 4110 13.4 0.050 0.053 6.330 6.988 make_images 4110 14.4 0.402 0.526 6.216 6.873 init_scf_run 11 5.9 0.000 0.001 6.788 6.790 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.788 6.790 cp_dbcsr_syevd 48 12.0 0.004 0.004 6.748 6.750 dbcsr_mm_accdrv_process 34862 16.1 5.468 5.894 6.222 6.425 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.334 5.750 apply_single 110 13.6 0.000 0.000 5.333 5.750 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.716 5.726 cp_fm_diag_elpa_base 48 14.0 5.625 5.670 5.713 5.722 mp_sum_l 6594 12.7 4.187 5.686 4.187 5.686 cp_fm_cholesky_invert 11 10.9 5.477 5.498 5.477 5.498 ot_diis_step 99 11.5 0.011 0.011 5.093 5.093 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.424 5.085 density_rs2pw 110 9.6 0.004 0.005 4.865 5.001 pw_transfer 1331 11.6 0.066 0.073 4.635 4.671 qs_ot_get_derivative_diag 47 12.0 0.001 0.002 4.486 4.595 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.527 4.564 make_images_data 4110 15.4 0.046 0.050 3.561 4.501 wfi_extrapolate 11 7.9 0.001 0.001 4.048 4.049 hybrid_alltoall_any 4261 16.3 0.110 0.386 3.078 3.912 potential_pw2rs 110 12.3 0.011 0.012 3.726 3.752 multiply_cannon_sync_h2d 16440 15.4 3.142 3.742 3.142 3.742 fft3d_ps 1111 14.6 1.094 1.107 3.472 3.507 fft_wrap_pw1pw2_140 451 13.1 0.576 0.583 3.454 3.494 grid_integrate_task_list 110 12.3 3.179 3.402 3.179 3.402 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.279 3.281 cp_fm_cholesky_decompose 22 10.9 3.092 3.115 3.092 3.115 mp_alltoall_d11v 2046 13.8 2.767 3.074 2.767 3.074 calculate_dm_sparse 110 9.5 0.001 0.001 3.037 3.070 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.669 2.857 transfer_rs2pw 451 10.6 0.005 0.006 2.676 2.834 mp_allgather_i34 2055 14.4 1.312 2.771 1.312 2.771 multiply_cannon_metrocomm4 14385 15.4 0.047 0.053 0.931 2.749 dbcsr_complete_redistribute 325 12.2 0.311 0.345 2.114 2.659 mp_irecv_dv 48980 15.7 0.858 2.616 0.858 2.616 calculate_first_density_matrix 1 7.0 0.000 0.002 2.608 2.615 transfer_pw2rs 451 13.1 0.005 0.005 2.363 2.395 grid_collocate_task_list 110 9.6 2.222 2.361 2.222 2.361 mp_alltoall_z22v 1111 16.6 2.161 2.198 2.161 2.198 make_images_sizes 4110 15.4 0.005 0.005 1.322 2.182 mp_alltoall_i44 4110 16.4 1.318 2.177 1.318 2.177 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.054 2.173 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.088 2.091 cp_fm_upper_to_full 70 14.2 1.481 1.966 1.481 1.966 copy_fm_to_dbcsr 174 11.2 0.001 0.002 1.393 1.945 mp_waitany 17072 13.8 1.681 1.928 1.681 1.928 mp_sum_d 3887 11.9 1.451 1.916 1.451 1.916 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.852 1.894 mp_alltoall_i22 605 13.7 1.013 1.694 1.013 1.694 acc_transpose_blocks 16440 15.4 0.078 0.080 1.601 1.672 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=83.480000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=633.090909, yerr=7.844564 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 741.416960E+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 0.012 0.032 89.571 89.572 qs_mol_dyn_low 1 2.0 0.003 0.003 89.281 89.291 qs_forces 11 3.9 0.003 0.003 89.196 89.197 qs_energies 11 4.9 0.002 0.002 85.427 85.433 scf_env_do_scf 11 5.9 0.000 0.001 75.688 75.690 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 57.675 57.676 velocity_verlet 10 3.0 0.002 0.002 49.494 49.497 qs_scf_new_mos 99 7.5 0.001 0.001 39.074 39.193 qs_scf_loop_do_ot 99 8.5 0.001 0.001 39.074 39.192 ot_scf_mini 99 9.5 0.003 0.003 37.054 37.148 dbcsr_multiply_generic 2055 12.4 0.116 0.122 36.200 36.819 multiply_cannon 2055 13.4 0.241 0.262 25.299 26.597 multiply_cannon_loop 2055 14.4 1.417 1.487 23.057 24.741 ot_mini 99 10.5 0.001 0.001 19.292 19.443 init_scf_loop 11 6.9 0.000 0.001 17.947 17.949 prepare_preconditioner 11 7.9 0.000 0.000 15.902 15.919 make_preconditioner 11 8.9 0.000 0.000 15.901 15.919 make_full_inverse_cholesky 11 9.9 0.000 0.000 13.860 15.478 qs_ot_get_derivative 99 11.5 0.001 0.001 14.553 14.661 rebuild_ks_matrix 110 8.3 0.000 0.000 14.372 14.524 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 14.371 14.524 multiply_cannon_multrec 24660 15.4 4.053 6.868 12.974 14.236 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.743 12.876 qs_ot_get_p 110 10.4 0.001 0.002 11.842 12.010 mp_waitall_1 121746 16.5 7.814 11.438 7.814 11.438 dbcsr_mm_accdrv_process 52304 16.0 7.537 8.846 8.767 9.735 qs_ot_p2m_diag 48 11.0 0.055 0.064 8.849 8.872 make_m2s 4110 13.4 0.060 0.062 8.028 8.551 make_images 4110 14.4 0.582 0.704 7.884 8.409 sum_up_and_integrate 110 10.3 0.001 0.002 8.224 8.237 integrate_v_rspace 110 11.3 0.003 0.003 8.198 8.211 cp_dbcsr_syevd 48 12.0 0.004 0.004 8.070 8.071 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.034 8.052 calculate_rho_elec 110 8.6 0.078 0.082 8.033 8.051 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.824 6.835 cp_fm_diag_elpa_base 48 14.0 6.605 6.694 6.819 6.831 init_scf_run 11 5.9 0.000 0.001 6.687 6.687 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.686 6.687 cp_fm_cholesky_invert 11 10.9 6.076 6.093 6.076 6.093 multiply_cannon_metrocomm3 24660 15.4 0.038 0.040 1.958 5.617 cp_fm_upper_to_full 70 14.2 3.681 5.365 3.681 5.365 make_images_data 4110 15.4 0.049 0.052 4.250 5.062 pw_transfer 1331 11.6 0.066 0.076 4.977 5.021 density_rs2pw 110 9.6 0.004 0.004 4.900 5.016 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 4.875 4.929 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.869 4.917 dbcsr_complete_redistribute 325 12.2 0.418 0.476 3.514 4.776 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.546 4.596 ot_diis_step 99 11.5 0.011 0.011 4.587 4.588 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.319 4.563 apply_single 110 13.6 0.000 0.000 4.319 4.563 hybrid_alltoall_any 4261 16.3 0.125 0.469 3.515 4.525 wfi_extrapolate 11 7.9 0.001 0.001 4.045 4.045 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.681 3.925 cp_fm_cholesky_decompose 22 10.9 3.774 3.824 3.774 3.824 potential_pw2rs 110 12.3 0.013 0.014 3.794 3.819 fft3d_ps 1111 14.6 1.094 1.131 3.753 3.778 fft_wrap_pw1pw2_140 451 13.1 0.608 0.628 3.702 3.753 calculate_dm_sparse 110 9.5 0.001 0.001 3.385 3.436 grid_integrate_task_list 110 12.3 3.227 3.398 3.227 3.398 mp_alltoall_i22 605 13.7 1.965 3.256 1.965 3.256 mp_alltoall_d11v 2046 13.8 2.948 3.239 2.948 3.239 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.033 3.236 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.084 3.089 mp_sum_l 6594 12.7 2.198 2.905 2.198 2.905 transfer_rs2pw 451 10.6 0.005 0.006 2.502 2.666 mp_allgather_i34 2055 14.4 1.324 2.584 1.324 2.584 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.508 2.550 acc_transpose_blocks 24660 15.4 0.114 0.116 2.411 2.541 calculate_first_density_matrix 1 7.0 0.000 0.000 2.522 2.525 mp_alltoall_z22v 1111 16.6 2.439 2.506 2.439 2.506 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.370 2.450 multiply_cannon_sync_h2d 24660 15.4 2.317 2.425 2.317 2.425 grid_collocate_task_list 110 9.6 2.270 2.411 2.270 2.411 make_images_sizes 4110 15.4 0.005 0.005 1.445 2.397 mp_alltoall_i44 4110 16.4 1.440 2.393 1.440 2.393 transfer_pw2rs 451 13.1 0.005 0.005 2.307 2.332 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.327 2.329 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.884 1.901 mp_waitany 13376 13.8 1.608 1.871 1.608 1.871 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=89.572000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=700.000000, yerr=11.724101 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 863.354880E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.041 76.540 76.542 qs_mol_dyn_low 1 2.0 0.003 0.003 76.230 76.255 qs_forces 11 3.9 0.003 0.003 76.150 76.151 qs_energies 11 4.9 0.002 0.002 72.184 72.194 scf_env_do_scf 11 5.9 0.000 0.001 62.462 62.462 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 49.287 49.289 velocity_verlet 10 3.0 0.002 0.003 42.146 42.149 qs_scf_new_mos 99 7.5 0.001 0.001 30.960 31.019 qs_scf_loop_do_ot 99 8.5 0.001 0.001 30.959 31.018 ot_scf_mini 