=== 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: e162f23508ef73105089f58ab0f087a87f283da0 ################# 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 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # 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. # # Author: Matthias Krack (19.04.2023) # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib 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__HDF5 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 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/01 job id: 46318841 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/02 job id: 46318842 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/03 job id: 46318843 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/04 job id: 46318844 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/05 job id: 46318845 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/06 job id: 46318848 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/07 job id: 46318850 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/08 job id: 46318852 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/09 job id: 46318853 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/10 job id: 46318857 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/11 job id: 46318858 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/12 job id: 46318859 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/13 job id: 46318860 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/14 job id: 46318861 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/15 job id: 46318862 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/16 job id: 46318863 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/17 job id: 46318864 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/18 job id: 46318867 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/19 job id: 46318869 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/20 job id: 46318872 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/21 job id: 46318873 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/22 job id: 46318874 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/23 job id: 46318875 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/24 job id: 46318876 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/25 job id: 46318877 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/26 job id: 46318878 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/27 job id: 46318879 --- 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/e162f23508ef73105089f58ab0f087a87f283da0_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 344 9. 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 0.035 134.416 134.416 farming_run 1 2.0 133.821 133.822 134.387 134.390 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.465197E+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 1103589. MP_Allreduce 491 2254389. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.024 115.723 115.723 qs_energies 1 2.0 0.000 0.000 115.538 115.541 mp2_main 1 3.0 0.000 0.000 113.472 113.475 mp2_gpw_main 1 4.0 0.019 0.025 112.465 112.468 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 93.273 93.857 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.401 55.986 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.168 41.714 47.225 get_2c_integrals 1 6.0 0.008 0.009 37.297 37.701 integrate_v_rspace 273 8.0 0.437 0.450 25.088 30.287 pw_transfer 6555 10.6 0.376 0.388 27.410 27.959 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 26.062 26.629 grid_integrate_task_list 273 9.0 20.901 26.572 20.901 26.572 fft_wrap_pw1pw2_100 2178 12.4 1.197 1.459 23.585 24.160 compute_2c_integrals 1 7.0 0.002 0.002 19.652 19.653 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.899 19.385 mp2_eri_2c_integrate_gpw 1 9.0 2.380 2.434 18.897 19.381 rpa_ri_compute_en 1 5.0 0.020 0.022 19.081 19.248 cp_fm_cholesky_decompose 12 8.2 17.590 17.964 17.590 17.964 cholesky_decomp 1 7.0 0.000 0.000 16.482 16.858 fft3d_s 5443 13.4 16.143 16.721 16.166 16.744 ao_to_mo_and_store_B_mult_1 272 7.0 10.854 15.555 10.854 15.555 calculate_wavefunction 272 8.0 5.425 5.580 12.549 13.171 rpa_num_int 1 6.0 0.000 0.000 10.643 10.644 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.631 10.643 calc_mat_Q 8 8.0 0.000 0.000 9.485 9.565 contract_S_to_Q 8 9.0 0.000 0.000 8.906 8.987 calc_potential_gpw 544 9.5 0.005 0.006 8.244 8.673 parallel_gemm_fm 14 9.1 0.000 0.000 8.489 8.597 parallel_gemm_fm_cosma 14 10.1 8.489 8.597 8.489 8.597 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.216 8.539 potential_pw2rs 545 10.0 0.108 0.110 7.695 8.305 create_integ_mat 1 6.0 0.021 0.028 7.757 7.758 collocate_single_gaussian 272 10.0 0.040 0.042 7.484 7.749 array2fm 1 7.0 0.000 0.000 6.765 7.222 pw_scatter_s 2720 13.7 4.439 4.578 4.439 4.578 pw_gather_s 2722 13.2 3.875 4.195 3.875 4.195 array2fm_buffer_send 1 8.0 3.015 3.138 3.015 3.138 pw_poisson_solve 545 10.5 1.100 1.142 2.151 2.415 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.467736, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2805.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 344 10. 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 0.028 0.040 395.190 395.192 farming_run 1 2.0 394.495 394.523 395.143 395.147 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.224090E+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 743 386399. MP_Allreduce 1941 22272. MP_Sync 37 MP_Alltoall 77 7589693. 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 0.010 0.029 209.498 209.499 qs_energies 1 2.0 0.001 0.010 209.293 209.301 scf_env_do_scf 1 3.0 0.000 0.000 106.472 106.472 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.599 105.608 rebuild_ks_matrix 4 6.0 0.000 0.000 105.598 105.606 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.066 105.598 105.606 hfx_ks_matrix 4 8.0 0.001 0.001 105.214 105.218 integrate_four_center 4 9.0 0.143 0.460 105.213 105.218 mp2_main 1 3.0 0.004 0.036 102.529 102.538 mp2_gpw_main 1 4.0 0.034 0.067 101.645 101.659 integrate_four_center_main 4 10.0 0.092 0.639 96.535 99.200 integrate_four_center_bin 263 11.0 96.443 99.177 96.443 99.177 init_scf_loop 1 4.0 0.000 0.000 92.173 92.174 mp2_ri_gpw_compute_in 1 5.0 0.069 0.104 74.863 75.887 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.418 55.417 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.158 42.064 47.169 integrate_v_rspace 95 8.0 0.397 0.572 28.448 33.386 pw_transfer 2240 10.6 0.145 0.159 29.895 30.415 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.922 29.430 grid_integrate_task_list 95 9.0 23.741 28.886 23.741 28.886 ao_to_mo_and_store_B_mult_1 91 7.0 10.663 28.317 10.663 28.317 mp2_ri_gpw_compute_en 1 5.0 0.056 0.062 26.569 28.228 fft_wrap_pw1pw2_100 730 12.4 1.254 1.414 26.624 27.114 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.828 1.903 24.884 24.895 get_2c_integrals 1 6.0 0.000 0.000 20.343 20.377 compute_2c_integrals 1 7.0 0.002 0.003 19.325 19.334 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.806 19.201 mp2_eri_2c_integrate_gpw 1 9.0 1.746 1.838 18.805 19.200 fft3d_s 1823 13.4 18.441 18.915 18.454 18.928 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.297 14.297 calculate_wavefunction 91 8.0 2.011 2.051 9.736 10.003 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.585 8.687 9.413 potential_pw2rs 186 10.0 0.034 0.035 8.627 9.183 local_gemm 172 8.0 8.130 8.858 8.130 8.858 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.207 8.592 mp2_ri_gpw_compute_en_comm 22 7.0 0.501 0.526 7.967 8.529 collocate_single_gaussian 91 10.0 0.018 0.036 7.856 8.169 calc_potential_gpw 182 9.5 0.002 0.002 7.880 8.050 mp_sendrecv_dm3 2068 8.0 6.006 6.560 6.006 6.560 mp2_ri_gpw_compute_en_ener 172 7.0 6.357 6.433 6.357 6.433 mp_sync 37 10.5 3.363 5.590 3.363 5.590 pw_gather_s 912 13.2 4.908 5.321 4.908 5.321 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.649192, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 448.442368E+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 62385. MP_Allreduce 10249 272. MP_Sync 530 MP_Alltoall 2083 588782. 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 0.032 0.044 53.533 53.534 qs_mol_dyn_low 1 2.0 0.003 0.004 53.255 53.261 qs_forces 11 3.9 0.002 0.003 53.197 53.197 qs_energies 11 4.9 0.001 0.001 51.694 51.710 scf_env_do_scf 11 5.9 0.001 0.001 45.686 45.687 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.651 43.652 qs_scf_new_mos 108 7.5 0.000 0.001 33.355 33.622 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.354 33.621 dbcsr_multiply_generic 2286 12.5 0.101 0.116 32.941 33.300 ot_scf_mini 108 9.5 0.002 0.002 31.679 31.861 multiply_cannon 2286 13.5 0.187 0.192 25.706 27.151 multiply_cannon_loop 2286 14.5 1.481 1.569 24.999 26.468 velocity_verlet 10 3.0 0.001 0.001 25.750 25.751 ot_mini 108 10.5 0.001 0.001 18.907 19.140 qs_ot_get_derivative 108 11.5 0.001 0.001 16.008 16.173 mp_waitall_1 245248 16.5 8.217 14.401 8.217 14.401 multiply_cannon_metrocomm3 54864 15.5 0.068 0.073 5.807 12.508 multiply_cannon_multrec 54864 15.5 4.274 6.701 7.623 11.082 qs_ot_get_p 119 10.4 0.001 0.001 8.174 8.468 rebuild_ks_matrix 119 8.3 0.000 0.000 7.958 8.083 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.012 7.957 8.083 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.011 7.125 multiply_cannon_sync_h2d 54864 15.5 5.963 7.077 5.963 7.077 mp_sum_l 7207 12.9 5.204 6.806 5.204 6.806 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.515 5.941 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.422 5.477 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.246 5.364 init_scf_run 11 5.9 0.000 0.001 4.795 4.795 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.794 4.795 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.628 4.629 dbcsr_mm_accdrv_process 76910 16.1 1.139 1.817 3.270 4.603 sum_up_and_integrate 119 10.3 0.012 0.015 4.596 4.602 integrate_v_rspace 119 11.3 0.002 0.003 4.584 4.591 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.323 4.456 calculate_rho_elec 119 8.7 0.012 0.017 4.322 4.455 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.405 4.406 cp_fm_redistribute_end 50 14.0 2.249 4.376 2.256 4.380 cp_fm_diag_elpa_base 50 14.0 2.118 4.260 2.122 4.268 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.885 3.071 apply_single 119 13.6 0.000 0.000 2.885 3.071 calculate_dm_sparse 119 9.5 0.000 0.001 2.932 3.050 rs_pw_transfer 974 11.9 0.012 0.013 2.869 3.000 multiply_cannon_metrocomm1 54864 15.5 0.052 0.057 1.677 2.749 jit_kernel_multiply 13 15.8 2.070 2.694 2.070 2.694 ot_diis_step 108 11.5 0.006 0.006 2.629 2.630 calculate_first_density_matrix 1 7.0 0.000 0.000 2.551 2.556 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.426 2.538 density_rs2pw 119 9.7 0.004 0.004 2.322 2.449 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.371 2.373 acc_transpose_blocks 54864 15.5 0.227 0.248 1.737 2.211 wfi_extrapolate 11 7.9 0.001 0.001 2.182 2.182 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.035 2.107 grid_integrate_task_list 119 12.3 2.014 2.097 2.014 2.097 init_scf_loop 11 6.9 0.000 0.000 2.014 2.015 potential_pw2rs 119 12.3 0.004 0.004 1.947 1.975 pw_transfer 1439 11.6 0.052 0.057 1.791 1.875 mp_sum_d 4135 12.0 1.258 1.828 1.258 1.828 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.715 1.803 make_m2s 4572 13.5 0.054 0.056 1.712 1.762 make_images 4572 14.5 0.133 0.138 1.630 1.680 mp_alltoall_d11v 2130 13.8 1.322 1.505 1.322 1.505 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.422 1.455 fft3d_ps 1201 14.6 0.373 0.480 1.361 1.439 mp_waitany 12084 13.8 1.243 1.410 1.243 1.410 fft_wrap_pw1pw2_140 487 13.2 0.181 0.198 1.318 1.405 grid_collocate_task_list 119 9.7 1.291 1.362 1.291 1.362 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.534000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=427.