99 9.5 0.002 0.002 29.407 29.439 dbcsr_multiply_generic 2055 12.4 0.146 0.151 28.249 28.493 multiply_cannon 2055 13.4 0.244 0.262 19.712 21.633 multiply_cannon_loop 2055 14.4 0.606 0.623 18.057 18.649 ot_mini 99 10.5 0.001 0.001 14.381 14.409 rebuild_ks_matrix 110 8.3 0.000 0.000 13.856 13.895 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.856 13.894 init_scf_loop 11 6.9 0.001 0.001 13.100 13.103 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.355 12.392 mp_waitall_1 103326 16.6 9.487 11.463 9.487 11.463 prepare_preconditioner 11 7.9 0.000 0.000 11.128 11.132 make_preconditioner 11 8.9 0.000 0.000 11.128 11.132 make_full_inverse_cholesky 11 9.9 0.000 0.000 10.508 10.691 qs_ot_get_derivative 99 11.5 0.001 0.001 10.418 10.447 qs_ot_get_p 110 10.4 0.002 0.002 10.417 10.446 multiply_cannon_multrec 8220 15.4 3.173 4.345 7.607 8.551 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.141 8.152 calculate_rho_elec 110 8.6 0.115 0.115 8.141 8.152 sum_up_and_integrate 110 10.3 0.002 0.002 7.987 8.001 qs_ot_p2m_diag 48 11.0 0.081 0.084 7.968 7.980 integrate_v_rspace 110 11.3 0.003 0.003 7.961 7.974 cp_dbcsr_syevd 48 12.0 0.004 0.004 7.413 7.415 make_m2s 4110 13.4 0.038 0.039 6.379 6.677 make_images 4110 14.4 0.656 0.719 6.249 6.548 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.349 6.357 cp_fm_diag_elpa_base 48 14.0 6.248 6.307 6.343 6.351 init_scf_run 11 5.9 0.000 0.001 6.234 6.234 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.234 6.234 cp_fm_cholesky_invert 11 10.9 5.788 5.795 5.788 5.795 pw_transfer 1331 11.6 0.066 0.071 5.322 5.351 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 5.214 5.246 dbcsr_mm_accdrv_process 17442 15.9 3.133 4.201 4.293 5.238 density_rs2pw 110 9.6 0.004 0.004 4.785 4.871 fft_wrap_pw1pw2_140 451 13.1 0.773 0.784 4.209 4.239 make_images_data 4110 15.4 0.042 0.047 3.464 4.193 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 2.765 4.077 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.934 3.985 apply_single 110 13.6 0.000 0.000 3.934 3.985 fft3d_ps 1111 14.6 1.145 1.195 3.883 3.918 ot_diis_step 99 11.5 0.012 0.012 3.890 3.890 hybrid_alltoall_any 4261 16.3 0.202 0.859 3.227 3.823 wfi_extrapolate 11 7.9 0.001 0.001 3.703 3.704 cp_fm_cholesky_decompose 22 10.9 3.581 3.609 3.581 3.609 grid_integrate_task_list 110 12.3 3.369 3.520 3.369 3.520 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 2.663 3.457 potential_pw2rs 110 12.3 0.015 0.016 3.401 3.421 mp_alltoall_d11v 2046 13.8 2.957 3.255 2.957 3.255 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.190 3.204 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.036 3.073 calculate_dm_sparse 110 9.5 0.001 0.001 2.837 2.874 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.800 2.803 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.798 2.799 mp_alltoall_z22v 1111 16.6 2.496 2.538 2.496 2.538 grid_collocate_task_list 110 9.6 2.381 2.477 2.381 2.477 multiply_cannon_sync_h2d 8220 15.4 2.353 2.437 2.353 2.437 make_images_sizes 4110 15.4 0.005 0.005 1.301 2.436 mp_alltoall_i44 4110 16.4 1.296 2.432 1.296 2.432 calculate_first_density_matrix 1 7.0 0.000 0.000 2.400 2.401 transfer_rs2pw 451 10.6 0.005 0.005 2.295 2.400 dbcsr_complete_redistribute 325 12.2 0.559 0.592 2.179 2.394 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.243 2.257 mp_allgather_i34 2055 14.4 0.818 2.254 0.818 2.254 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.787 1.998 transfer_pw2rs 451 13.1 0.004 0.004 1.864 1.887 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.840 1.843 mp_sum_l 6594 12.7 1.395 1.840 1.395 1.840 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.620 1.741 mp_waitany 9240 13.8 1.521 1.625 1.521 1.625 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.588 1.605 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.545 1.558 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=76.542000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=811.545455, yerr=12.815925 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 1.390686E+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 0.017 0.057 110.686 110.698 qs_mol_dyn_low 1 2.0 0.003 0.003 110.257 110.275 qs_forces 11 3.9 0.003 0.004 110.179 110.182 qs_energies 11 4.9 0.002 0.002 105.693 105.713 scf_env_do_scf 11 5.9 0.001 0.001 93.748 93.748 velocity_verlet 10 3.0 0.006 0.008 68.315 68.322 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 59.643 59.644 dbcsr_multiply_generic 2055 12.4 0.122 0.128 37.403 38.316 qs_scf_new_mos 99 7.5 0.001 0.001 37.927 38.034 qs_scf_loop_do_ot 99 8.5 0.001 0.001 37.926 38.033 ot_scf_mini 99 9.5 0.002 0.002 35.689 35.791 init_scf_loop 11 6.9 0.001 0.001 34.011 34.013 prepare_preconditioner 11 7.9 0.000 0.000 31.655 31.676 make_preconditioner 11 8.9 0.000 0.000 31.655 31.676 make_full_inverse_cholesky 11 9.9 0.000 0.000 25.728 30.944 multiply_cannon 2055 13.4 0.350 0.378 26.248 27.188 multiply_cannon_loop 2055 14.4 0.837 0.860 23.668 25.649 ot_mini 99 10.5 0.001 0.001 19.766 19.945 cp_fm_upper_to_full 70 14.2 12.604 18.108 12.604 18.108 rebuild_ks_matrix 110 8.3 0.000 0.001 16.394 16.605 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 16.393 16.605 mp_waitall_1 84994 16.7 12.456 16.232 12.456 16.232 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.811 15.001 qs_ot_get_derivative 99 11.5 0.001 0.001 14.284 14.377 dbcsr_complete_redistribute 325 12.2 1.021 1.045 8.001 11.150 multiply_cannon_multrec 8220 15.4 4.325 4.799 10.252 10.987 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.692 9.845 qs_ot_get_p 110 10.4 0.002 0.002 9.568 9.760 qs_rho_update_rho_low 110 7.6 0.001 0.001 9.542 9.640 calculate_rho_elec 110 8.6 0.225 0.226 9.542 9.639 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 7.185 9.161 make_m2s 4110 13.4 0.043 0.044 8.028 9.130 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.911 9.046 make_images 4110 14.4 0.911 0.970 7.837 8.938 mp_alltoall_i22 605 13.7 5.553 8.811 5.553 8.811 cp_fm_cholesky_invert 11 10.9 8.709 8.717 8.709 8.717 sum_up_and_integrate 110 10.3 0.002 0.002 8.661 8.681 integrate_v_rspace 110 11.3 0.003 0.003 8.633 8.653 init_scf_run 11 5.9 0.000 0.001 7.333 7.333 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.332 7.333 qs_ot_p2m_diag 48 11.0 0.151 0.156 7.056 7.072 pw_transfer 1331 11.6 0.075 0.075 6.806 6.839 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 6.689 6.722 make_images_data 4110 15.4 0.046 0.049 4.544 6.373 cp_dbcsr_syevd 48 12.0 0.004 0.004 6.323 6.324 dbcsr_mm_accdrv_process 11614 15.7 3.769 4.139 5.776 6.036 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.320 5.897 apply_single 110 13.6 0.000 0.000 5.320 5.897 hybrid_alltoall_any 4261 16.3 0.265 0.565 4.185 5.862 fft_wrap_pw1pw2_140 451 13.1 1.280 1.282 5.574 5.606 density_rs2pw 110 9.6 0.004 0.004 5.402 5.467 ot_diis_step 99 11.5 0.016 0.016 5.336 5.337 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.331 5.332 cp_fm_diag_elpa_base 48 14.0 4.749 4.989 5.328 5.329 fft3d_ps 1111 14.6 1.305 1.316 4.741 4.772 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.140 4.603 wfi_extrapolate 11 7.9 0.001 0.001 4.455 4.456 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 4.093 4.156 cp_fm_cholesky_decompose 22 10.9 3.910 3.929 3.910 3.929 calculate_dm_sparse 110 9.5 0.001 0.001 3.887 3.922 grid_integrate_task_list 110 12.3 3.836 3.881 3.836 3.881 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.781 3.783 mp_alltoall_d11v 2046 13.8 3.559 3.689 3.559 3.689 potential_pw2rs 110 12.3 0.021 0.022 3.645 3.665 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.406 3.414 mp_sum_l 6594 12.7 2.237 3.336 2.237 3.336 mp_alltoall_z22v 1111 16.6 3.181 3.204 3.181 3.204 multiply_cannon_sync_h2d 8220 15.4 3.119 3.154 3.119 3.154 mp_allgather_i34 2055 14.4 1.312 3.047 1.312 3.047 grid_collocate_task_list 110 9.6 2.793 2.893 2.793 2.893 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.665 2.735 calculate_first_density_matrix 1 7.0 0.000 0.000 2.718 2.719 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.299 2.355 qs_ot_get_orbitals 99 10.5 0.000 0.000 2.270 2.348 copy_dbcsr_to_fm 151 11.3 0.003 0.003 2.161 2.234 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=110.698000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1248.545455, yerr=56.000590 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 628.584448E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2529110. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.032 239.320 239.321 qs_mol_dyn_low 1 2.0 0.003 0.004 238.838 238.854 qs_forces 11 3.9 0.005 0.006 238.735 238.737 qs_energies 11 4.9 0.002 0.003 232.669 232.683 scf_env_do_scf 11 5.9 0.001 0.001 214.388 214.392 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 185.904 185.907 velocity_verlet 10 3.0 0.001 0.002 143.737 143.739 qs_scf_new_mos 117 7.6 0.001 0.001 139.942 140.272 qs_scf_loop_do_ot 117 8.6 0.001 0.001 139.941 140.272 ot_scf_mini 117 9.6 0.003 0.003 133.194 133.536 dbcsr_multiply_generic 2507 12.6 0.178 0.183 128.381 129.154 multiply_cannon 2507 13.6 0.249 0.258 102.210 103.882 multiply_cannon_loop 2507 14.6 2.347 2.502 99.214 100.944 ot_mini 117 10.6 0.001 0.001 68.566 68.899 multiply_cannon_multrec 60168 15.6 32.082 35.065 41.706 43.983 qs_ot_get_derivative 117 11.6 0.001 0.001 43.470 43.810 qs_ot_get_p 128 10.4 0.007 0.018 42.051 42.450 rebuild_ks_matrix 128 8.3 0.001 0.001 37.484 38.018 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 37.483 38.017 mp_waitall_1 267128 16.5 32.398 35.798 32.398 35.798 qs_ks_update_qs_env 128 7.6 0.001 0.001 33.725 34.183 qs_ot_p2m_diag 83 11.4 0.079 0.092 34.041 34.160 cp_dbcsr_syevd 83 12.4 0.005 0.006 31.317 31.321 multiply_cannon_sync_h2d 60168 15.6 26.591 29.584 26.591 29.584 init_scf_loop 11 6.9 0.000 0.001 28.374 28.377 cp_fm_diag_elpa 83 13.4 0.000 0.001 26.835 26.867 cp_fm_diag_elpa_base 83 14.4 26.664 26.733 26.828 26.858 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.570 25.380 apply_single 128 13.6 0.001 0.001 24.570 25.380 ot_diis_step 117 11.6 0.008 0.008 24.869 24.870 prepare_preconditioner 11 7.9 0.000 0.000 23.347 23.401 make_preconditioner 11 8.9 0.000 0.000 23.347 23.401 make_full_inverse_cholesky 11 9.9 0.000 0.000 22.480 22.673 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.181 20.477 make_m2s 5014 13.6 0.102 0.109 17.059 17.486 sum_up_and_integrate 128 10.3 0.002 0.004 17.366 17.382 integrate_v_rspace 128 11.3 0.004 0.004 17.306 17.326 make_images 5014 14.6 0.406 0.425 16.884 17.324 qs_rho_update_rho_low 128 7.7 0.001 0.001 16.778 16.875 calculate_rho_elec 128 8.7 0.045 0.064 16.777 16.875 multiply_cannon_metrocomm3 60168 15.6 0.120 0.126 15.091 16.795 init_scf_run 11 5.9 0.000 0.001 13.766 13.766 scf_env_initial_rho_setup 11 6.9 0.001 0.002 13.765 13.766 cp_fm_cholesky_invert 11 10.9 13.687 13.698 13.687 13.698 density_rs2pw 128 9.7 0.006 0.007 10.416 10.947 wfi_extrapolate 11 7.9 0.001 0.001 10.370 10.370 dbcsr_mm_accdrv_process 124484 16.2 4.464 4.896 9.168 10.026 pw_transfer 1547 11.6 0.075 0.093 9.538 9.947 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 9.334 9.748 make_images_data 5014 15.6 0.072 0.077 8.383 9.659 mp_sum_l 7950 12.9 7.843 8.978 7.843 8.978 multiply_cannon_metrocomm1 60168 15.6 0.096 0.101 6.897 8.965 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.763 8.892 calculate_dm_sparse 128 9.5 0.001 0.001 8.646 8.781 hybrid_alltoall_any 5200 16.5 0.300 2.287 7.106 8.592 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.155 8.264 fft3d_ps 1291 14.7 2.186 2.827 7.357 7.924 potential_pw2rs 128 12.3 0.009 0.010 7.586 7.792 fft_wrap_pw1pw2_140 523 13.2 1.105 1.170 7.243 7.662 grid_integrate_task_list 128 12.3 6.912 7.591 6.912 7.591 mp_alltoall_d11v 2415 14.1 5.303 7.410 5.303 7.410 cp_fm_cholesky_decompose 22 10.9 7.258 7.275 7.258 7.275 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.931 6.943 transfer_rs2pw 523 10.6 0.008 0.009 5.638 6.233 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.896 6.011 make_images_sizes 5014 15.6 0.007 0.008 3.273 5.513 mp_alltoall_i44 5014 16.6 3.266 5.507 3.266 5.507 grid_collocate_task_list 128 9.7 4.838 5.232 4.838 5.232 transfer_pw2rs 523 13.3 0.008 0.010 4.813 5.023 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=239.321000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.272727, yerr=5.925153 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 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 11528896499712 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 93.514757E+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 6755939872 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 842.772480E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+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 1463834560144 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 4985592. 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 0.051 0.188 235.678 235.679 qs_mol_dyn_low 1 2.0 0.005 0.016 234.929 234.957 qs_forces 11 3.9 0.006 0.019 234.802 234.804 qs_energies 11 4.9 0.008 0.048 227.541 227.570 scf_env_do_scf 11 5.9 0.001 0.002 209.069 209.079 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 164.661 164.664 velocity_verlet 10 3.0 0.002 0.002 147.199 147.202 qs_scf_new_mos 117 7.6 0.001 0.001 119.773 120.264 qs_scf_loop_do_ot 117 8.6 0.001 0.002 119.772 120.263 ot_scf_mini 117 9.6 0.004 0.006 114.536 115.102 dbcsr_multiply_generic 2507 12.6 0.191 0.200 108.476 111.972 multiply_cannon 2507 13.6 0.508 0.570 81.202 90.578 multiply_cannon_loop 2507 14.6 1.589 1.664 76.455 85.409 ot_mini 117 10.6 0.001 0.001 59.427 60.032 mp_waitall_1 214728 16.6 31.775 53.767 31.775 53.767 init_scf_loop 11 6.9 0.001 0.005 44.279 44.280 multiply_cannon_metrocomm3 30084 15.6 0.099 0.108 19.438 42.052 prepare_preconditioner 11 7.9 0.000 0.001 39.507 39.579 make_preconditioner 11 8.9 0.000 0.001 39.507 39.579 make_full_inverse_cholesky 11 9.9 0.000 0.000 37.974 38.706 qs_ot_get_derivative 117 11.6 0.001 0.002 36.247 36.829 qs_ot_get_p 128 10.4 0.001 0.002 35.790 36.550 rebuild_ks_matrix 128 8.3 0.001 0.001 35.685 36.475 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 35.684 36.475 multiply_cannon_multrec 30084 15.6 19.920 25.287 30.658 36.341 qs_ks_update_qs_env 128 7.6 0.001 0.001 32.120 32.844 qs_ot_p2m_diag 83 11.4 0.189 0.216 29.283 29.512 cp_dbcsr_syevd 83 12.4 0.006 0.007 27.307 27.309 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.565 24.510 apply_single 128 13.6 0.001 0.001 22.565 24.510 cp_fm_cholesky_invert 11 10.9 23.399 23.418 23.399 23.418 cp_fm_diag_elpa 83 13.4 0.000 0.001 22.808 22.834 cp_fm_diag_elpa_base 83 14.4 22.460 22.619 22.800 22.822 ot_diis_step 117 11.6 0.014 0.015 22.801 22.804 multiply_cannon_sync_h2d 30084 15.6 17.023 20.307 17.023 20.307 make_m2s 5014 13.6 0.091 0.097 17.261 19.047 make_images 5014 14.6 1.195 1.387 17.059 18.842 qs_rho_update_rho_low 128 7.7 0.001 0.001 17.216 17.360 calculate_rho_elec 128 8.7 0.088 0.105 17.215 17.359 sum_up_and_integrate 128 10.3 0.002 0.004 16.802 16.836 integrate_v_rspace 128 11.3 0.003 0.004 16.742 16.779 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 15.282 15.682 init_scf_run 11 5.9 0.000 0.001 13.234 13.236 scf_env_initial_rho_setup 11 6.9 0.001 0.002 13.234 13.236 make_images_data 5014 15.6 0.069 0.078 10.036 12.405 multiply_cannon_metrocomm4 27577 15.6 0.104 0.121 4.058 11.682 mp_sum_l 7950 12.9 8.319 11.341 8.319 11.341 hybrid_alltoall_any 5200 16.5 0.359 1.553 8.400 11.285 mp_irecv_dv 69486 16.3 3.855 11.269 3.855 11.269 density_rs2pw 128 9.7 0.006 0.007 10.493 11.170 cp_fm_cholesky_decompose 22 10.9 10.974 11.052 10.974 11.052 dbcsr_mm_accdrv_process 62242 16.2 5.518 6.323 10.236 10.861 pw_transfer 1547 11.6 0.085 0.100 10.121 10.234 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 9.897 10.014 wfi_extrapolate 11 7.9 0.001 0.001 9.739 9.740 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.643 8.352 fft_wrap_pw1pw2_140 523 13.2 1.208 1.232 8.051 8.182 grid_integrate_task_list 