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 483.205120E+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 62664. MP_Allreduce 10226 305. MP_Sync 54 MP_Alltoall 2060 519000. 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 0.014 0.031 39.411 39.412 qs_mol_dyn_low 1 2.0 0.003 0.004 39.187 39.194 qs_forces 11 3.9 0.002 0.003 39.129 39.130 qs_energies 11 4.9 0.001 0.001 37.425 37.429 scf_env_do_scf 11 5.9 0.001 0.001 32.330 32.331 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.824 29.824 dbcsr_multiply_generic 2286 12.5 0.099 0.102 21.397 21.755 qs_scf_new_mos 108 7.5 0.001 0.001 20.406 20.656 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.405 20.655 ot_scf_mini 108 9.5 0.002 0.003 19.514 19.686 velocity_verlet 10 3.0 0.001 0.001 18.633 18.634 multiply_cannon 2286 13.5 0.208 0.216 16.287 17.800 multiply_cannon_loop 2286 14.5 0.893 0.965 15.134 16.544 ot_mini 108 10.5 0.001 0.001 11.887 12.123 mp_waitall_1 200699 16.5 5.714 10.918 5.714 10.918 qs_ot_get_derivative 108 11.5 0.001 0.001 9.425 9.601 multiply_cannon_metrocomm3 27432 15.5 0.066 0.068 4.128 9.436 multiply_cannon_multrec 27432 15.5 1.979 4.517 5.858 8.673 rebuild_ks_matrix 119 8.3 0.000 0.000 7.320 7.458 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.319 7.458 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.473 6.598 dbcsr_mm_accdrv_process 47894 16.0 3.040 5.118 3.808 5.637 qs_ot_get_p 119 10.4 0.001 0.001 4.709 4.939 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.654 4.500 sum_up_and_integrate 119 10.3 0.024 0.027 4.314 4.319 integrate_v_rspace 119 11.3 0.002 0.003 4.289 4.298 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.045 4.116 apply_single 119 13.6 0.000 0.000 3.045 4.116 mp_sum_l 7207 12.9 2.103 4.091 2.103 4.091 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.947 3.983 calculate_rho_elec 119 8.7 0.021 0.024 3.946 3.982 init_scf_run 11 5.9 0.000 0.001 3.848 3.849 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.848 3.849 qs_ot_p2m_diag 50 11.0 0.009 0.014 3.207 3.227 rs_pw_transfer 974 11.9 0.010 0.011 2.791 3.158 multiply_cannon_sync_h2d 27432 15.5 2.187 2.862 2.187 2.862 make_m2s 4572 13.5 0.052 0.054 2.569 2.801 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.759 2.760 make_images 4572 14.5 0.200 0.236 2.481 2.712 density_rs2pw 119 9.7 0.004 0.004 2.272 2.695 init_scf_loop 11 6.9 0.000 0.000 2.480 2.481 ot_diis_step 108 11.5 0.011 0.011 2.413 2.414 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.378 2.379 cp_fm_redistribute_end 50 14.0 1.203 2.344 1.207 2.347 cp_fm_diag_elpa_base 50 14.0 1.106 2.243 1.135 2.290 calculate_first_density_matrix 1 7.0 0.000 0.000 2.283 2.285 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.175 2.266 calculate_dm_sparse 119 9.5 0.000 0.001 2.038 2.115 pw_transfer 1439 11.6 0.065 0.069 2.020 2.062 potential_pw2rs 119 12.3 0.006 0.006 2.026 2.040 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.928 1.972 grid_integrate_task_list 119 12.3 1.838 1.935 1.838 1.935 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.887 1.890 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.666 1.706 jit_kernel_multiply 9 16.1 0.716 1.680 0.716 1.680 prepare_preconditioner 11 7.9 0.000 0.000 1.555 1.581 make_preconditioner 11 8.9 0.000 0.000 1.555 1.581 make_images_data 4572 15.5 0.045 0.051 1.154 1.576 fft3d_ps 1201 14.6 0.521 0.576 1.481 1.519 fft_wrap_pw1pw2_140 487 13.2 0.201 0.210 1.477 1.518 wfi_extrapolate 11 7.9 0.001 0.001 1.513 1.513 acc_transpose_blocks 27432 15.5 0.109 0.114 1.189 1.513 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.454 1.513 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.005 1.458 mp_allgather_i34 2286 14.5 0.595 1.358 0.595 1.358 grid_collocate_task_list 119 9.7 1.229 1.348 1.229 1.348 mp_alltoall_d11v 2130 13.8 1.238 1.338 1.238 1.338 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.255 1.263 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.194 1.243 mp_sum_d 4135 12.0 0.589 1.040 0.589 1.040 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.970 0.971 rs_pw_transfer_RS2PW_140 130 11.5 0.140 0.149 0.560 0.938 mp_waitany 5720 13.7 0.532 0.925 0.532 0.925 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.906 0.920 acc_transpose_blocks_kernels 27432 16.5 0.180 0.269 0.654 0.881 make_images_sizes 4572 15.5 0.005 0.005 0.601 0.832 mp_alltoall_i44 4572 16.5 0.596 0.827 0.596 0.827 mp_alltoall_z22v 1201 16.6 0.729 0.810 0.729 0.810 rs_pw_transfer_PW2RS_50 119 14.3 0.590 0.609 0.747 0.801 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.412000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=459.636364, yerr=1.366663 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 515.948544E+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 63497. MP_Allreduce 10075 307. MP_Sync 54 MP_Alltoall 1821 2512315. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.044 31.843 31.844 qs_mol_dyn_low 1 2.0 0.003 0.007 31.559 31.566 qs_forces 11 3.9 0.003 0.008 31.501 31.501 qs_energies 11 4.9 0.001 0.005 29.938 29.942 scf_env_do_scf 11 5.9 0.001 0.003 25.266 25.266 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 22.775 22.775 dbcsr_multiply_generic 2286 12.5 0.095 0.100 16.171 16.266 velocity_verlet 10 3.0 0.001 0.001 15.010 15.011 qs_scf_new_mos 108 7.5 0.001 0.001 14.658 14.672 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.657 14.671 ot_scf_mini 108 9.5 0.002 0.003 13.942 13.952 multiply_cannon 2286 13.5 0.196 0.202 12.959 13.745 multiply_cannon_loop 2286 14.5 0.627 0.661 12.184 13.007 ot_mini 108 10.5 0.001 0.001 8.599 8.610 qs_ot_get_derivative 108 11.5 0.001 0.003 7.117 7.127 multiply_cannon_multrec 18288 15.5 1.926 2.899 6.731 7.025 rebuild_ks_matrix 119 8.3 0.000 0.000 6.336 6.351 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.336 6.351 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.597 5.611 dbcsr_mm_accdrv_process 38222 16.0 3.945 5.503 4.720 5.561 sum_up_and_integrate 119 10.3 0.031 0.032 3.934 3.940 integrate_v_rspace 119 11.3 0.003 0.003 3.903 3.913 mp_waitall_1 158411 16.6 2.513 3.677 2.513 3.677 init_scf_run 11 5.9 0.000 0.001 3.472 3.472 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.472 3.472 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.419 3.425 calculate_rho_elec 119 8.7 0.031 0.032 3.418 3.425 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.658 3.313 qs_ot_get_p 119 10.4 0.001 0.001 3.199 3.228 rs_pw_transfer 974 11.9 0.009 0.010 2.238 2.493 init_scf_loop 11 6.9 0.000 0.003 2.474 2.475 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.371 2.413 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.058 2.368 apply_single 119 13.6 0.000 0.000 2.058 2.368 calculate_first_density_matrix 1 7.0 0.000 0.002 2.257 2.258 density_rs2pw 119 9.7 0.004 0.006 1.938 2.186 qs_ot_p2m_diag 50 11.0 0.012 0.014 2.172 2.178 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.911 1.911 pw_transfer 1439 11.6 0.065 0.070 1.895 1.903 jit_kernel_multiply 10 16.2 0.724 1.892 0.724 1.892 grid_integrate_task_list 119 12.3 1.806 1.884 1.806 1.884 make_m2s 4572 13.5 0.044 0.045 1.727 1.881 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.802 1.813 calculate_dm_sparse 119 9.5 0.001 0.001 1.799 1.811 make_images 4572 14.5 0.190 0.201 1.642 1.794 potential_pw2rs 119 12.3 0.007 0.008 1.695 1.701 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.699 1.701 prepare_preconditioner 11 7.9 0.000 0.000 1.677 1.680 make_preconditioner 11 8.9 0.000 0.002 1.677 1.680 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.670 1.679 cp_fm_diag_elpa_base 50 14.0 1.647 1.658 1.668 1.678 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.534 1.616 multiply_cannon_sync_h2d 18288 15.5 1.417 1.587 1.417 1.587 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.555 1.563 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 1.518 1.522 mp_sum_l 7207 12.9 1.108 1.473 1.108 1.473 ot_diis_step 108 11.5 0.011 0.011 1.462 1.462 fft_wrap_pw1pw2_140 487 13.2 0.254 0.264 1.448 1.458 grid_collocate_task_list 119 9.7 1.213 1.369 1.213 1.369 fft3d_ps 1201 14.6 0.529 0.548 1.289 1.300 acc_transpose_blocks 18288 15.5 0.075 0.079 1.233 1.260 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.175 1.180 wfi_extrapolate 11 7.9 0.001 0.001 1.155 1.155 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.978 0.979 make_images_data 4572 15.5 0.045 0.049 0.774 0.932 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.861 0.884 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.653 0.862 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.810 0.811 acc_transpose_blocks_kernels 18288 16.5 0.208 0.217 0.792 0.810 mp_waitany 9880 13.7 0.541 0.806 0.541 0.806 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.123 0.527 0.780 mp_alltoall_d11v 2130 13.8 0.638 0.742 0.638 0.742 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.665 0.725 cp_fm_cholesky_invert 11 10.9 0.698 0.702 0.698 0.702 mp_alltoall_z22v 1201 16.6 0.613 0.683 0.613 0.683 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.844000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=491.818182, yerr=1.585054 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 554.835968E+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 63496. MP_Allreduce 10074 349. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.050 34.848 34.848 qs_mol_dyn_low 1 2.0 0.003 0.004 34.469 34.476 qs_forces 11 3.9 0.003 0.008 34.412 34.412 qs_energies 11 4.9 0.001 0.005 32.697 32.703 scf_env_do_scf 11 5.9 0.001 0.001 27.481 27.482 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.133 24.134 dbcsr_multiply_generic 2286 12.5 0.098 0.102 17.920 18.035 velocity_verlet 10 3.0 0.001 0.001 17.514 17.516 qs_scf_new_mos 108 7.5 0.001 0.001 15.803 15.860 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.803 15.859 ot_scf_mini 108 9.5 0.002 0.003 14.876 14.927 multiply_cannon 2286 13.5 0.238 0.270 14.496 14.883 multiply_cannon_loop 2286 14.5 0.931 0.966 13.574 13.885 multiply_cannon_multrec 27432 15.5 2.333 2.981 8.747 9.129 ot_mini 108 10.5 0.001 0.001 9.005 9.067 dbcsr_mm_accdrv_process 47916 15.9 5.343 7.193 6.320 7.496 qs_ot_get_derivative 108 11.5 0.001 0.001 7.217 7.266 rebuild_ks_matrix 119 8.3 0.000 0.000 6.499 6.551 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.498 6.550 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.767 5.813 init_scf_run 11 5.9 0.000 0.001 3.851 3.852 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.851 3.851 sum_up_and_integrate 119 10.3 0.035 0.038 3.753 3.761 integrate_v_rspace 119 11.3 0.003 0.003 3.717 3.726 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.475 3.504 calculate_rho_elec 119 8.7 0.040 0.046 3.475 3.503 qs_ot_get_p 119 10.4 0.001 0.002 3.295 3.365 init_scf_loop 11 6.9 0.001 0.005 3.329 3.329 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.717 3.114 prepare_preconditioner 11 7.9 0.000 0.000 2.496 2.504 make_preconditioner 11 8.9 0.000 0.001 2.496 2.504 calculate_first_density_matrix 1 7.0 0.000 0.003 2.500 2.501 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.106 2.434 mp_waitall_1 137007 16.6 1.661 2.227 1.661 2.227 make_m2s 4572 13.5 0.054 0.056 2.095 2.211 calculate_dm_sparse 119 9.5 0.000 0.001 2.147 2.202 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.018 2.150 apply_single 119 13.6 0.000 0.000 2.017 2.150 rs_pw_transfer 974 11.9 0.009 0.009 1.954 2.144 make_images 4572 14.5 0.271 0.333 1.989 2.103 density_rs2pw 119 9.7 0.004 0.004 1.892 2.085 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.062 2.070 pw_transfer 1439 11.6 0.065 0.069 1.967 2.004 jit_kernel_multiply 10 15.9 0.916 1.998 0.916 1.998 grid_integrate_task_list 119 12.3 1.809 1.922 1.809 1.922 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.875 1.916 qs_ot_get_derivative_diag 49 12.0 0.001 0.003 1.867 1.894 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.892 1.893 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.755 1.755 ot_diis_step 108 11.5 0.012 0.012 1.747 1.748 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.695 1.710 fft_wrap_pw1pw2_140 487 13.2 0.288 0.302 1.551 1.595 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.520 1.527 cp_fm_diag_elpa_base 50 14.0 1.486 1.501 1.518 1.524 potential_pw2rs 119 12.3 0.009 0.010 1.518 1.521 acc_transpose_blocks 27432 15.5 0.109 0.113 1.474 1.506 grid_collocate_task_list 119 9.7 1.222 1.355 1.222 1.355 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.339 1.349 fft3d_ps 1201 14.6 0.557 0.605 1.313 1.336 multiply_cannon_metrocomm3 27432 15.5 0.038 0.040 0.726 1.335 wfi_extrapolate 11 7.9 0.001 0.001 1.305 1.305 mp_sum_l 7207 12.9 0.911 1.201 0.911 1.201 cp_fm_upper_to_full 72 14.2 0.814 1.162 0.814 1.162 multiply_cannon_sync_h2d 27432 15.5 1.007 1.108 1.007 1.108 qs_energies_init_hamiltonians 11 5.9 0.001 0.008 1.104 1.105 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.078 1.100 dbcsr_complete_redistribute 329 12.2 0.131 0.150 0.772 1.044 make_images_data 4572 15.5 0.044 0.049 0.813 0.932 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.872 hybrid_alltoall_any 4725 16.4 0.062 0.153 0.695 0.863 acc_transpose_blocks_kernels 27432 16.5 0.266 0.274 0.849 0.862 copy_fm_to_dbcsr 176 11.2 0.001 0.002 0.582 0.848 mp_alltoall_d11v 2130 13.8 0.698 0.845 0.698 0.845 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.821 0.826 cp_fm_cholesky_invert 11 10.9 0.722 0.725 0.722 0.725 mp_alltoall_i22 627 13.8 0.426 0.724 0.426 0.724 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=34.848000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=523.636364, yerr=4.578065 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 599.711744E+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 63495. MP_Allreduce 10074 348. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.029 27.317 27.317 qs_mol_dyn_low 1 2.0 0.003 0.014 27.155 27.161 qs_forces 11 3.9 0.006 0.006 27.097 27.098 qs_energies 11 4.9 0.001 0.001 25.393 25.395 scf_env_do_scf 11 5.9 0.001 0.001 20.801 20.801 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 18.392 18.393 velocity_verlet 10 3.0 0.001 0.001 14.106 14.109 dbcsr_multiply_generic 2286 12.5 0.093 0.100 11.755 11.828 qs_scf_new_mos 108 7.5 0.001 0.001 10.499 10.526 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.498 10.525 ot_scf_mini 108 9.5 0.002 0.002 9.838 9.867 multiply_cannon 2286 13.5 0.230 0.235 9.346 9.770 multiply_cannon_loop 2286 14.5 0.327 0.337 8.448 8.618 rebuild_ks_matrix 119 8.3 0.000 0.000 5.903 5.924 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.903 5.923 multiply_cannon_multrec 9144 15.5 1.583 1.847 5.708 5.914 ot_mini 108 10.5 0.001 0.001 5.451 5.487 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.258 5.277 qs_ot_get_derivative 108 11.5 0.001 0.002 4.168 4.196 dbcsr_mm_accdrv_process 12550 15.8 2.917 3.723 4.024 4.117 sum_up_and_integrate 119 10.3 0.037 0.041 3.616 3.620 integrate_v_rspace 119 11.3 0.003 0.003 3.579 3.584 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.475 3.479 calculate_rho_elec 119 8.7 0.060 0.061 3.474 3.479 init_scf_run 11 5.9 0.000 0.001 3.172 3.172 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.171 3.171 qs_ot_get_p 119 10.4 0.001 0.001 2.642 2.677 init_scf_loop 11 6.9 0.000 0.000 2.389 2.391 calculate_first_density_matrix 1 7.0 0.000 0.000 2.068 2.068 pw_transfer 1439 11.6 0.066 0.069 2.004 2.014 density_rs2pw 119 9.7 0.004 0.004 1.789 1.934 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.910 1.921 grid_integrate_task_list 119 12.3 1.848 1.918 1.848 1.918 make_m2s 4572 13.5 0.035 0.036 1.671 1.840 mp_waitall_1 115863 16.7 1.243 1.765 1.243 1.765 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.761 1.763 make_images 4572 14.5 0.268 0.301 1.582 1.749 rs_pw_transfer 974 11.9 0.008 0.008 1.617 1.748 jit_kernel_multiply 10 15.8 1.070 1.735 1.070 1.735 calculate_dm_sparse 119 9.5 0.000 0.000 1.681 1.699 prepare_preconditioner 11 7.9 0.000 0.000 1.657 1.661 make_preconditioner 11 8.9 0.000 0.000 1.657 1.661 fft_wrap_pw1pw2_140 487 13.2 0.364 0.372 1.578 1.589 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.550 1.576 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.570 1.571 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.512 1.513 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.460 1.480 grid_collocate_task_list 119 9.7 1.271 1.391 1.271 1.391 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.364 1.374 potential_pw2rs 119 12.3 0.010 0.011 1.352 1.354 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.284 1.292 cp_fm_diag_elpa_base 50 14.0 1.258 1.273 1.283 1.291 ot_diis_step 108 11.5 0.012 0.013 1.272 1.273 fft3d_ps 1201 14.6 0.558 0.572 1.244 1.254 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.214 1.215 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.152 1.175 apply_single 119 13.6 0.000 0.000 1.152 1.175 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.112 1.117 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.083 1.096 wfi_extrapolate 11 7.9 0.001 0.001 1.057 1.057 hybrid_alltoall_any 4725 16.4 0.062 0.174 0.752 0.995 make_images_data 4572 15.5 0.038 0.042 0.768 0.952 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.867 0.921 acc_transpose_blocks 9144 15.5 0.038 0.039 0.897 0.907 mp_alltoall_d11v 2130 13.8 0.761 0.866 0.761 0.866 cp_fm_cholesky_invert 11 10.9 0.825 0.827 0.825 0.827 multiply_cannon_sync_h2d 9144 15.5 0.708 0.798 0.708 0.798 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.765 0.767 qs_env_update_s_mstruct 11 6.9 0.001 0.003 0.667 0.719 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.695 0.704 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 0.656 0.661 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.305 0.629 mp_allgather_i34 2286 14.5 0.230 0.621 0.230 0.621 mp_alltoall_z22v 1201 16.6 0.560 0.592 0.560 0.592 mp_waitany 5200 13.7 0.451 0.578 0.451 0.578 qs_create_task_list 11 7.9 0.000 0.000 0.540 0.568 generate_qs_task_list 11 8.9 0.188 0.210 0.540 0.567 jit_kernel_transpose 5 15.6 0.540 0.546 0.540 0.546 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.317000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=566.000000, yerr=8.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 749.993984E+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 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.033 41.965 41.965 qs_mol_dyn_low 1 2.0 0.003 0.004 41.670 41.676 qs_forces 11 3.9 0.002 0.002 41.594 41.595 qs_energies 11 4.9 0.001 0.001 39.596 39.600 scf_env_do_scf 11 5.9 0.001 0.001 33.979 33.980 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.976 25.977 velocity_verlet 10 3.0 0.001 0.001 23.744 23.752 dbcsr_multiply_generic 2286 12.5 0.099 0.102 17.389 17.623 qs_scf_new_mos 108 7.5 0.001 0.001 16.125 16.222 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.124 16.221 ot_scf_mini 108 9.5 0.002 0.002 15.036 15.135 multiply_cannon 2286 13.5 0.303 0.310 13.436 14.388 multiply_cannon_loop 2286 14.5 0.342 0.348 12.157 13.100 ot_mini 108 10.5 0.001 0.001 8.862 8.978 multiply_cannon_multrec 9144 15.5 3.392 4.698 8.502 8.626 init_scf_loop 11 6.9 0.000 0.000 7.974 7.977 rebuild_ks_matrix 119 8.3 0.000 0.000 7.326 7.464 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.326 7.464 prepare_preconditioner 11 7.9 0.000 0.000 7.009 7.022 make_preconditioner 11 8.9 0.000 0.000 7.009 7.022 qs_ot_get_derivative 108 11.5 0.001 0.001 6.824 6.923 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.521 6.894 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.620 6.744 dbcsr_mm_accdrv_process 12550 15.8 4.094 5.559 4.985 6.327 cp_fm_upper_to_full 72 14.2 3.218 4.644 3.218 4.644 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.307 4.313 calculate_rho_elec 119 8.7 0.118 0.121 4.306 4.312 sum_up_and_integrate 119 10.3 0.065 0.067 4.040 4.047 integrate_v_rspace 119 11.3 0.003 0.004 3.975 3.982 init_scf_run 11 5.9 0.000 0.001 3.578 3.578 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.578 3.578 qs_ot_get_p 119 10.4 0.001 0.001 3.424 3.558 mp_waitall_1 94719 16.7 2.355 3.336 2.355 3.336 dbcsr_complete_redistribute 329 12.2 0.289 0.294 2.067 2.933 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.501 2.909 pw_transfer 1439 11.6 0.069 0.070 2.734 2.738 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.635 2.639 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.723 2.576 make_m2s 4572 13.5 0.038 0.038 2.373 2.548 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.184 2.456 apply_single 119 13.6 0.000 0.000 2.184 2.455 make_images 4572 14.5 0.353 0.383 2.252 2.427 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.483 2.328 mp_alltoall_i22 627 13.8 1.454 2.279 1.454 2.279 density_rs2pw 119 9.7 0.004 0.004 2.232 2.248 fft_wrap_pw1pw2_140 487 13.2 0.616 0.623 2.199 2.205 calculate_dm_sparse 119 9.5 0.000 0.000 2.162 2.180 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.220 2.143 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.102 2.103 calculate_first_density_matrix 1 7.0 0.000 0.000 2.088 2.088 grid_integrate_task_list 119 12.3 2.057 2.073 2.057 2.073 ot_diis_step 108 11.5 0.014 0.015 2.012 2.012 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.820 1.820 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.793 1.793 mp_sum_l 7207 12.9 1.036 1.774 1.036 1.774 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.713 1.762 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.760 1.760 fft3d_ps 1201 14.6 0.597 0.606 1.652 1.656 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.527 1.567 rs_pw_transfer 974 11.9 0.009 0.009 1.501 1.524 cp_fm_cholesky_invert 11 10.9 1.515 1.519 1.515 1.519 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.517 1.517 cp_fm_diag_elpa_base 50 14.0 1.367 1.422 1.515 1.515 grid_collocate_task_list 119 9.7 1.489 1.499 1.489 1.499 hybrid_alltoall_any 4725 16.4 0.088 0.149 1.196 1.453 potential_pw2rs 119 12.3 0.014 0.015 1.437 1.441 wfi_extrapolate 11 7.9 0.001 0.001 1.427 1.427 make_images_data 4572 15.5 0.042 0.045 1.151 1.385 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.223 1.231 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.148 1.168 mp_alltoall_d11v 2130 13.8 1.107 1.134 1.107 1.134 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.096 1.114 multiply_cannon_sync_h2d 9144 15.5 1.042 1.045 1.042 1.045 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.992 1.028 jit_kernel_multiply 6 15.7 0.864 1.017 0.864 1.017 qs_create_task_list 11 7.9 0.000 0.000 0.942 0.953 generate_qs_task_list 11 8.9 0.369 0.387 0.942 0.953 mp_alltoall_z22v 1201 16.6 0.920 0.934 0.920 0.934 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.918 0.931 acc_transpose_blocks 9144 15.5 0.038 0.039 0.901 0.909 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.965000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=703.454545, yerr=13.193412 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 499.707904E+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 66218. MP_Allreduce 9696 492. MP_Sync 52 MP_Alltoall 1938 527415. 