128 12.3 7.092 7.777 7.092 7.777 fft3d_ps 1291 14.7 2.846 3.015 7.461 7.556 calculate_dm_sparse 128 9.5 0.001 0.001 7.044 7.257 potential_pw2rs 128 12.3 0.016 0.019 7.081 7.147 mp_allgather_i34 2507 14.6 3.256 6.952 3.256 6.952 mp_alltoall_d11v 2415 14.1 5.493 6.902 5.493 6.902 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 6.514 6.522 qs_ot_get_orbitals 117 10.6 0.001 0.001 6.368 6.484 transfer_rs2pw 523 10.6 0.007 0.008 5.556 6.183 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 5.401 5.643 grid_collocate_task_list 128 9.7 4.997 5.331 4.997 5.331 mp_sum_d 4465 12.1 3.740 5.300 3.740 5.300 dbcsr_complete_redistribute 395 12.7 0.777 0.886 4.127 5.144 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=235.679000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=800.727273, yerr=3.886829 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 947.580928E+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 0.059 0.069 218.750 218.751 qs_mol_dyn_low 1 2.0 0.003 0.003 218.260 218.297 qs_forces 11 3.9 0.004 0.005 218.156 218.159 qs_energies 11 4.9 0.007 0.011 211.144 211.174 scf_env_do_scf 11 5.9 0.001 0.001 193.645 193.649 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 147.036 147.037 velocity_verlet 10 3.0 0.001 0.002 139.719 139.722 qs_scf_new_mos 118 7.6 0.001 0.001 103.961 104.440 qs_scf_loop_do_ot 118 8.6 0.001 0.001 103.960 104.439 ot_scf_mini 118 9.6 0.003 0.004 99.234 99.723 dbcsr_multiply_generic 2527 12.6 0.181 0.191 92.628 95.763 multiply_cannon 2527 13.6 0.513 0.540 66.456 73.435 multiply_cannon_loop 2527 14.6 1.149 1.213 62.270 70.398 ot_mini 118 10.6 0.001 0.001 50.635 51.178 mp_waitall_1 171878 16.6 31.401 46.966 31.401 46.966 init_scf_loop 11 6.9 0.000 0.001 46.473 46.474 prepare_preconditioner 11 7.9 0.000 0.000 42.047 42.119 make_preconditioner 11 8.9 0.000 0.000 42.047 42.119 make_full_inverse_cholesky 11 9.9 0.000 0.000 39.422 41.006 rebuild_ks_matrix 129 8.3 0.001 0.001 33.770 34.783 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.018 33.769 34.782 multiply_cannon_metrocomm3 20216 15.6 0.064 0.072 17.715 33.340 qs_ot_get_p 129 10.4 0.001 0.002 32.432 33.028 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.489 31.412 qs_ot_get_derivative 118 11.6 0.001 0.002 29.760 30.248 qs_ot_p2m_diag 83 11.4 0.266 0.272 26.529 26.641 multiply_cannon_multrec 20216 15.6 12.776 15.839 22.286 25.331 cp_dbcsr_syevd 83 12.4 0.006 0.006 24.862 24.867 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.328 21.691 apply_single 129 13.6 0.001 0.001 20.328 21.691 cp_fm_cholesky_invert 11 10.9 21.550 21.577 21.550 21.577 ot_diis_step 118 11.6 0.018 0.018 20.573 20.574 cp_fm_diag_elpa 83 13.4 0.000 0.001 20.205 20.224 cp_fm_diag_elpa_base 83 14.4 19.651 19.909 20.199 20.219 make_m2s 5054 13.6 0.082 0.086 17.783 19.096 make_images 5054 14.6 1.181 1.286 17.554 18.861 qs_rho_update_rho_low 129 7.7 0.001 0.001 16.876 16.985 calculate_rho_elec 129 8.7 0.133 0.148 16.875 16.984 sum_up_and_integrate 129 10.3 0.002 0.003 16.783 16.809 integrate_v_rspace 129 11.3 0.003 0.004 16.722 16.750 multiply_cannon_sync_h2d 20216 15.6 13.476 15.215 13.476 15.215 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 12.468 12.826 make_images_data 5054 15.6 0.064 0.073 10.689 12.495 init_scf_run 11 5.9 0.000 0.001 12.171 12.171 scf_env_initial_rho_setup 11 6.9 0.003 0.017 12.171 12.171 hybrid_alltoall_any 5240 16.5 0.458 2.075 9.065 11.729 cp_fm_cholesky_decompose 22 10.9 10.881 10.920 10.881 10.920 pw_transfer 1559 11.6 0.087 0.103 10.354 10.527 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 10.128 10.307 multiply_cannon_metrocomm4 17689 15.6 0.066 0.079 3.583 10.144 density_rs2pw 129 9.7 0.006 0.006 9.796 10.091 mp_irecv_dv 50625 16.2 3.456 9.880 3.456 9.880 dbcsr_mm_accdrv_process 41830 16.2 5.745 6.091 9.005 9.129 mp_sum_l 8010 12.9 6.558 9.053 6.558 9.053 wfi_extrapolate 11 7.9 0.001 0.001 8.827 8.827 fft_wrap_pw1pw2_140 527 13.2 1.313 1.346 8.369 8.555 grid_integrate_task_list 129 12.3 7.360 7.956 7.360 7.956 cp_fm_upper_to_full 105 14.8 6.005 7.812 6.005 7.812 fft3d_ps 1301 14.7 2.782 3.015 7.482 7.605 dbcsr_complete_redistribute 395 12.7 1.202 1.266 5.566 7.501 potential_pw2rs 129 12.3 0.021 0.023 6.870 6.946 calculate_dm_sparse 129 9.5 0.001 0.001 6.515 6.652 mp_alltoall_d11v 2423 14.1 5.561 6.576 5.561 6.576 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 5.817 6.443 mp_allgather_i34 2527 14.6 2.657 6.370 2.657 6.370 copy_fm_to_dbcsr 209 11.7 0.002 0.002 4.078 6.016 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.980 5.990 grid_collocate_task_list 129 9.7 5.203 5.636 5.203 5.636 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.975 5.055 transfer_rs2pw 527 10.6 0.007 0.007 4.724 5.050 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.705 5.011 mp_alltoall_i22 718 14.1 2.560 4.688 2.560 4.688 transfer_fm_to_dbcsr 11 9.9 0.020 0.026 2.605 4.519 make_images_sizes 5054 15.6 0.007 0.007 2.193 4.394 mp_alltoall_i44 5054 16.6 2.186 4.388 2.186 4.388 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=218.751000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=893.454545, yerr=10.404227 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 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 11613083000832 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 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 4.387043E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 6024576 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.8 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.155391E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1152312 MPI messages size (bytes): total size 2.039354E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.769793E+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 321600 36333158400 131072 < size <= 4194304 721560 792041881600 4194304 < size <= 16777216 70800 669921130560 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57809. MP_Allreduce 11152 1080. MP_Sync 87 MP_Alltoall 1724 12509439. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.070 221.047 221.047 qs_mol_dyn_low 1 2.0 0.003 0.003 220.616 220.631 qs_forces 11 3.9 0.004 0.004 220.491 220.498 qs_energies 11 4.9 0.003 0.004 213.041 213.048 scf_env_do_scf 11 5.9 0.001 0.001 194.689 194.698 velocity_verlet 10 3.0 0.005 0.010 145.508 145.512 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 138.830 138.831 qs_scf_new_mos 118 7.6 0.001 0.001 97.151 97.448 qs_scf_loop_do_ot 118 8.6 0.001 0.001 97.151 97.447 ot_scf_mini 118 9.6 0.003 0.004 92.351 92.697 dbcsr_multiply_generic 2527 12.6 0.187 0.193 85.667 86.308 multiply_cannon 2527 13.6 0.547 0.578 57.253 60.196 init_scf_loop 11 6.9 0.000 0.001 55.705 55.707 multiply_cannon_loop 2527 14.6 1.842 1.937 53.127 54.765 prepare_preconditioner 11 7.9 0.000 0.000 51.204 51.230 make_preconditioner 11 8.9 0.000 0.000 51.204 51.230 make_full_inverse_cholesky 11 9.9 0.011 0.023 44.668 49.851 ot_mini 118 10.6 0.001 0.001 47.056 47.376 rebuild_ks_matrix 129 8.3 0.001 0.001 32.559 32.849 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 32.558 32.849 multiply_cannon_multrec 30324 15.6 13.837 18.987 26.934 31.661 qs_ot_get_p 129 10.4 0.001 0.002 30.884 31.206 mp_waitall_1 149060 16.7 21.037 30.685 21.037 30.685 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.430 29.687 qs_ot_get_derivative 118 11.6 0.002 0.002 27.154 27.492 qs_ot_p2m_diag 83 11.4 0.343 0.390 25.643 25.698 make_m2s 5054 13.6 0.098 0.100 23.370 24.485 make_images 5054 14.6 1.976 2.305 23.064 24.181 cp_dbcsr_syevd 83 12.4 0.006 0.007 23.930 23.933 cp_fm_cholesky_invert 11 10.9 21.223 21.235 21.223 21.235 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.270 19.809 apply_single 129 13.6 0.001 0.001 19.269 19.809 cp_fm_diag_elpa 83 13.4 0.000 0.001 19.764 19.779 cp_fm_diag_elpa_base 83 14.4 18.600 18.970 19.755 19.768 ot_diis_step 118 11.6 0.018 0.018 19.764 19.767 cp_fm_upper_to_full 105 14.8 11.974 17.602 11.974 17.602 sum_up_and_integrate 129 