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 0.011 0.028 82.148 82.148 qs_mol_dyn_low 1 2.0 0.003 0.004 81.851 81.860 qs_forces 11 3.9 0.003 0.003 81.786 81.787 qs_energies 11 4.9 0.001 0.001 78.901 78.930 scf_env_do_scf 11 5.9 0.001 0.001 70.075 70.078 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 64.541 64.541 dbcsr_multiply_generic 2055 12.4 0.105 0.107 50.119 50.341 qs_scf_new_mos 99 7.5 0.000 0.001 46.913 47.066 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.912 47.065 ot_scf_mini 99 9.5 0.002 0.002 44.530 44.659 velocity_verlet 10 3.0 0.001 0.002 42.659 42.660 multiply_cannon 2055 13.4 0.183 0.189 41.877 42.497 multiply_cannon_loop 2055 14.4 1.516 1.549 40.903 41.539 ot_mini 99 10.5 0.001 0.001 26.023 26.143 qs_ot_get_derivative 99 11.5 0.001 0.001 19.323 19.459 multiply_cannon_multrec 49320 15.4 12.486 13.182 17.319 18.233 rebuild_ks_matrix 110 8.3 0.000 0.001 14.723 14.845 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.723 14.845 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.921 13.025 mp_waitall_1 220248 16.4 10.480 11.433 10.480 11.433 multiply_cannon_sync_h2d 49320 15.4 10.349 10.867 10.349 10.867 qs_ot_get_p 110 10.4 0.001 0.001 10.210 10.375 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.304 7.832 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.148 7.727 apply_single 110 13.6 0.000 0.000 7.148 7.727 sum_up_and_integrate 110 10.3 0.036 0.043 7.261 7.276 multiply_cannon_metrocomm3 49320 15.4 0.079 0.082 6.248 7.262 integrate_v_rspace 110 11.3 0.003 0.004 7.225 7.248 qs_ot_p2m_diag 48 11.0 0.012 0.019 7.011 7.052 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.722 6.862 calculate_rho_elec 110 8.6 0.021 0.026 6.722 6.862 init_scf_run 11 5.9 0.000 0.001 6.736 6.736 scf_env_initial_rho_setup 11 6.9 0.005 0.006 6.736 6.736 ot_diis_step 99 11.5 0.006 0.006 6.520 6.520 cp_dbcsr_syevd 48 12.0 0.002 0.003 6.175 6.176 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.534 5.571 cp_fm_diag_elpa_base 48 14.0 5.519 5.557 5.532 5.569 init_scf_loop 11 6.9 0.000 0.000 5.506 5.507 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.345 5.400 dbcsr_mm_accdrv_process 87628 16.1 1.894 1.961 4.708 4.961 mp_sum_l 6514 12.8 3.893 4.776 3.893 4.776 rs_pw_transfer 902 11.9 0.012 0.013 3.940 4.643 density_rs2pw 110 9.6 0.004 0.005 3.542 4.294 wfi_extrapolate 11 7.9 0.001 0.001 4.052 4.052 make_m2s 4110 13.4 0.061 0.066 3.899 3.996 make_images 4110 14.4 0.177 0.190 3.803 3.904 calculate_dm_sparse 110 9.5 0.001 0.001 3.745 3.837 multiply_cannon_metrocomm1 49320 15.4 0.060 0.063 2.286 3.586 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.536 3.540 grid_integrate_task_list 110 12.3 3.243 3.415 3.243 3.415 prepare_preconditioner 11 7.9 0.000 0.000 3.311 3.337 make_preconditioner 11 8.9 0.000 0.000 3.311 3.337 pw_transfer 1331 11.6 0.056 0.066 3.220 3.307 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.193 3.235 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.131 3.218 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.095 3.162 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.079 3.133 potential_pw2rs 110 12.3 0.006 0.007 2.782 2.808 fft_wrap_pw1pw2_140 451 13.1 0.449 0.494 2.632 2.725 mp_waitany 14300 13.8 1.841 2.696 1.841 2.696 calculate_first_density_matrix 1 7.0 0.000 0.000 2.594 2.600 jit_kernel_multiply 13 15.9 2.538 2.552 2.538 2.552 mp_alltoall_d11v 2046 13.8 2.070 2.493 2.070 2.493 fft3d_ps 1111 14.6 0.792 0.873 2.327 2.397 grid_collocate_task_list 110 9.6 2.085 2.276 2.085 2.276 acc_transpose_blocks 49320 15.4 0.209 0.219 2.079 2.138 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.932 1.953 mp_sum_d 3889 11.9 1.349 1.914 1.349 1.914 make_images_data 4110 15.4 0.042 0.045 1.757 1.867 cp_fm_cholesky_invert 11 10.9 1.842 1.846 1.842 1.846 hybrid_alltoall_any 4261 16.3 0.082 0.483 1.533 1.801 rs_pw_transfer_RS2PW_140 121 11.5 0.168 0.177 0.936 1.674 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.644 1.673 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.148000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=473.909091, yerr=3.260077 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 584.966144E+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 66437. MP_Allreduce 9695 570. MP_Sync 52 MP_Alltoall 1717 2755812. 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 0.014 0.030 69.923 69.924 qs_mol_dyn_low 1 2.0 0.003 0.004 69.693 69.702 qs_forces 11 3.9 0.004 0.009 69.629 69.631 qs_energies 11 4.9 0.001 0.005 66.269 66.275 scf_env_do_scf 11 5.9 0.001 0.001 57.699 57.702 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 50.006 50.006 dbcsr_multiply_generic 2055 12.4 0.114 0.117 37.619 37.832 velocity_verlet 10 3.0 0.001 0.001 36.555 36.556 qs_scf_new_mos 99 7.5 0.001 0.001 33.366 33.526 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.366 33.525 multiply_cannon 2055 13.4 0.222 0.247 31.004 32.199 ot_scf_mini 99 9.5 0.003 0.003 31.705 31.848 multiply_cannon_loop 2055 14.4 0.914 0.941 29.744 30.677 ot_mini 99 10.5 0.001 0.001 18.253 18.402 multiply_cannon_multrec 24660 15.4 7.691 9.367 13.866 15.302 rebuild_ks_matrix 110 8.3 0.000 0.000 13.963 14.110 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.963 14.110 qs_ot_get_derivative 99 11.5 0.001 0.001 12.397 12.537 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.291 12.423 mp_waitall_1 176588 16.5 7.687 10.475 7.687 10.475 multiply_cannon_metrocomm3 24660 15.4 0.067 0.072 5.156 8.300 multiply_cannon_sync_h2d 24660 15.4 7.087 8.226 7.087 8.226 init_scf_loop 11 6.9 0.000 0.000 7.655 7.656 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.544 7.318 apply_single 110 13.6 0.000 0.001 6.543 7.317 qs_ot_get_p 110 10.4 0.001 0.001 6.820 7.010 sum_up_and_integrate 110 10.3 0.052 0.059 6.791 6.801 integrate_v_rspace 110 11.3 0.003 0.003 6.738 6.752 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.356 6.365 calculate_rho_elec 110 8.6 0.040 0.047 6.355 6.365 dbcsr_mm_accdrv_process 52282 16.1 4.584 5.443 6.014 6.342 init_scf_run 11 5.9 0.000 0.001 6.207 6.208 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.207 6.207 ot_diis_step 99 11.5 0.010 0.010 5.807 5.808 prepare_preconditioner 11 7.9 0.000 0.000 5.599 5.623 make_preconditioner 11 8.9 0.000 0.000 5.599 5.623 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.678 5.375 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.176 5.338 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.896 4.915 make_m2s 4110 13.4 0.057 0.060 4.209 4.636 make_images 4110 14.4 0.399 0.444 4.100 4.523 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.453 4.453 pw_transfer 1331 11.6 0.067 0.073 3.773 3.916 density_rs2pw 110 9.6 0.004 0.005 3.439 3.902 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.838 3.856 cp_fm_diag_elpa_base 48 14.0 3.790 3.806 3.835 3.853 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.666 3.809 rs_pw_transfer 902 11.9 0.012 0.014 3.136 3.635 wfi_extrapolate 11 7.9 0.001 0.001 3.561 3.561 grid_integrate_task_list 110 12.3 3.135 3.348 3.135 3.348 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.194 3.269 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.254 3.255 fft_wrap_pw1pw2_140 451 13.1 0.520 0.534 3.091 3.234 calculate_dm_sparse 110 9.5 0.001 0.001 2.957 2.993 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.904 2.956 fft3d_ps 1111 14.6 1.106 1.328 2.640 2.795 make_images_data 4110 15.4 0.046 0.051 2.351 2.775 cp_fm_cholesky_invert 11 10.9 2.676 2.685 2.676 2.685 hybrid_alltoall_any 4261 16.3 0.102 0.449 2.011 2.641 potential_pw2rs 110 12.3 0.008 0.009 2.566 2.585 calculate_first_density_matrix 1 7.0 0.000 0.001 2.552 2.555 grid_collocate_task_list 110 9.6 2.056 2.475 2.056 2.475 mp_sum_l 6514 12.8 1.782 2.460 1.782 2.460 mp_alltoall_d11v 2046 13.8 1.808 2.176 1.808 2.176 jit_kernel_multiply 11 16.2 1.078 2.115 1.078 2.115 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.967 1.988 qs_energies_init_hamiltonians 11 5.9 0.001 0.006 1.850 1.851 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.828 1.845 mp_waitany 10164 13.8 1.232 1.751 1.232 1.751 multiply_cannon_metrocomm4 22605 15.4 0.077 0.082 0.764 1.697 mp_allgather_i34 2055 14.4 0.586 1.595 0.586 1.595 acc_transpose_blocks 24660 15.4 0.111 0.118 1.533 1.558 mp_irecv_dv 57340 16.2 0.638 1.544 0.638 1.544 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.528 1.538 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.367 1.472 rs_pw_transfer_RS2PW_140 121 11.5 0.208 0.220 0.972 1.455 dbcsr_complete_redistribute 325 12.2 0.236 0.308 1.147 1.412 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.924000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=552.090909, yerr=7.279542 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 655.028224E+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 66428. MP_Allreduce 9694 567. 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 0.017 0.033 59.855 59.856 qs_mol_dyn_low 1 2.0 0.003 0.004 59.474 59.483 qs_forces 11 3.9 0.003 0.004 59.403 59.404 qs_energies 11 4.9 0.004 0.023 56.203 56.207 scf_env_do_scf 11 5.9 0.004 0.014 48.291 48.291 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 39.801 39.803 velocity_verlet 10 3.0 0.001 0.001 32.407 32.409 dbcsr_multiply_generic 2055 12.4 0.106 0.109 28.292 28.543 qs_scf_new_mos 99 7.5 0.001 0.001 24.697 24.805 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.696 24.805 ot_scf_mini 99 9.5 0.002 0.003 23.477 23.606 multiply_cannon 2055 13.4 0.220 0.228 21.978 23.210 multiply_cannon_loop 2055 14.4 0.612 0.627 20.766 21.853 ot_mini 99 10.5 0.001 0.001 13.389 13.519 rebuild_ks_matrix 110 8.3 0.000 0.000 12.407 12.564 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.407 12.563 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.921 11.065 multiply_cannon_multrec 16440 15.4 3.894 5.080 9.598 10.722 mp_waitall_1 139946 16.5 7.044 9.766 7.044 9.766 qs_ot_get_derivative 99 11.5 0.001 0.001 8.885 9.012 init_scf_loop 11 6.9 0.005 0.037 8.453 8.456 multiply_cannon_metrocomm3 16440 15.4 0.043 0.047 4.316 7.009 prepare_preconditioner 11 7.9 0.000 0.000 6.615 6.633 make_preconditioner 11 8.9 0.000 0.001 6.615 6.633 sum_up_and_integrate 110 10.3 0.059 0.060 6.553 6.568 integrate_v_rspace 110 11.3 0.003 0.003 6.493 6.508 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.944 6.288 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.007 6.016 calculate_rho_elec 110 8.6 0.059 0.059 6.006 6.016 dbcsr_mm_accdrv_process 34862 16.1 4.656 5.180 5.557 5.685 init_scf_run 11 5.9 0.000 0.001 5.489 5.491 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.489 5.491 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.036 5.488 apply_single 110 13.6 0.000 0.000 5.036 5.488 qs_ot_get_p 110 10.4 0.001 0.001 5.329 5.458 make_m2s 4110 13.4 0.050 0.051 4.250 4.640 make_images 4110 14.4 0.390 0.507 4.136 4.528 ot_diis_step 99 11.5 0.011 0.011 4.477 4.477 density_rs2pw 110 9.6 0.004 0.005 3.118 4.410 multiply_cannon_sync_h2d 16440 15.4 3.665 4.250 3.665 4.250 rs_pw_transfer 902 11.9 0.011 0.012 2.691 3.974 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.718 3.721 pw_transfer 1331 11.6 0.066 0.074 3.676 3.682 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.027 3.656 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.569 3.578 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.382 3.382 grid_integrate_task_list 110 12.3 3.164 3.341 3.164 3.341 fft_wrap_pw1pw2_140 451 13.1 0.639 0.654 3.086 3.098 make_images_data 4110 15.4 0.043 0.047 2.527 3.042 wfi_extrapolate 11 7.9 0.001 0.001 2.956 2.956 hybrid_alltoall_any 4261 16.3 0.105 0.374 2.247 2.908 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.853 2.862 cp_fm_diag_elpa_base 48 14.0 2.788 2.820 2.851 2.860 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.822 2.824 cp_fm_cholesky_invert 11 10.9 2.589 2.595 2.589 2.595 mp_waitany 17072 13.8 1.225 2.575 1.225 2.575 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.517 2.563 calculate_dm_sparse 110 9.5 0.001 0.001 2.490 2.523 multiply_cannon_metrocomm4 14385 15.4 0.045 0.050 0.882 2.483 grid_collocate_task_list 110 9.6 2.084 2.470 2.084 2.470 calculate_first_density_matrix 1 7.0 0.000 0.001 2.444 2.447 fft3d_ps 1111 14.6 1.091 1.103 2.390 2.402 mp_irecv_dv 48980 15.7 0.812 2.356 0.812 2.356 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.281 2.343 potential_pw2rs 110 12.3 0.011 0.011 2.278 2.283 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.184 0.908 2.189 mp_alltoall_d11v 2046 13.8 1.752 2.137 1.752 2.137 mp_sum_l 6514 12.8 1.454 2.034 1.454 2.034 qs_energies_init_hamiltonians 11 5.9 0.002 0.007 1.992 1.993 dbcsr_complete_redistribute 325 12.2 0.323 0.348 1.438 1.868 mp_allgather_i34 2055 14.4 0.529 1.753 0.529 1.753 cp_fm_upper_to_full 70 14.2 1.388 1.750 1.388 1.750 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.680 1.694 cp_fm_cholesky_decompose 22 10.9 1.509 1.531 1.509 1.531 jit_kernel_multiply 8 16.6 0.512 1.471 0.512 1.471 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.452 1.464 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.378 1.463 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.975 1.412 rs_gather_matrices 110 12.3 0.232 0.261 0.973 1.319 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.245 1.250 acc_transpose_blocks 16440 15.4 0.074 0.077 1.212 1.224 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.856000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=620.818182, yerr=8.054566 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 720.359424E+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 66426. MP_Allreduce 9694 608. 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 0.031 64.971 64.971 qs_mol_dyn_low 1 2.0 0.003 0.004 64.750 64.758 qs_forces 11 3.9 0.003 0.003 64.687 64.687 qs_energies 11 4.9 0.001 0.001 61.271 61.274 scf_env_do_scf 11 5.9 0.001 0.001 52.768 52.771 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.238 41.239 velocity_verlet 10 3.0 0.001 0.001 36.748 36.750 dbcsr_multiply_generic 2055 12.4 0.112 0.119 29.284 29.517 qs_scf_new_mos 99 7.5 0.001 0.001 26.209 26.322 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.208 26.322 ot_scf_mini 99 9.5 0.003 0.003 24.599 24.717 multiply_cannon 2055 13.4 0.245 0.270 22.498 23.709 multiply_cannon_loop 2055 14.4 0.879 0.893 21.082 21.665 ot_mini 99 10.5 0.001 0.001 14.092 14.233 multiply_cannon_multrec 24660 15.4 4.230 6.783 12.625 13.871 rebuild_ks_matrix 110 8.3 0.000 0.000 12.203 12.304 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.203 12.304 init_scf_loop 11 6.9 0.000 0.000 11.491 11.492 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.771 10.860 qs_ot_get_derivative 99 11.5 0.002 0.007 9.935 10.059 prepare_preconditioner 11 7.9 0.000 0.000 9.754 9.770 make_preconditioner 11 8.9 0.000 0.000 9.754 9.770 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.992 9.445 dbcsr_mm_accdrv_process 52304 16.0 6.789 8.153 8.247 9.168 sum_up_and_integrate 110 10.3 0.068 0.071 6.507 6.523 integrate_v_rspace 110 11.3 0.003 0.003 6.440 6.453 mp_waitall_1 121746 16.5 4.305 6.234 4.305 6.234 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.086 6.093 calculate_rho_elec 110 8.6 0.078 0.082 6.085 6.093 init_scf_run 11 5.9 0.000 0.001 5.786 5.786 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.786 5.786 qs_ot_get_p 110 10.4 0.001 0.001 5.544 5.717 make_m2s 4110 13.4 0.060 0.062 5.261 5.622 make_images 4110 14.4 0.576 0.701 5.121 5.480 cp_fm_upper_to_full 70 14.2 3.332 4.834 3.332 4.834 ot_diis_step 99 11.5 0.011 0.011 4.120 4.120 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.027 4.109 apply_single 110 13.6 0.000 0.000 4.026 4.109 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.813 3.828 pw_transfer 1331 11.6 0.066 0.075 3.741 3.772 dbcsr_complete_redistribute 325 12.2 0.415 0.454 2.638 3.769 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.635 3.669 density_rs2pw 110 9.6 0.004 0.004 3.015 3.526 grid_integrate_task_list 110 12.3 3.276 3.417 3.276 3.417 multiply_cannon_sync_h2d 24660 15.4 3.207 3.346 3.207 3.346 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.342 3.342 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.230 3.286 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.098 3.207 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.551 3.203 fft_wrap_pw1pw2_140 451 13.1 0.667 0.690 3.123 3.161 hybrid_alltoall_any 4261 16.3 0.120 0.458 2.277 3.041 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.963 3.030 make_images_data 4110 15.4 0.045 0.049 2.656 3.006 wfi_extrapolate 11 7.9 0.001 0.001 2.994 2.994 calculate_dm_sparse 110 9.5 0.001 0.001 2.893 2.925 rs_pw_transfer 902 11.9 0.010 0.011 2.348 2.866 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.754 2.847 mp_alltoall_i22 605 13.7 1.641 2.815 1.641 2.815 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.806 2.815 cp_fm_diag_elpa_base 48 14.0 2.660 2.717 2.803 2.813 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.769 2.772 calculate_first_density_matrix 1 7.0 0.000 0.000 2.700 2.704 cp_fm_cholesky_invert 11 10.9 2.612 2.620 2.612 2.620 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.420 2.455 grid_collocate_task_list 110 9.6 2.180 2.439 2.180 2.439 fft3d_ps 1111 14.6 1.084 1.119 2.396 2.412 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.234 2.234 jit_kernel_multiply 12 15.9 1.128 2.184 1.128 2.184 potential_pw2rs 110 12.3 0.013 0.013 2.154 2.162 mp_alltoall_d11v 2046 13.8 1.720 2.074 1.720 2.074 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.678 1.717 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.598 1.699 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.673 1.684 cp_fm_cholesky_decompose 22 10.9 1.613 1.662 1.613 1.662 mp_allgather_i34 2055 14.4 0.509 1.602 0.509 1.602 mp_waitany 13376 13.8 1.040 1.535 1.040 1.535 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.519 1.534 acc_transpose_blocks 24660 15.4 0.103 0.105 1.496 1.513 mp_sum_l 6514 12.8 0.851 1.463 0.851 1.463 multiply_cannon_metrocomm4 20550 15.4 0.059 0.063 0.849 1.458 mp_irecv_dv 62702 16.1 0.750 1.383 0.750 1.383 rs_pw_transfer_RS2PW_140 121 11.5 0.168 0.177 0.794 1.317 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=64.971000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=682.090909, yerr=6.359739 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 847.015936E+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 66424. MP_Allreduce 9694 649. 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 0.030 54.855 54.855 qs_mol_dyn_low 1 2.0 0.003 0.004 54.626 54.635 qs_forces 11 3.9 0.003 0.003 54.562 54.563 qs_energies 11 4.9 0.001 0.001 50.870 50.875 scf_env_do_scf 11 5.9 0.000 0.001 42.439 42.439 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 34.826 34.827 velocity_verlet 10 3.0 0.001 0.001 30.993 30.996 dbcsr_multiply_generic 2055 12.4 0.104 0.109 22.472 22.587 qs_scf_new_mos 99 7.5 0.001 0.001 19.806 19.861 qs_scf_loop_do_ot 99 8.5 0.001 0.001 19.806 19.861 ot_scf_mini 99 9.5 0.002 0.002 18.579 18.627 multiply_cannon 2055 13.4 0.246 0.256 17.197 18.262 multiply_cannon_loop 2055 14.4 0.318 0.332 15.904 16.093 rebuild_ks_matrix 110 8.3 0.000 0.000 11.903 11.929 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.903 11.929 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.574 10.599 ot_mini 99 10.5 0.001 0.001 10.167 10.211 multiply_cannon_multrec 8220 15.4 3.200 4.511 7.387 8.566 mp_waitall_1 103326 16.6 5.980 7.614 5.980 7.614 init_scf_loop 11 6.9 0.000 0.000 7.567 7.568 sum_up_and_integrate 110 10.3 0.079 0.080 6.521 6.531 integrate_v_rspace 110 11.3 0.003 0.003 6.441 6.451 qs_ot_get_derivative 99 11.5 0.001 0.001 6.376 6.423 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.324 6.336 calculate_rho_elec 110 8.6 0.114 0.115 6.324 6.335 prepare_preconditioner 11 7.9 0.000 0.000 5.916 5.926 make_preconditioner 11 8.9 0.000 0.000 5.916 5.926 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.485 5.554 init_scf_run 11 5.9 0.000 0.001 5.303 5.303 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.302 5.302 dbcsr_mm_accdrv_process 17442 15.9 2.751 3.586 4.057 4.940 qs_ot_get_p 110 10.4 0.001 0.001 4.508 4.562 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.139 4.560 make_m2s 4110 13.4 0.039 0.040 4.072 4.306 make_images 4110 14.4 0.638 0.692 3.943 4.179 pw_transfer 1331 11.6 0.066 0.073 3.932 3.948 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.824 3.844 ot_diis_step 99 11.5 0.012 0.012 3.771 3.771 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.695 3.738 apply_single 110 13.6 0.000 0.000 3.695 3.738 grid_integrate_task_list 110 12.3 3.369 3.508 3.369 3.508 density_rs2pw 110 9.6 0.004 0.005 3.067 3.440 fft_wrap_pw1pw2_140 451 13.1 0.834 0.844 3.309 3.336 qs_ot_p2m_diag 48 11.0 0.081 