10.3 0.002 0.003 16.729 16.753 integrate_v_rspace 129 11.3 0.003 0.004 16.669 16.695 qs_rho_update_rho_low 129 7.7 0.001 0.001 16.639 16.690 calculate_rho_elec 129 8.7 0.177 0.193 16.638 16.689 make_images_data 5054 15.6 0.068 0.075 12.827 14.697 multiply_cannon_metrocomm3 30324 15.6 0.050 0.054 6.209 14.487 dbcsr_complete_redistribute 395 12.7 1.512 1.684 10.019 13.999 dbcsr_mm_accdrv_process 62756 16.2 8.579 9.472 12.666 13.156 hybrid_alltoall_any 5240 16.5 0.540 2.242 11.312 13.065 copy_fm_to_dbcsr 209 11.7 0.001 0.002 8.235 12.151 init_scf_run 11 5.9 0.000 0.001 12.118 12.119 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.118 12.119 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.586 11.824 cp_fm_cholesky_decompose 22 10.9 10.963 11.043 10.963 11.043 multiply_cannon_sync_h2d 30324 15.6 10.520 10.994 10.520 10.994 pw_transfer 1559 11.6 0.087 0.102 10.630 10.713 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 10.402 10.489 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.514 10.316 mp_alltoall_i22 718 14.1 5.825 9.856 5.825 9.856 density_rs2pw 129 9.7 0.006 0.006 9.290 9.562 fft_wrap_pw1pw2_140 527 13.2 1.457 1.493 8.792 8.887 wfi_extrapolate 11 7.9 0.001 0.001 8.798 8.799 grid_integrate_task_list 129 12.3 7.582 7.900 7.582 7.900 fft3d_ps 1301 14.7 2.863 2.925 7.556 7.605 mp_alltoall_d11v 2423 14.1 6.477 6.895 6.477 6.895 calculate_dm_sparse 129 9.5 0.001 0.001 6.634 6.722 potential_pw2rs 129 12.3 0.023 0.024 6.476 6.521 multiply_cannon_metrocomm4 25270 15.6 0.083 0.092 2.799 6.235 mp_irecv_dv 76685 16.2 2.647 5.972 2.647 5.972 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.734 5.806 grid_collocate_task_list 129 9.7 5.337 5.686 5.337 5.686 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 5.235 5.347 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.591 4.656 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.493 4.497 transfer_rs2pw 527 10.6 0.007 0.007 4.212 4.480 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.378 4.470 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=221.047000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1086.909091, yerr=17.390983 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.554768E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3995 58302. MP_Allreduce 11065 1167. MP_Sync 86 MP_Alltoall 1700 18828174. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.048 211.272 211.280 qs_mol_dyn_low 1 2.0 0.003 0.003 210.782 210.805 qs_forces 11 3.9 0.004 0.005 210.159 210.163 qs_energies 11 4.9 0.002 0.003 202.381 202.390 scf_env_do_scf 11 5.9 0.001 0.001 183.085 183.094 velocity_verlet 10 3.0 0.002 0.002 136.513 136.549 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 135.743 135.744 qs_scf_new_mos 116 7.6 0.001 0.001 94.330 94.460 qs_scf_loop_do_ot 116 8.6 0.001 0.001 94.329 94.459 ot_scf_mini 116 9.6 0.003 0.004 89.496 89.587 dbcsr_multiply_generic 2485 12.5 0.180 0.190 79.368 80.490 multiply_cannon 2485 13.5 0.567 0.609 56.592 59.878 multiply_cannon_loop 2485 14.5 0.805 0.848 52.266 54.845 init_scf_loop 11 6.9 0.001 0.001 47.140 47.142 ot_mini 116 10.6 0.001 0.001 43.441 43.576 prepare_preconditioner 11 7.9 0.000 0.000 42.916 42.951 make_preconditioner 11 8.9 0.000 0.000 42.916 42.951 make_full_inverse_cholesky 11 9.9 0.016 0.025 40.427 40.932 mp_waitall_1 124680 16.7 29.094 38.693 29.094 38.693 qs_ot_get_p 127 10.4 0.002 0.002 32.551 32.678 rebuild_ks_matrix 127 8.3 0.001 0.001 31.712 32.071 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 31.712 32.070 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.891 29.218 qs_ot_p2m_diag 82 11.4 0.490 0.495 27.436 27.466 cp_fm_cholesky_invert 11 10.9 25.906 25.921 25.906 25.921 qs_ot_get_derivative 116 11.6 0.001 0.002 25.779 25.875 cp_dbcsr_syevd 82 12.4 0.012 0.039 25.710 25.713 multiply_cannon_metrocomm3 9940 15.5 0.026 0.027 12.707 21.883 cp_fm_diag_elpa 82 13.4 0.000 0.001 21.408 21.422 cp_fm_diag_elpa_base 82 14.4 21.104 21.210 21.375 21.382 multiply_cannon_multrec 9940 15.5 10.380 14.418 18.121 20.909 make_m2s 4970 13.5 0.066 0.071 17.463 19.159 make_images 4970 14.5 2.178 2.616 17.165 18.864 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.341 17.858 apply_single 127 13.6 0.001 0.001 17.341 17.858 ot_diis_step 116 11.6 0.020 0.020 17.475 17.475 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.789 16.868 calculate_rho_elec 127 8.7 0.256 0.267 16.788 16.867 sum_up_and_integrate 127 10.3 0.002 0.002 16.382 16.427 integrate_v_rspace 127 11.3 0.004 0.004 16.322 16.369 make_images_data 4970 15.5 0.057 0.064 10.082 12.711 init_scf_run 11 5.9 0.000 0.001 12.117 12.117 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.116 12.117 hybrid_alltoall_any 5155 16.4 0.839 3.740 9.443 12.035 pw_transfer 1535 11.6 0.085 0.094 11.331 11.392 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 11.109 11.175 multiply_cannon_sync_h2d 9940 15.5 10.667 11.063 10.667 11.063 cp_fm_cholesky_decompose 22 10.9 10.886 11.000 10.886 11.000 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 10.457 10.533 fft_wrap_pw1pw2_140 519 13.2 1.787 1.830 9.389 9.460 density_rs2pw 127 9.7 0.005 0.006 8.913 9.228 multiply_cannon_metrocomm1 9940 15.5 0.030 0.030 6.158 9.181 wfi_extrapolate 11 7.9 0.001 0.001 8.856 8.856 dbcsr_mm_accdrv_process 20590 16.1 3.356 4.584 7.401 8.134 grid_integrate_task_list 127 12.3 7.692 8.075 7.692 8.075 fft3d_ps 1281 14.7 2.752 2.832 7.864 7.924 mp_alltoall_d11v 2401 14.1 6.728 7.301 6.728 7.301 dbcsr_complete_redistribute 393 12.7 2.102 2.143 6.467 6.872 calculate_dm_sparse 127 9.5 0.001 0.001 6.619 6.746 mp_allgather_i34 2485 14.5 2.504 6.394 2.504 6.394 potential_pw2rs 127 12.3 0.027 0.028 6.035 6.064 grid_collocate_task_list 127 9.7 5.527 5.774 5.527 5.774 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.485 5.498 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.350 5.352 mp_alltoall_z22v 1281 16.7 4.500 4.659 4.500 4.659 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.563 4.642 copy_fm_to_dbcsr 208 11.6 0.001 0.002 4.140 4.586 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.303 4.443 make_images_sizes 4970 15.5 0.007 0.007 2.020 4.228 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=211.280000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1452.909091, yerr=32.295383 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456868864 0.0% 0.0% 100.0% flops 32 x 32 x 32 1967095021568 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992007581696 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753954734080 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 11613074706432 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176888320 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176888320 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911117062144 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.237301E+12 0.0% 0.0% 100.0% flops max/rank 11.786605E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806449600 0.0% 0.0% 100.0% number of processed stacks 1981024 0.0% 0.0% 100.0% average stack size 0.0 0.0 3435.8 marketing flops 145.655226E+12 ------------------------------------------------------------------------------- # multiplications 2531 max memory usage/rank 3.225235E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101240 MPI messages size (bytes): total size 1.145036E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.310119E+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 45728 35500589056 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592330608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4080 58046. MP_Allreduce 11277 1498. MP_Sync 87 MP_Alltoall 1724 36993608. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.061 354.842 354.854 qs_mol_dyn_low 1 2.0 0.003 0.003 354.199 354.220 qs_forces 11 3.9 0.004 0.005 354.080 354.085 qs_energies 11 4.9 0.002 0.003 344.437 344.449 scf_env_do_scf 11 5.9 0.001 0.001 319.308 319.316 velocity_verlet 10 3.0 0.005 0.006 253.080 253.088 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 172.209 172.212 init_scf_loop 11 6.9 0.001 0.001 146.766 146.776 prepare_preconditioner 11 7.9 0.000 0.000 141.246 141.265 make_preconditioner 11 8.9 0.000 0.000 141.246 141.265 make_full_inverse_cholesky 11 9.9 0.037 0.039 116.372 137.944 qs_scf_new_mos 118 7.6 0.001 0.001 119.143 119.299 qs_scf_loop_do_ot 118 8.6 0.001 0.001 119.142 119.298 ot_scf_mini 118 9.6 0.004 0.004 113.695 113.844 dbcsr_multiply_generic 2531 12.6 0.215 0.227 94.314 94.986 cp_fm_upper_to_full 105 14.8 53.601 77.201 53.601 77.201 multiply_cannon 2531 13.6 0.708 0.761 65.551 67.864 multiply_cannon_loop 2531 14.6 1.076 1.098 60.783 63.097 ot_mini 118 10.6 0.001 0.001 53.271 53.487 cp_fm_cholesky_invert 11 10.9 48.296 48.305 48.296 48.305 dbcsr_complete_redistribute 395 12.7 4.011 4.072 32.050 45.674 qs_ot_get_p 129 10.4 0.002 0.002 43.534 43.800 mp_waitall_1 104660 16.8 37.074 41.835 37.074 41.835 copy_fm_to_dbcsr 209 11.7 0.001 0.002 27.863 41.492 rebuild_ks_matrix 129 8.3 0.001 0.001 40.122 40.300 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 40.121 40.300 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 24.830 38.317 qs_ot_p2m_diag 83 11.4 0.880 0.887 37.583 37.644 qs_ks_update_qs_env 129 7.6 0.001 0.001 37.141 37.311 mp_alltoall_i22 718 14.1 22.365 36.277 22.365 36.277 cp_dbcsr_syevd 83 12.4 0.007 0.007 35.204 35.209 qs_ot_get_derivative 118 11.6 0.002 0.002 31.823 31.982 cp_fm_diag_elpa 83 13.4 0.000 0.000 30.501 30.502 cp_fm_diag_elpa_base 83 14.4 25.756 27.514 30.492 30.493 multiply_cannon_metrocomm3 10124 15.6 0.025 0.027 24.024 26.403 make_m2s 5062 13.6 0.078 0.081 22.587 24.549 make_images 5062 14.6 3.108 3.299 22.106 24.067 qs_rho_update_rho_low 129 7.7 0.001 0.001 21.382 21.397 calculate_rho_elec 129 8.7 0.484 0.485 21.381 21.397 ot_diis_step 118 11.6 0.022 0.023 21.334 21.335 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.379 20.733 apply_single 129 13.6 0.001 0.001 20.379 20.733 sum_up_and_integrate 129 10.3 0.002 0.002 20.324 20.412 integrate_v_rspace 129 11.3 0.004 0.005 20.260 20.349 multiply_cannon_multrec 10124 15.6 10.547 12.466 18.491 18.588 make_images_data 5062 15.6 0.065 0.071 13.353 16.501 pw_transfer 1559 11.6 0.096 0.097 15.934 15.955 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 15.694 15.715 hybrid_alltoall_any 5248 16.5 1.324 3.081 12.753 15.640 multiply_cannon_sync_h2d 10124 15.6 14.419 14.446 14.419 14.446 cp_fm_cholesky_decompose 22 10.9 14.386 14.416 14.386 14.416 init_scf_run 11 5.9 0.000 0.001 14.395 14.398 scf_env_initial_rho_setup 11 6.9 0.001 0.001 14.395 14.398 fft_wrap_pw1pw2_140 527 13.2 3.082 3.199 13.558 13.577 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 12.965 13.061 mp_alltoall_d11v 2423 14.1 11.354 11.501 11.354 11.501 density_rs2pw 129 9.7 0.005 0.005 11.131 11.177 wfi_extrapolate 11 7.9 0.001 0.001 10.971 10.971 fft3d_ps 1301 14.7 2.886 2.906 10.891 10.941 dbcsr_mm_accdrv_process 20942 16.1 4.373 6.403 7.685 9.710 grid_integrate_task_list 129 12.3 8.637 8.830 8.637 8.830 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.258 8.259 calculate_dm_sparse 129 9.5 0.001 0.001 7.408 7.493 mp_alltoall_z22v 1301 16.7 7.352 7.401 7.352 7.401 copy_dbcsr_to_fm 186 11.8 0.004 0.004 7.212 7.333 potential_pw2rs 129 12.3 0.042 0.042 7.193 7.214 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.086 7.163 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=354.854000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2939.636364, yerr=148.748887 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 1.259860E+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 39729645. 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 0.014 0.033 86.122 86.123 qs_energies 1 2.0 0.000 0.000 85.633 85.639 ls_scf 1 3.0 0.000 0.000 84.725 84.731 dbcsr_multiply_generic 111 6.7 0.016 0.017 73.106 73.300 multiply_cannon 111 7.7 0.017 0.020 56.073 57.313 multiply_cannon_loop 111 8.7 0.221 0.245 52.595 54.034 ls_scf_main 1 4.0 0.000 0.000 52.875 52.875 density_matrix_trs4 2 5.0 0.002 0.003 47.191 47.252 ls_scf_init_scf 1 4.0 0.000 0.000 28.816 28.817 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.624 27.682 mp_waitall_1 11031 10.9 22.755 25.690 22.755 25.690 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.466 25.480 multiply_cannon_multrec 2664 9.7 8.145 8.907 15.647 17.436 multiply_cannon_sync_h2d 2664 9.7 13.473 15.211 13.473 15.211 make_m2s 222 7.7 0.009 0.012 13.343 13.822 make_images 222 8.7 0.101 0.112 13.321 13.802 multiply_cannon_metrocomm1 2664 9.7 0.011 0.012 9.774 12.597 make_images_data 222 9.7 0.005 0.006 7.836 8.401 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.461 8.317 dbcsr_mm_accdrv_process 4760 10.4 0.587 0.708 7.122 8.106 hybrid_alltoall_any 227 10.6 0.220 1.850 6.685 7.896 dbcsr_mm_accdrv_process_sort 4760 11.4 6.333 7.232 6.333 7.232 calculate_norms 4752 9.8 5.509 6.144 5.509 6.144 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.010 5.199 mp_sum_l 887 5.1 3.154 4.678 3.154 4.678 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.068 4.026 mp_irecv_dv 6231 10.9 2.051 3.994 2.051 3.994 arnoldi_extremal 4 6.8 0.000 0.000 3.567 3.593 arnoldi_normal_ev 4 7.8 0.001 0.003 3.567 3.593 make_images_sizes 222 9.7 0.000 0.000 0.833 3.576 mp_alltoall_i44 222 10.7 0.833 3.576 0.833 3.576 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.387 3.517 build_subspace 16 8.4 0.009 0.012 3.469 3.473 ls_scf_post 1 4.0 0.000 0.000 3.035 3.040 ls_scf_store_result 1 5.0 0.000 0.000 2.847 2.887 dbcsr_special_finalize 555 9.7 0.005 0.006 2.310 2.806 dbcsr_merge_single_wm 555 10.7 0.478 0.591 2.301 2.797 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.454 2.677 make_images_pack 222 9.7 2.213 2.646 2.215 2.648 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.429 2.539 dbcsr_sort_data 658 11.4 2.078 2.478 2.078 2.478 dbcsr_matrix_vector_mult_local 304 10.0 2.042 2.440 2.044 2.442 buffer_matrices_ensure_size 222 8.7 1.798 2.098 1.798 2.098 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.833 1.835 rebuild_ks_matrix 3 7.3 0.000 0.000 1.824 1.825 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.823 1.825 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.123000, 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 2.191245E+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 0.025 0.076 90.590 90.591 qs_energies 1 2.0 0.000 0.000 90.035 90.040 ls_scf 1 3.0 0.000 0.000 88.691 88.697 dbcsr_multiply_generic 111 6.7 0.016 0.017 74.306 74.632 multiply_cannon 111 7.7 0.027 0.041 52.854 56.779 ls_scf_main 1 4.0 0.000 0.000 54.874 54.882 multiply_cannon_loop 111 8.7 0.136 0.146 50.108 53.920 density_matrix_trs4 2 5.0 0.002 0.003 49.033 49.232 ls_scf_init_scf 1 4.0 0.000 0.000 30.119 30.120 mp_waitall_1 9105 10.9 20.672 29.701 20.672 29.701 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.888 28.971 multiply_cannon_multrec 1332 9.7 13.275 16.495 22.685 27.057 matrix_sqrt_Newton_Schulz 2 6.5 0.007 0.014 26.512 26.522 multiply_cannon_metrocomm3 1332 9.7 0.007 0.009 11.565 21.313 make_m2s 222 7.7 0.007 0.008 14.837 15.316 make_images 222 8.7 1.384 1.710 14.806 15.287 dbcsr_mm_accdrv_process 4041 10.4 0.335 0.514 9.005 10.588 dbcsr_mm_accdrv_process_sort 4041 11.4 8.520 10.061 8.520 10.061 make_images_data 222 9.7 0.005 0.006 8.509 9.243 hybrid_alltoall_any 227 10.6 0.546 2.573 7.877 9.056 mp_sum_l 887 5.1 5.156 8.160 5.156 8.160 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.204 7.683 mp_irecv_dv 3311 11.0 3.184 7.626 3.184 7.626 calculate_norms 2376 9.8 6.060 6.789 6.060 6.789 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.930 6.451 multiply_cannon_sync_h2d 1332 9.7 4.853 6.009 4.853 6.009 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.988 5.260 arnoldi_extremal 4 6.8 0.000 0.000 5.000 5.022 arnoldi_normal_ev 4 7.8 0.006 0.015 5.000 5.022 build_subspace 16 8.4 0.014 0.021 4.740 4.744 ls_scf_post 1 4.0 0.000 0.000 3.698 3.703 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.328 