0.085 3.209 3.213 multiply_cannon_sync_h2d 8220 15.4 2.901 3.028 2.901 3.028 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.914 2.914 cp_fm_cholesky_invert 11 10.9 2.839 2.843 2.839 2.843 wfi_extrapolate 11 7.9 0.001 0.001 2.801 2.801 make_images_data 4110 15.4 0.038 0.043 2.309 2.704 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.678 2.679 rs_pw_transfer 902 11.9 0.010 0.011 2.228 2.626 hybrid_alltoall_any 4261 16.3 0.200 0.863 2.251 2.622 grid_collocate_task_list 110 9.6 2.277 2.539 2.277 2.539 calculate_dm_sparse 110 9.5 0.001 0.001 2.443 2.487 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.480 2.482 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.409 2.417 cp_fm_diag_elpa_base 48 14.0 2.355 2.376 2.408 2.415 calculate_first_density_matrix 1 7.0 0.000 0.000 2.406 2.408 fft3d_ps 1111 14.6 1.140 1.188 2.368 2.394 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.142 2.153 potential_pw2rs 110 12.3 0.016 0.016 2.068 2.072 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.774 2.003 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.842 1.870 mp_alltoall_d11v 2046 13.8 1.609 1.764 1.609 1.764 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.741 1.763 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.659 1.661 cp_fm_cholesky_decompose 22 10.9 1.640 1.653 1.640 1.653 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.513 1.635 mp_waitany 9240 13.8 1.133 1.582 1.133 1.582 mp_allgather_i34 2055 14.4 0.479 1.522 0.479 1.522 dbcsr_complete_redistribute 325 12.2 0.563 0.586 1.420 1.505 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.418 1.426 jit_kernel_multiply 8 15.9 0.999 1.361 0.999 1.361 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.480 1.351 qs_create_task_list 11 7.9 0.000 0.001 1.223 1.327 generate_qs_task_list 11 8.9 0.375 0.444 1.223 1.326 mp_irecv_dv 24056 15.7 0.456 1.308 0.456 1.308 rs_pw_transfer_RS2PW_140 121 11.5 0.163 0.194 0.826 1.235 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.706 1.145 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.096 1.131 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.855000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=798.181818, yerr=12.452223 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 1.312965E+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 67104. MP_Allreduce 9672 819. 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 0.037 86.430 86.431 qs_mol_dyn_low 1 2.0 0.003 0.004 86.102 86.110 qs_forces 11 3.9 0.003 0.003 86.037 86.038 qs_energies 11 4.9 0.001 0.001 81.860 81.861 scf_env_do_scf 11 5.9 0.001 0.001 71.944 71.944 velocity_verlet 10 3.0 0.001 0.001 55.450 55.456 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 43.798 43.799 dbcsr_multiply_generic 2055 12.4 0.119 0.124 28.199 28.348 init_scf_loop 11 6.9 0.000 0.000 28.076 28.077 prepare_preconditioner 11 7.9 0.000 0.000 26.063 26.075 make_preconditioner 11 8.9 0.000 0.000 26.063 26.075 qs_scf_new_mos 99 7.5 0.001 0.001 25.674 25.760 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.674 25.760 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.241 25.545 ot_scf_mini 99 9.5 0.002 0.002 23.944 24.022 multiply_cannon 2055 13.4 0.340 0.367 21.308 22.010 multiply_cannon_loop 2055 14.4 0.340 0.344 19.532 19.904 cp_fm_upper_to_full 70 14.2 12.603 18.156 12.603 18.156 rebuild_ks_matrix 110 8.3 0.000 0.000 14.012 14.097 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.012 14.097 ot_mini 99 10.5 0.001 0.001 13.379 13.454 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.687 12.764 dbcsr_complete_redistribute 325 12.2 1.026 1.053 7.368 10.582 multiply_cannon_multrec 8220 15.4 4.366 4.573 9.508 9.611 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.352 9.552 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.806 8.965 qs_ot_get_derivative 99 11.5 0.001 0.001 8.767 8.845 mp_alltoall_i22 605 13.7 5.452 8.684 5.452 8.684 mp_waitall_1 84994 16.7 7.346 8.300 7.346 8.300 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.669 7.704 calculate_rho_elec 110 8.6 0.225 0.225 7.668 7.703 sum_up_and_integrate 110 10.3 0.150 0.152 7.253 7.268 integrate_v_rspace 110 11.3 0.004 0.004 7.103 7.118 init_scf_run 11 5.9 0.000 0.001 5.663 5.663 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.662 5.663 make_m2s 4110 13.4 0.043 0.043 5.137 5.660 qs_ot_get_p 110 10.4 0.001 0.001 5.434 5.515 make_images 4110 14.4 0.879 0.932 4.951 5.471 dbcsr_mm_accdrv_process 11614 15.7 3.202 3.604 5.001 5.270 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.693 5.207 apply_single 110 13.6 0.000 0.000 4.693 5.207 cp_fm_cholesky_invert 11 10.9 5.151 5.156 5.151 5.156 pw_transfer 1331 11.6 0.075 0.076 5.012 5.025 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.685 4.945 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.894 4.907 ot_diis_step 99 11.5 0.015 0.016 4.596 4.596 fft_wrap_pw1pw2_140 451 13.1 1.338 1.343 4.272 4.282 multiply_cannon_sync_h2d 8220 15.4 3.948 3.955 3.948 3.955 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.906 3.913 density_rs2pw 110 9.6 0.004 0.005 3.708 3.743 grid_integrate_task_list 110 12.3 3.663 3.733 3.663 3.733 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.654 3.655 hybrid_alltoall_any 4261 16.3 0.257 0.549 2.819 3.555 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.472 3.472 make_images_data 4110 15.4 0.041 0.044 2.790 3.407 wfi_extrapolate 11 7.9 0.001 0.001 3.309 3.309 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.798 3.267 calculate_dm_sparse 110 9.5 0.001 0.001 3.086 3.110 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.886 2.887 cp_fm_diag_elpa_base 48 14.0 2.351 2.549 2.885 2.885 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.835 2.841 fft3d_ps 1111 14.6 1.302 1.327 2.822 2.833 grid_collocate_task_list 110 9.6 2.631 2.651 2.631 2.651 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.339 2.381 potential_pw2rs 110 12.3 0.022 0.022 2.373 2.379 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.335 2.369 rs_pw_transfer 902 11.9 0.011 0.011 2.237 2.266 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.182 2.250 calculate_first_density_matrix 1 7.0 0.000 0.000 2.236 2.237 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.104 2.192 mp_alltoall_d11v 2046 13.8 1.905 1.991 1.905 1.991 cp_fm_cholesky_decompose 22 10.9 1.941 1.961 1.941 1.961 qs_create_task_list 11 7.9 0.000 0.001 1.888 1.935 generate_qs_task_list 11 8.9 0.734 0.787 1.888 1.935 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.924 1.933 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.774 1.820 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.431000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1193.272727, yerr=46.378073 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 630.693888E+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 57910. MP_Allreduce 10979 803. MP_Sync 87 MP_Alltoall 2226 1785057. 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 0.021 0.066 205.347 205.348 qs_mol_dyn_low 1 2.0 0.003 0.010 204.885 204.900 qs_forces 11 3.9 0.006 0.012 204.793 204.795 qs_energies 11 4.9 0.001 0.003 199.211 199.226 scf_env_do_scf 11 5.9 0.001 0.001 182.679 182.683 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 162.635 162.637 dbcsr_multiply_generic 2507 12.6 0.179 0.183 124.095 124.582 qs_scf_new_mos 117 7.6 0.001 0.001 122.931 123.127 qs_scf_loop_do_ot 117 8.6 0.001 0.001 122.930 123.127 velocity_verlet 10 3.0 0.001 0.002 122.096 122.098 ot_scf_mini 117 9.6 0.003 0.003 116.334 116.523 multiply_cannon 2507 13.6 0.239 0.249 100.921 102.190 multiply_cannon_loop 2507 14.6 2.066 2.116 98.828 100.160 ot_mini 117 10.6 0.001 0.002 65.411 65.591 multiply_cannon_multrec 60168 15.6 33.515 35.385 41.584 43.525 qs_ot_get_derivative 117 11.6 0.001 0.002 40.479 40.658 rebuild_ks_matrix 128 8.3 0.001 0.001 33.566 33.845 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.020 33.565 33.845 mp_waitall_1 267128 16.5 28.096 31.143 28.096 31.143 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.132 30.368 multiply_cannon_sync_h2d 60168 15.6 27.651 29.266 27.651 29.266 qs_ot_get_p 128 10.4 0.001 0.001 28.899 29.173 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.497 25.058 apply_single 128 13.6 0.001 0.001 24.496 25.058 ot_diis_step 117 11.6 0.008 0.011 24.648 24.649 qs_ot_p2m_diag 83 11.4 0.079 0.092 22.478 22.542 init_scf_loop 11 6.9 0.000 0.001 19.972 19.973 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.945 19.946 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.355 19.506 multiply_cannon_metrocomm3 60168 15.6 0.113 0.117 15.776 17.779 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.014 17.041 cp_fm_diag_elpa_base 83 14.4 16.947 16.980 17.010 17.035 prepare_preconditioner 11 7.9 0.000 0.000 15.412 15.450 make_preconditioner 11 8.9 0.000 0.000 15.412 15.450 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.657 14.834 make_m2s 5014 13.6 0.105 0.112 13.732 14.135 sum_up_and_integrate 128 10.3 0.090 0.111 14.078 14.096 integrate_v_rspace 128 11.3 0.004 0.005 13.988 14.010 make_images 5014 14.6 0.399 0.418 13.550 13.963 qs_rho_update_rho_low 128 7.7 0.001 0.002 13.503 13.630 calculate_rho_elec 128 8.7 0.046 0.065 13.502 13.629 init_scf_run 11 5.9 0.000 0.001 12.278 12.278 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.277 12.278 density_rs2pw 128 9.7 0.006 0.007 6.889 10.132 mp_sum_l 7870 13.0 8.193 9.305 8.193 9.305 wfi_extrapolate 11 7.9 0.004 0.029 9.060 9.060 rs_pw_transfer 1046 11.9 0.017 0.019 5.550 8.803 cp_fm_cholesky_invert 11 10.9 8.699 8.706 8.699 8.706 calculate_dm_sparse 128 9.5 0.001 0.001 8.423 8.537 dbcsr_mm_accdrv_process 124484 16.2 3.064 3.198 7.632 8.156 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.857 8.007 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.914 7.988 pw_transfer 1547 11.6 0.074 0.093 7.675 7.842 make_images_data 5014 15.6 0.066 0.074 6.777 7.712 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 7.473 7.639 grid_integrate_task_list 128 12.3 7.054 7.560 7.054 7.560 multiply_cannon_metrocomm1 60168 15.6 0.088 0.091 5.913 7.489 hybrid_alltoall_any 5200 16.5 0.288 2.252 5.952 7.218 fft_wrap_pw1pw2_140 523 13.2 1.273 1.318 6.589 6.771 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.631 6.644 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.801 5.925 mp_waitany 16020 13.9 2.659 5.895 2.659 5.895 grid_collocate_task_list 128 9.7 4.556 5.785 4.556 5.785 fft3d_ps 1291 14.7 2.140 2.784 5.241 5.541 mp_alltoall_d11v 2415 14.1 4.185 5.424 4.185 5.424 rs_pw_transfer_RS2PW_140 139 11.5 0.279 0.299 2.149 5.403 cp_fm_cholesky_decompose 22 10.9 4.552 4.565 4.552 4.565 potential_pw2rs 128 12.3 0.009 0.011 4.453 4.471 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.348000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.363636, yerr=7.535042 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 828.342272E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57909. MP_Allreduce 11058 964. MP_Sync 87 MP_Alltoall 1983 5004302. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.082 189.830 189.831 qs_mol_dyn_low 1 2.0 0.004 0.010 189.361 189.374 qs_forces 11 3.9 0.005 0.013 189.275 189.276 qs_energies 11 4.9 0.002 0.007 182.480 182.490 scf_env_do_scf 11 5.9 0.001 0.003 166.050 166.060 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 133.412 133.415 velocity_verlet 10 3.0 0.001 0.005 119.826 119.827 dbcsr_multiply_generic 2527 12.6 0.191 0.204 97.532 98.689 qs_scf_new_mos 118 7.6 0.001 0.001 93.198 93.659 qs_scf_loop_do_ot 118 8.6 0.001 0.002 93.197 93.659 ot_scf_mini 118 9.6 0.004 0.006 88.374 88.876 multiply_cannon 2527 13.6 0.483 0.535 77.806 82.104 multiply_cannon_loop 2527 14.6 1.244 1.272 74.482 76.917 ot_mini 118 10.6 0.001 0.002 49.877 50.344 mp_waitall_1 216438 16.6 24.334 39.284 24.334 39.284 multiply_cannon_multrec 30324 15.6 22.123 26.144 31.736 36.264 rebuild_ks_matrix 129 8.3 0.001 0.001 33.464 33.945 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 33.463 33.945 init_scf_loop 11 6.9 0.001 0.004 32.549 32.550 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.119 30.555 multiply_cannon_metrocomm3 30324 15.6 0.091 0.097 15.827 28.885 qs_ot_get_derivative 118 11.6 0.001 0.005 27.825 28.321 prepare_preconditioner 11 7.9 0.000 0.000 28.161 28.230 make_preconditioner 11 8.9 0.000 0.001 28.161 28.230 make_full_inverse_cholesky 11 9.9 0.000 0.001 26.873 27.381 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 22.080 23.118 apply_single 129 13.6 0.001 0.001 22.080 23.118 ot_diis_step 118 11.6 0.015 0.018 21.870 21.872 multiply_cannon_sync_h2d 30324 15.6 19.352 21.365 19.352 21.365 qs_ot_get_p 129 10.4 0.001 0.001 20.169 20.784 cp_fm_cholesky_invert 11 10.9 16.389 16.402 16.389 16.402 make_m2s 5054 13.6 0.090 0.095 14.076 15.752 make_images 5054 14.6 1.162 1.353 13.868 15.545 qs_ot_p2m_diag 83 11.4 0.190 0.223 15.405 15.439 sum_up_and_integrate 129 10.3 0.116 0.131 14.709 14.743 integrate_v_rspace 129 11.3 0.004 0.004 14.592 14.628 cp_dbcsr_syevd 83 12.4 0.006 0.008 14.281 14.283 qs_rho_update_rho_low 129 7.7 0.001 0.002 14.047 14.077 calculate_rho_elec 129 8.7 0.088 0.105 14.046 14.077 init_scf_run 11 5.9 0.000 0.001 11.574 11.576 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.574 11.575 qs_ot_get_derivative_diag 77 12.4 0.013 0.092 10.982 11.337 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.187 11.215 cp_fm_diag_elpa_base 83 14.4 10.935 11.032 11.180 11.211 multiply_cannon_metrocomm4 27797 15.6 0.103 0.116 3.801 10.883 mp_irecv_dv 70031 16.3 3.599 10.486 3.599 10.486 make_images_data 5054 15.6 0.065 0.073 8.523 10.479 density_rs2pw 129 9.7 0.007 0.008 7.384 9.876 hybrid_alltoall_any 5240 16.5 0.343 1.507 7.283 9.827 dbcsr_mm_accdrv_process 62734 16.2 4.424 5.036 9.067 9.649 pw_transfer 1559 11.6 0.087 0.101 8.780 8.838 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 8.553 8.614 wfi_extrapolate 11 7.9 0.001 0.001 8.369 8.369 rs_pw_transfer 1054 12.0 0.015 0.017 5.643 8.146 fft_wrap_pw1pw2_140 527 13.2 1.348 1.369 7.583 7.664 grid_integrate_task_list 129 12.3 7.238 7.636 7.238 7.636 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.252 7.008 cp_fm_cholesky_decompose 22 10.9 6.834 6.922 6.834 6.922 calculate_dm_sparse 129 9.5 0.001 0.002 6.476 6.625 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.201 6.210 grid_collocate_task_list 129 9.7 4.738 5.921 4.738 5.921 fft3d_ps 1301 14.7 2.831 3.012 5.837 5.888 mp_sum_l 7930 13.1 3.891 5.733 3.891 5.733 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.350 5.507 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.332 5.399 mp_waitany 11836 13.9 2.512 5.016 2.512 5.016 mp_alltoall_d11v 2423 14.1 4.095 4.955 4.095 4.955 potential_pw2rs 129 12.3 0.015 0.018 4.893 4.912 mp_allgather_i34 2527 14.6 1.885 4.907 1.885 4.907 rs_pw_transfer_RS2PW_140 140 11.5 0.354 0.378 2.113 4.583 mp_sum_d 4499 12.1 2.665 3.975 2.665 3.975 dbcsr_complete_redistribute 395 12.7 0.773 0.861 3.116 3.962 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.831000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=789.818182, yerr=0.574960 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239167967232 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239167967232 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233007E+12 0.0% 0.0% 100.0% flops max/rank 2.951022E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383584 0.0% 0.0% 100.0% number of processed stacks 4017216 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.3 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 946.794496E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1052064 MPI messages size (bytes): total size 2.737208E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.601750E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 284544 37295751168 131072 < size <= 4194304 665856 1004519030784 4194304 < size <= 16777216 66080 937890820720 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57635. MP_Allreduce 11104 1003. MP_Sync 88 MP_Alltoall 1724 9394234. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.076 0.168 175.849 175.851 qs_mol_dyn_low 1 2.0 0.003 0.004 175.276 175.289 qs_forces 11 3.9 0.004 0.005 175.177 175.180 qs_energies 11 4.9 0.002 0.006 168.540 168.552 scf_env_do_scf 11 5.9 0.001 0.003 152.820 152.822 scf_env_do_scf_inner_loop 118 6.6 0.008 0.042 118.149 118.151 velocity_verlet 10 3.0 0.001 0.002 113.371 113.374 dbcsr_multiply_generic 2529 12.6 0.182 0.187 82.026 83.275 qs_scf_new_mos 118 7.6 0.001 0.001 79.639 79.949 qs_scf_loop_do_ot 118 8.6 0.001 0.001 79.638 79.948 ot_scf_mini 118 9.6 0.004 0.004 75.465 75.879 multiply_cannon 2529 13.6 0.507 0.525 62.067 67.660 multiply_cannon_loop 2529 14.6 0.857 0.883 59.084 61.824 ot_mini 118 10.6 0.001 0.001 42.709 43.125 init_scf_loop 11 6.9 0.001 0.005 34.574 34.577 mp_waitall_1 172006 16.6 25.096 34.481 25.096 34.481 rebuild_ks_matrix 129 8.3 0.001 0.001 30.989 31.496 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.024 30.988 31.496 prepare_preconditioner 11 7.9 0.000 0.000 30.514 30.558 make_preconditioner 11 8.9 0.000 0.001 30.514 30.558 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.184 29.532 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.926 28.388 multiply_cannon_multrec 20232 15.6 13.713 16.792 22.373 25.432 multiply_cannon_metrocomm3 20232 15.6 0.058 0.062 15.252 24.646 qs_ot_get_derivative 118 11.6 0.001 0.002 22.891 23.320 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.898 20.932 apply_single 129 13.6 0.001 0.001 19.898 20.931 ot_diis_step 118 11.6 0.018 0.019 19.719 19.719 qs_ot_get_p 129 10.4 0.001 0.002 18.036 18.559 multiply_cannon_sync_h2d 20232 15.6 14.436 16.421 14.436 16.421 make_m2s 5058 13.6 0.080 0.092 14.850 15.710 make_images 5058 14.6 1.187 1.281 14.618 15.475 sum_up_and_integrate 129 10.3 0.134 0.145 14.672 14.703 integrate_v_rspace 129 11.3 0.004 0.004 14.537 14.564 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.368 14.406 calculate_rho_elec 129 8.7 0.131 0.147 14.367 14.406 cp_fm_cholesky_invert 11 10.9 14.172 14.181 14.172 14.181 qs_ot_p2m_diag 84 11.4 0.269 0.276 13.738 13.747 cp_dbcsr_syevd 84 12.4 0.008 0.022 12.740 12.740 init_scf_run 11 5.9 0.000 0.001 10.612 10.612 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.612 10.612 make_images_data 5058 15.6 0.060 0.067 9.192 10.497 hybrid_alltoall_any 5245 16.5 0.437 1.994 7.992 9.702 cp_fm_diag_elpa 84 13.4 0.000 0.001 9.580 9.601 cp_fm_diag_elpa_base 84 14.4 9.175 9.329 9.577 9.596 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.010 9.326 density_rs2pw 129 9.7 0.006 0.008 7.148 9.254 multiply_cannon_metrocomm4 17703 15.6 0.062 0.072 3.443 9.168 pw_transfer 1559 11.6 0.086 0.103 8.901 9.010 mp_irecv_dv 50659 16.2 3.321 8.927 3.321 8.927 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.675 8.786 dbcsr_mm_accdrv_process 41846 16.2 4.531 5.217 8.104 8.215 grid_integrate_task_list 129 12.3 7.304 7.874 7.304 7.874 fft_wrap_pw1pw2_140 527 13.2 1.425 1.453 7.722 7.847 wfi_extrapolate 11 7.9 0.001 0.001 7.451 7.451 cp_fm_cholesky_decompose 22 10.9 7.232 7.244 7.232 7.244 cp_fm_upper_to_full 106 14.8 5.745 7.202 5.745 7.202 rs_pw_transfer 1054 12.0 0.014 0.015 5.086 7.158 dbcsr_complete_redistribute 397 12.7 1.173 1.212 4.564 6.312 calculate_dm_sparse 129 9.5 0.001 0.001 5.772 5.880 fft3d_ps 1301 14.7 2.757 2.982 5.775 5.851 grid_collocate_task_list 129 9.7 4.936 5.800 4.936 5.800 mp_alltoall_d11v 2429 14.1 4.445 5.713 4.445 5.713 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.572 5.576 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.511 5.252 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.429 5.168 mp_sum_l 7936 13.1 3.325 5.050 3.325 5.050 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.629 4.744 potential_pw2rs 129 12.3 0.021 0.023 4.665 4.676 mp_allgather_i34 2529 14.6 1.479 4.501 1.479 4.501 mp_waitany 11836 13.9 2.365 4.460 2.365 4.460 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.994 4.019 rs_pw_transfer_RS2PW_140 140 11.5 0.333 0.366 1.946 4.004 transfer_fm_to_dbcsr 11 9.9 0.019 0.026 2.310 4.000 mp_alltoall_i22 720 14.1 1.955 3.846 1.955 3.846 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.832 3.833 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.568 3.614 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.851000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=895.909091, yerr=6.841777 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.159999E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58210. MP_Allreduce 11002 1090. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.034 185.961 185.962 qs_mol_dyn_low 1 2.0 0.004 0.010 185.486 185.522 qs_forces 11 3.9 0.010 0.012 185.379 185.384 qs_energies 11 4.9 0.002 0.011 178.225 178.235 scf_env_do_scf 11 5.9 0.001 0.002 161.557 161.564 velocity_verlet 10 3.0 0.002 0.002 123.722 123.725 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 115.612 115.613 dbcsr_multiply_generic 2507 12.6 0.188 0.194 78.802 79.647 qs_scf_new_mos 117 7.6 0.001 0.001 78.253 78.620 qs_scf_loop_do_ot 117 8.6 0.001 0.001 78.252 78.619 ot_scf_mini 117 9.6 0.003 0.004 73.795 74.118 multiply_cannon 2507 13.6 0.563 0.600 54.658 57.872 multiply_cannon_loop 2507 14.6 1.174 1.204 50.996 52.939 init_scf_loop 11 6.9 0.001 0.002 45.824 45.826 ot_mini 117 10.6 0.001 0.001 42.002 42.317 prepare_preconditioner 11 7.9 0.000 0.000 41.732 41.755 make_preconditioner 11 8.9 0.000 0.001 41.732 41.755 make_full_inverse_cholesky 11 9.9 0.013 0.023 35.418 40.398 multiply_cannon_multrec 30084 15.6 13.991 19.550 25.841 30.867 rebuild_ks_matrix 128 8.3 0.001 0.001 29.816 30.170 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 29.816 30.170 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.866 27.183 mp_waitall_1 147882 16.7 17.143 26.193 17.143 26.193 qs_ot_get_derivative 117 11.6 0.002 0.002 22.471 22.798 make_m2s 5014 13.6 0.096 0.101 19.797 20.759 make_images 5014 14.6 1.946 2.254 19.490 20.455 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.889 19.523 apply_single 128 13.6 0.001 0.001 18.888 19.522 ot_diis_step 117 11.6 0.018 0.018 19.418 19.419 qs_ot_get_p 128 10.4 0.001 0.002 18.171 18.522 cp_fm_upper_to_full 105 14.8 11.469 16.844 11.469 16.844 cp_fm_cholesky_invert 11 10.9 15.814 