3.526 ls_scf_store_result 1 5.0 0.000 0.000 3.404 3.513 dbcsr_matrix_vector_mult_local 304 10.0 2.771 3.260 2.773 3.261 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.239 2.999 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.707 2.839 mp_allgather_i34 111 8.7 0.726 2.776 0.726 2.776 dbcsr_data_new 4174 10.1 2.117 2.407 2.117 2.407 make_images_pack 222 9.7 1.824 2.137 1.827 2.139 dbcsr_sort_data 436 11.2 1.838 2.087 1.838 2.087 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.923 1.925 rebuild_ks_matrix 3 7.3 0.000 0.000 1.910 1.912 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.910 1.912 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.591000, 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/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 2.921832E+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 0.024 0.043 94.403 94.403 qs_energies 1 2.0 0.000 0.000 93.836 93.840 ls_scf 1 3.0 0.000 0.000 92.424 92.428 dbcsr_multiply_generic 111 6.7 0.017 0.018 76.401 76.750 ls_scf_main 1 4.0 0.000 0.000 57.693 57.697 multiply_cannon 111 7.7 0.033 0.104 52.728 56.682 multiply_cannon_loop 111 8.7 0.118 0.132 49.921 54.348 density_matrix_trs4 2 5.0 0.002 0.003 51.669 51.856 mp_waitall_1 7281 11.0 24.187 34.582 24.187 34.582 ls_scf_init_scf 1 4.0 0.000 0.000 30.991 30.993 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.738 29.833 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 27.364 27.375 multiply_cannon_multrec 888 9.7 12.712 15.285 21.416 24.572 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 11.491 23.946 make_m2s 222 7.7 0.007 0.008 16.649 17.349 make_images 222 8.7 1.597 1.875 16.611 17.313 make_images_data 222 9.7 0.004 0.005 9.666 10.809 hybrid_alltoall_any 227 10.6 0.643 2.940 9.097 10.397 dbcsr_mm_accdrv_process 3754 10.4 0.312 0.520 8.226 9.548 dbcsr_mm_accdrv_process_sort 3754 11.4 7.791 9.028 7.791 9.028 mp_sum_l 887 5.1 5.330 8.901 5.330 8.901 multiply_cannon_sync_h2d 888 9.7 6.051 7.339 6.051 7.339 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.473 7.287 mp_irecv_dv 2335 11.1 2.457 7.229 2.457 7.229 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.150 7.055 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.666 6.660 arnoldi_extremal 4 6.8 0.000 0.000 5.648 5.663 arnoldi_normal_ev 4 7.8 0.001 0.005 5.648 5.663 build_subspace 16 8.4 0.014 0.020 5.344 5.351 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.833 5.091 calculate_norms 1584 9.8 4.346 4.749 4.346 4.749 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.709 3.969 mp_allgather_i34 111 8.7 0.901 3.856 0.901 3.856 ls_scf_post 1 4.0 0.000 0.000 3.741 3.747 dbcsr_matrix_vector_mult_local 304 10.0 3.050 3.628 3.052 3.630 ls_scf_store_result 1 5.0 0.000 0.000 3.476 3.565 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.904 2.999 dbcsr_data_new 4116 9.9 2.112 2.450 2.112 2.450 make_images_sizes 222 9.7 0.000 0.000 1.149 2.448 mp_alltoall_i44 222 10.7 1.149 2.447 1.149 2.447 dbcsr_sort_data 325 11.1 1.856 2.142 1.856 2.142 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.936 1.938 dbcsr_finalize 304 7.8 0.027 0.033 1.615 1.927 rebuild_ks_matrix 3 7.3 0.000 0.000 1.915 1.917 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.915 1.917 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.403000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2288.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 3.331244E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.061 98.585 98.585 qs_energies 1 2.0 0.000 0.000 97.950 97.956 ls_scf 1 3.0 0.000 0.000 96.294 96.300 dbcsr_multiply_generic 111 6.7 0.018 0.020 79.144 79.371 ls_scf_main 1 4.0 0.000 0.000 59.781 59.784 multiply_cannon 111 7.7 0.045 0.087 51.847 56.213 density_matrix_trs4 2 5.0 0.002 0.003 53.521 53.635 multiply_cannon_loop 111 8.7 0.153 0.168 46.916 50.016 ls_scf_init_scf 1 4.0 0.000 0.000 33.248 33.250 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.980 32.063 mp_waitall_1 6369 11.0 23.152 29.799 23.152 29.799 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.490 29.503 multiply_cannon_multrec 1332 9.7 14.264 17.057 22.299 25.253 make_m2s 222 7.7 0.008 0.010 21.627 23.017 make_images 222 8.7 3.159 3.613 21.575 22.967 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.129 16.925 make_images_data 222 9.7 0.004 0.005 12.154 13.789 hybrid_alltoall_any 227 10.6 0.802 3.800 11.448 12.936 dbcsr_mm_accdrv_process 3641 10.4 0.288 0.484 7.681 9.241 dbcsr_mm_accdrv_process_sort 3641 11.4 7.239 8.744 7.239 8.744 mp_sum_l 887 5.1 4.079 7.565 4.079 7.565 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.103 6.091 mp_irecv_dv 3229 10.9 2.078 5.996 2.078 5.996 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.108 5.981 arnoldi_extremal 4 6.8 0.000 0.000 5.893 5.906 arnoldi_normal_ev 4 7.8 0.001 0.005 5.893 5.906 multiply_cannon_sync_h2d 1332 9.7 5.401 5.889 5.401 5.889 build_subspace 16 8.4 0.014 0.021 5.542 5.551 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.693 5.502 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.550 4.735 calculate_norms 2376 9.8 4.183 4.463 4.183 4.463 mp_allgather_i34 111 8.7 2.023 4.319 2.023 4.319 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.891 4.164 dbcsr_matrix_vector_mult_local 304 10.0 3.228 3.735 3.230 3.737 dbcsr_sort_data 658 11.4 3.012 3.349 3.012 3.349 ls_scf_post 1 4.0 0.000 0.000 3.265 3.271 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.142 3.230 dbcsr_special_finalize 555 9.7 0.006 0.007 2.750 3.089 dbcsr_merge_single_wm 555 10.7 0.528 0.651 2.742 3.081 ls_scf_store_result 1 5.0 0.000 0.000 3.006 3.062 dbcsr_data_release 10477 10.7 1.601 2.447 1.601 2.447 dbcsr_finalize 304 7.8 0.050 0.062 1.819 1.996 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.986 1.988 make_images_sizes 222 9.7 0.000 0.000 0.847 1.972 mp_alltoall_i44 222 10.7 0.847 1.972 0.847 1.972 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.585000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2714.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 4.760957E+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 0.036 0.053 93.816 93.816 qs_energies 1 2.0 0.000 0.000 93.033 93.037 ls_scf 1 3.0 0.000 0.000 91.064 91.068 dbcsr_multiply_generic 111 6.7 0.019 0.020 71.395 71.616 ls_scf_main 1 4.0 0.000 0.000 57.452 57.453 multiply_cannon 111 7.7 0.095 0.188 52.685 56.206 multiply_cannon_loop 111 8.7 0.088 0.095 50.008 51.229 density_matrix_trs4 2 5.0 0.002 0.003 50.222 50.305 ls_scf_init_scf 1 4.0 0.000 0.000 30.173 30.175 mp_waitall_1 5436 11.0 24.590 29.206 24.590 29.206 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.828 28.852 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 26.690 26.699 multiply_cannon_multrec 444 9.7 13.744 16.242 20.928 22.639 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.764 15.331 make_m2s 222 7.7 0.006 0.006 14.059 14.952 make_images 222 8.7 2.053 2.493 13.990 14.882 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.104 13.623 hybrid_alltoall_any 227 10.6 0.800 3.823 8.336 10.002 make_images_data 222 9.7 0.003 0.004 8.529 9.965 dbcsr_mm_accdrv_process 3003 10.4 0.319 0.408 6.882 8.019 multiply_cannon_sync_h2d 444 9.7 6.652 7.916 6.652 7.916 dbcsr_mm_accdrv_process_sort 3003 11.4 6.515 7.619 6.515 7.619 arnoldi_extremal 4 6.8 0.000 0.000 6.685 6.697 arnoldi_normal_ev 4 7.8 0.002 0.006 6.685 6.697 build_subspace 16 8.4 0.014 0.019 6.269 6.280 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.555 4.782 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.560 4.706 dbcsr_matrix_vector_mult_local 304 10.0 3.749 4.237 3.751 4.239 mp_sum_l 887 5.1 2.645 4.153 2.645 4.153 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.579 3.919 mp_irecv_dv 1241 11.2 1.563 3.882 1.563 3.882 mp_allgather_i34 111 8.7 1.235 3.817 1.235 3.817 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.679 3.752 calculate_norms 792 9.8 3.633 3.749 3.633 3.749 make_images_sizes 222 9.7 0.000 0.000 1.117 3.701 mp_alltoall_i44 222 10.7 1.117 3.701 1.117 3.701 ls_scf_post 1 4.0 0.000 0.000 3.439 3.443 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.846 3.325 ls_scf_store_result 1 5.0 0.000 0.000 3.225 3.262 dbcsr_finalize 304 7.8 0.062 0.077 2.200 2.271 dbcsr_data_new 4608 9.7 1.776 2.202 1.776 2.202 dbcsr_merge_all 275 8.9 0.482 0.524 2.057 2.108 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.082 2.083 rebuild_ks_matrix 3 7.3 0.000 0.000 2.050 2.051 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.050 2.051 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 1.953 1.953 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=93.816000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3769.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 8.986608E+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 0.083 0.097 107.254 107.254 qs_energies 1 2.0 0.000 0.000 105.830 105.836 ls_scf 1 3.0 0.000 0.000 102.845 102.851 dbcsr_multiply_generic 111 6.7 0.025 0.028 75.179 75.342 ls_scf_main 1 4.0 0.000 0.000 64.517 64.518 density_matrix_trs4 2 5.0 0.002 0.003 55.289 55.353 multiply_cannon 111 7.7 0.141 0.196 48.315 50.813 multiply_cannon_loop 111 8.7 0.098 0.100 45.256 45.587 ls_scf_init_scf 1 4.0 0.000 0.000 34.507 34.507 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.936 32.951 matrix_sqrt_Newton_Schulz 2 6.5 0.009 0.014 30.179 30.184 mp_waitall_1 4527 11.1 21.996 25.841 21.996 25.841 make_m2s 222 7.7 0.007 0.007 23.272 24.421 make_images 222 8.7 3.589 3.895 23.162 24.310 multiply_cannon_multrec 444 9.7 17.842 18.452 22.555 23.194 hybrid_alltoall_any 227 10.6 1.661 3.639 13.098 16.413 make_images_data 222 9.7 0.004 0.004 13.429 15.828 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.927 10.319 multiply_cannon_sync_h2d 444 9.7 8.793 8.852 8.793 8.852 arnoldi_extremal 4 6.8 0.000 0.000 8.015 8.030 arnoldi_normal_ev 4 7.8 0.003 0.009 8.015 8.030 build_subspace 16 8.4 0.026 0.036 7.452 7.465 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.767 5.920 dbcsr_matrix_vector_mult_local 304 10.0 5.139 5.443 5.142 5.445 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.290 5.381 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.904 5.143 dbcsr_mm_accdrv_process 1814 10.4 0.245 0.358 4.533 4.665 dbcsr_mm_accdrv_process_sort 1814 11.4 4.196 4.334 4.196 4.334 make_images_sizes 222 9.7 0.000 0.000 1.667 4.285 mp_alltoall_i44 222 10.7 1.666 4.285 1.666 4.285 ls_scf_post 1 4.0 0.000 0.000 3.821 3.828 mp_allgather_i34 111 8.7 1.159 3.650 1.159 3.650 ls_scf_store_result 1 5.0 0.000 0.000 3.540 3.580 calculate_norms 792 9.8 3.229 3.282 3.229 3.282 dbcsr_finalize 304 7.8 0.082 0.090 3.085 3.146 dbcsr_complete_redistribute 5 7.6 1.437 1.476 2.856 2.977 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.954 2.954 dbcsr_merge_all 275 8.9 0.889 0.923 2.868 2.923 dbcsr_new_transposed 4 7.5 0.262 0.293 2.598 2.620 mp_alltoall_d11v 48 9.2 2.481 2.577 2.481 2.577 matrix_ls_to_qs 2 6.0 0.000 0.000 2.445 2.563 dbcsr_sort_data 325 11.1 2.443 2.502 2.443 2.502 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.497 2.499 rebuild_ks_matrix 3 7.3 0.000 0.000 2.431 2.433 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.431 2.433 dbcsr_data_new 6591 9.6 1.868 2.392 1.868 2.392 dbcsr_redistribute 4 8.5 1.378 1.458 2.309 2.336 dbcsr_frobenius_norm 74 6.6 2.056 2.131 2.184 2.218 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.208 dbcsr_add_anytype 103 7.2 0.859 0.891 2.131 2.208 dbcsr_data_release 12724 10.6 1.973 2.183 1.973 2.183 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.254000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6947.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6636911b7db0cba9d55e8e44d46088c49b56cb27_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 592.908288E+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 693046797. 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 0.038 0.072 374.176 374.187 qs_mol_dyn_low 1 2.0 0.003 0.004 372.763 372.883 qs_forces 5 3.8 0.006 0.041 372.650 372.653 qs_energies 5 4.8 0.005 0.044 368.597 368.640 scf_env_do_scf 5 5.8 0.002 0.036 351.841 351.845 scf_env_do_scf_inner_loop 105 6.6 0.004 0.053 302.051 302.059 qs_scf_new_mos 105 7.6 0.000 0.001 238.543 239.122 qs_scf_loop_do_ot 105 8.6 0.001 0.001 238.543 239.121 ot_scf_mini 105 9.6 0.003 0.005 227.439 227.856 dbcsr_multiply_generic 1445 12.2 0.124 0.133 149.170 151.114 velocity_verlet 4 3.0 0.009 0.040 142.850 142.852 multiply_cannon 1445 13.2 0.293 0.309 122.186 126.941 qs_ot_get_p 112 10.4 0.001 0.001 124.153 124.628 multiply_cannon_loop 1445 14.2 2.824 3.040 118.328 123.288 qs_ot_p2m_diag 40 11.0 0.020 0.033 109.028 109.133 cp_dbcsr_syevd 40 12.0 0.002 0.003 104.657 104.663 cp_fm_syevd 40 13.0 0.000 0.001 96.524 96.754 cp_fm_redistribute_end 40 14.0 44.913 89.581 44.937 89.600 cp_fm_syevd_base 40 14.0 44.620 89.305 44.620 89.305 ot_mini 105 10.6 0.001 0.001 70.216 70.423 mp_waitall_1 488190 16.1 51.122 65.394 51.122 65.394 multiply_cannon_metrocomm3 69360 15.2 0.205 0.219 32.067 50.504 init_scf_loop 7 6.6 0.001 0.008 49.694 49.758 qs_ot_get_derivative 55 11.6 0.001 0.001 47.937 48.261 multiply_cannon_multrec 69360 15.2 28.573 35.058 38.301 43.985 prepare_preconditioner 7 7.6 0.000 0.001 43.370 43.421 make_preconditioner 7 8.6 0.000 0.003 43.370 43.421 rebuild_ks_matrix 110 8.4 0.000 0.000 38.907 39.127 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.014 38.906 39.127 qs_ks_update_qs_env 112 7.6 0.001 0.001 35.694 35.888 qs_rho_update_rho_low 110 7.6 0.001 0.001 32.705 33.117 calculate_rho_elec 110 8.6 0.029 0.032 32.705 33.117 make_full_inverse_cholesky 7 9.6 0.000 0.000 33.028 33.102 multiply_cannon_sync_h2d 69360 15.2 27.348 31.617 27.348 31.617 density_rs2pw 110 9.6 0.006 0.007 25.201 25.739 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.394 23.706 apply_single 62 13.6 0.000 0.000 23.394 23.706 cp_fm_cholesky_invert 7 10.6 21.818 21.840 21.818 21.840 ot_new_cg_direction 55 11.6 0.001 0.001 21.398 21.401 transfer_rs2pw 445 10.6 0.007 0.008 20.475 21.064 multiply_cannon_metrocomm1 69360 15.2 0.095 0.105 7.222 20.009 qs_ot_get_orbitals 105 10.6 0.001 0.001 17.368 17.725 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 16.936 17.158 mp_sum_l 4764 12.2 15.261 17.141 15.261 17.141 sum_up_and_integrate 60 10.3 0.001 0.002 16.753 16.768 integrate_v_rspace 60 11.3 0.002 0.003 16.735 16.752 pw_transfer 1645 12.4 0.081 0.103 15.054 15.647 fft_wrap_pw1pw2 1425 13.5 0.013 0.017 14.914 15.513 mp_sendrecv_dv 168740 12.6 14.282 14.543 14.282 14.543 transfer_rs2pw_30 110 11.6 1.224 1.284 12.747 14.308 check_diag 80 13.5 8.600 8.964 13.599 13.798 init_scf_run 5 5.8 0.000 0.001 13.563 13.564 scf_env_initial_rho_setup 5 6.8 0.001 0.003 13.562 13.563 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 13.308 13.438 calculate_dm_sparse 110 9.5 0.000 0.001 12.797 13.115 fft_wrap_pw1pw2_240 915 15.0 1.138 1.244 12.170 12.753 make_m2s 2890 13.2 0.079 0.084 11.341 12.096 potential_pw2rs 60 12.3 0.003 0.003 11.969 12.064 make_images 2890 14.2 0.246 0.267 11.237 11.994 qs_vxc_create 110 10.4 0.002 0.005 11.775 11.848 fft3d_pb 915 16.0 2.385 2.715 10.360 10.967 dbcsr_mm_accdrv_process 154766 15.8 6.035 6.446 9.604 10.623 cp_fm_cholesky_decompose 14 10.2 10.458 10.470 10.458 10.470 transfer_pw2rs 245 13.2 0.003 0.004 10.213 10.305 make_full_single_inverse 7 9.6 0.001 0.005 9.829 9.880 mp_alltoall_z22v 2340 17.7 7.687 8.465 7.687 8.465 mp_alltoall_d11v 1300 13.8 7.479 8.325 7.479 8.325 xc_rho_set_and_dset_create 110 12.4 0.077 0.099 8.001 8.271 xc_vxc_pw_create 60 11.3 0.039 0.052 7.890 7.960 acc_transpose_blocks 69360 15.2 0.341 0.369 7.343 7.895 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 7.592 7.608 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=374.187000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.200000, yerr=3.487119 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 6636911b7db0cba9d55e8e44d46088c49b56cb27 Summary: empty Status: OK