15.823 15.814 15.823 multiply_cannon_metrocomm3 30084 15.6 0.048 0.051 6.540 15.716 sum_up_and_integrate 128 10.3 0.140 0.152 14.683 14.712 integrate_v_rspace 128 11.3 0.004 0.004 14.543 14.576 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.489 14.524 calculate_rho_elec 128 8.7 0.173 0.189 14.489 14.523 qs_ot_p2m_diag 83 11.4 0.343 0.390 14.055 14.106 multiply_cannon_sync_h2d 30084 15.6 11.716 13.088 11.716 13.088 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.797 12.799 dbcsr_complete_redistribute 395 12.7 1.507 1.619 8.824 12.633 make_images_data 5014 15.6 0.062 0.065 10.608 12.107 dbcsr_mm_accdrv_process 62264 16.2 7.298 8.021 11.423 11.963 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.488 11.285 hybrid_alltoall_any 5200 16.5 0.525 2.209 9.574 11.082 init_scf_run 11 5.9 0.000 0.001 10.752 10.753 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.752 10.753 transfer_fm_to_dbcsr 11 9.9 0.001 0.003 6.292 9.973 cp_fm_diag_elpa 83 13.4 0.000 0.001 9.774 9.785 cp_fm_diag_elpa_base 83 14.4 8.824 9.143 9.768 9.779 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.300 9.531 mp_alltoall_i22 716 14.1 5.569 9.455 5.569 9.455 pw_transfer 1547 11.6 0.087 0.103 9.090 9.174 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.863 8.955 density_rs2pw 128 9.7 0.006 0.007 6.918 8.549 fft_wrap_pw1pw2_140 523 13.2 1.559 1.587 7.901 8.014 grid_integrate_task_list 128 12.3 7.496 7.971 7.496 7.971 wfi_extrapolate 11 7.9 0.001 0.001 7.689 7.689 cp_fm_cholesky_decompose 22 10.9 7.387 7.526 7.387 7.526 multiply_cannon_metrocomm4 25070 15.6 0.079 0.090 2.853 7.176 mp_irecv_dv 76098 16.2 2.705 6.905 2.705 6.905 rs_pw_transfer 1046 11.9 0.013 0.014 4.550 6.254 calculate_dm_sparse 128 9.5 0.001 0.001 6.150 6.229 fft3d_ps 1291 14.7 2.831 2.940 5.787 5.844 grid_collocate_task_list 128 9.7 5.060 5.826 5.060 5.826 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.416 5.471 mp_alltoall_d11v 2415 14.1 4.817 5.452 4.817 5.452 potential_pw2rs 128 12.3 0.024 0.024 4.443 4.461 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.453 4.453 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.361 4.447 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.333 4.435 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.116 4.170 mp_waitany 14952 13.9 2.229 3.788 2.229 3.788 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=185.962000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1096.272727, yerr=16.074208 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 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 1.511305E+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 716108580288 ------------------------------------------------------------------------------- - - - 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 4003 58193. MP_Allreduce 11005 1175. MP_Sync 86 MP_Alltoall 1700 18828189. 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 0.016 0.032 171.295 171.296 qs_mol_dyn_low 1 2.0 0.003 0.004 170.660 170.672 qs_forces 11 3.9 0.007 0.009 170.553 170.559 qs_energies 11 4.9 0.001 0.001 162.956 162.966 scf_env_do_scf 11 5.9 0.001 0.001 145.592 145.594 velocity_verlet 10 3.0 0.002 0.003 112.091 112.095 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 110.264 110.265 dbcsr_multiply_generic 2485 12.5 0.181 0.193 73.170 73.622 qs_scf_new_mos 116 7.6 0.001 0.001 72.269 72.393 qs_scf_loop_do_ot 116 8.6 0.001 0.001 72.268 72.393 ot_scf_mini 116 9.6 0.003 0.004 67.896 68.012 multiply_cannon 2485 13.5 0.582 0.609 53.790 58.438 multiply_cannon_loop 2485 14.5 0.437 0.447 49.234 49.971 ot_mini 116 10.6 0.001 0.001 38.960 39.058 init_scf_loop 11 6.9 0.000 0.000 35.183 35.184 mp_waitall_1 124680 16.7 25.497 32.359 25.497 32.359 prepare_preconditioner 11 7.9 0.000 0.000 31.284 31.305 make_preconditioner 11 8.9 0.000 0.000 31.284 31.305 rebuild_ks_matrix 127 8.3 0.001 0.001 29.518 29.622 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 29.518 29.621 make_full_inverse_cholesky 11 9.9 0.016 0.027 29.196 29.482 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.767 26.864 multiply_cannon_multrec 9940 15.5 10.319 14.222 17.525 20.606 ot_diis_step 116 11.6 0.020 0.022 19.698 19.698 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.304 19.596 apply_single 127 13.6 0.001 0.001 19.304 19.596 qs_ot_get_derivative 116 11.6 0.002 0.002 19.191 19.299 multiply_cannon_metrocomm3 9940 15.5 0.023 0.025 12.139 18.858 make_m2s 4970 13.5 0.066 0.071 15.890 18.284 make_images 4970 14.5 2.269 2.589 15.588 17.978 cp_fm_cholesky_invert 11 10.9 17.850 17.855 17.850 17.855 qs_ot_get_p 127 10.4 0.001 0.001 16.202 16.321 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.087 15.127 calculate_rho_elec 127 8.7 0.254 0.265 15.086 15.127 sum_up_and_integrate 127 10.3 0.178 0.187 14.798 14.848 integrate_v_rspace 127 11.3 0.004 0.004 14.620 14.674 qs_ot_p2m_diag 82 11.4 0.490 0.496 12.551 12.567 make_images_data 4970 15.5 0.051 0.059 9.716 12.187 multiply_cannon_sync_h2d 9940 15.5 11.554 12.125 11.554 12.125 hybrid_alltoall_any 5155 16.4 0.814 3.633 9.592 12.011 cp_dbcsr_syevd 82 12.4 0.005 0.005 11.451 11.452 init_scf_run 11 5.9 0.000 0.001 10.426 10.426 scf_env_initial_rho_setup 11 6.9 0.008 0.015 10.426 10.426 pw_transfer 1535 11.6 0.085 0.093 9.306 9.326 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.084 9.108 cp_fm_diag_elpa 82 13.4 0.000 0.000 8.545 8.556 cp_fm_diag_elpa_base 82 14.4 8.301 8.385 8.536 8.547 fft_wrap_pw1pw2_140 519 13.2 1.897 1.932 8.068 8.098 grid_integrate_task_list 127 12.3 7.706 8.071 7.706 8.071 cp_fm_cholesky_decompose 22 10.9 7.802 7.923 7.802 7.923 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 7.650 7.725 density_rs2pw 127 9.7 0.006 0.006 6.716 7.611 mp_allgather_i34 2485 14.5 2.762 7.610 2.762 7.610 dbcsr_mm_accdrv_process 20590 16.1 2.731 3.574 6.836 7.455 wfi_extrapolate 11 7.9 0.001 0.001 7.430 7.430 multiply_cannon_metrocomm1 9940 15.5 0.028 0.029 4.419 7.357 calculate_dm_sparse 127 9.5 0.001 0.001 6.027 6.107 grid_collocate_task_list 127 9.7 5.352 6.015 5.352 6.015 mp_alltoall_d11v 2401 14.1 4.861 5.708 4.861 5.708 fft3d_ps 1281 14.7 2.739 2.807 5.586 5.614 dbcsr_complete_redistribute 393 12.7 2.120 2.168 5.031 5.471 qs_energies_init_hamiltonians 11 5.9 0.014 0.017 5.402 5.403 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.104 5.120 rs_pw_transfer 1038 11.9 0.013 0.013 4.020 4.935 potential_pw2rs 127 12.3 0.026 0.027 4.314 4.327 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.052 4.069 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.560 3.910 multiply_cannon_metrocomm4 7455 15.5 0.024 0.027 1.843 3.881 mp_irecv_dv 28618 15.9 1.807 3.813 1.807 3.813 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.641 3.671 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.647 3.654 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.299 3.622 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.449 3.581 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=171.296000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1411.454545, yerr=40.533631 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1389586876416 0.0% 0.0% 100.0% flops 32 x 32 x 32 1887638126592 0.0% 0.0% 100.0% flops 22 x 9 x 32 1929499951104 0.0% 0.0% 100.0% flops 9 x 22 x 32 1935094689792 0.0% 0.0% 100.0% flops 22 x 22 x 32 2675276685312 0.0% 0.0% 100.0% flops 32 x 32 x 9 4300336005120 0.0% 0.0% 100.0% flops 32 x 32 x 22 5255966228480 0.0% 0.0% 100.0% flops 9 x 32 x 32 5299016564736 0.0% 0.0% 100.0% flops 22 x 32 x 32 6476575801344 0.0% 0.0% 100.0% flops 9 x 32 x 9 11276327374848 0.0% 0.0% 100.0% flops 22 x 32 x 9 14799245598720 0.0% 0.0% 100.0% flops 9 x 32 x 22 14799245598720 0.0% 0.0% 100.0% flops 22 x 32 x 22 19338589388800 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 91.362399E+12 0.0% 0.0% 100.0% flops max/rank 11.427044E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6604656480 0.0% 0.0% 100.0% number of processed stacks 1915696 0.0% 0.0% 100.0% average stack size 0.0 0.0 3447.7 marketing flops 141.366701E+12 ------------------------------------------------------------------------------- # multiplications 2442 max memory usage/rank 3.131855E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 97680 MPI messages size (bytes): total size 1.109909E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.362705E+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 43928 34091302912 4194304 < size <= 16777216 43248 370189271040 16777216 < size 9888 705625874544 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4013 58983. MP_Allreduce 11030 1515. MP_Sync 86 MP_Alltoall 1676 36913749. MP_SendRecv 1750 218624. MP_ISendRecv 1750 218624. MP_Wait 9580 MP_ISend 6264 1080795. MP_IRecv 6264 1080795. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.046 296.118 296.125 qs_mol_dyn_low 1 2.0 0.005 0.019 295.527 295.539 qs_forces 11 3.9 0.005 0.005 295.437 295.441 qs_energies 11 4.9 0.001 0.001 286.184 286.190 scf_env_do_scf 11 5.9 0.001 0.001 263.833 263.845 velocity_verlet 10 3.0 0.001 0.001 211.483 211.491 scf_env_do_scf_inner_loop 114 6.6 0.003 0.008 138.670 138.672 init_scf_loop 11 6.9 0.000 0.000 124.911 124.915 prepare_preconditioner 11 7.9 0.000 0.000 119.853 119.883 make_preconditioner 11 8.9 0.000 0.000 119.853 119.883 make_full_inverse_cholesky 11 9.9 0.037 0.039 95.593 116.996 qs_scf_new_mos 114 7.6 0.001 0.001 89.422 89.580 qs_scf_loop_do_ot 114 8.6 0.001 0.001 89.421 89.579 ot_scf_mini 114 9.6 0.003 0.004 84.678 84.781 dbcsr_multiply_generic 2442 12.5 0.210 0.220 82.021 82.883 cp_fm_upper_to_full 104 14.8 53.468 76.723 53.468 76.723 multiply_cannon 2442 13.5 0.700 0.748 57.647 58.606 multiply_cannon_loop 2442 14.5 0.456 0.465 53.975 55.951 ot_mini 114 10.6 0.001 0.001 43.904 44.036 dbcsr_complete_redistribute 393 12.7 4.004 4.038 29.914 43.329 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.538 39.927 rebuild_ks_matrix 125 8.3 0.001 0.001 37.417 37.572 qs_ks_build_kohn_sham_matrix 125 9.3 0.017 0.017 37.416 37.572 transfer_fm_to_dbcsr 11 9.9 0.027 0.032 24.214 37.482 mp_alltoall_i22 708 14.1 22.042 35.283 22.042 35.283 qs_ks_update_qs_env 125 7.6 0.001 0.001 34.443 34.586 cp_fm_cholesky_invert 11 10.9 33.145 33.151 33.145 33.151 mp_waitall_1 100996 16.8 28.165 32.484 28.165 32.484 qs_ot_get_p 125 10.4 0.001 0.001 25.676 25.814 qs_ot_get_derivative 114 11.6 0.002 0.002 23.695 23.796 qs_ot_p2m_diag 82 11.4 0.867 0.873 21.653 21.683 make_m2s 4884 13.5 0.074 0.078 19.612 20.590 multiply_cannon_metrocomm3 9768 15.5 0.023 0.025 18.744 20.516 ot_diis_step 114 11.6 0.021 0.022 20.166 20.166 make_images 4884 14.5 3.660 3.857 19.145 20.127 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.951 19.954 qs_rho_update_rho_low 125 7.7 0.001 0.001 19.907 19.923 calculate_rho_elec 125 8.7 0.470 0.470 19.906 19.922 apply_preconditioner_dbcsr 125 12.6 0.000 0.000 19.339 19.745 apply_single 125 13.6 0.001 0.001 19.339 19.745 sum_up_and_integrate 125 10.3 0.314 0.316 19.102 19.178 integrate_v_rspace 125 11.3 0.004 0.004 18.788 18.866 multiply_cannon_multrec 9768 15.5 10.222 11.946 17.438 17.541 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.807 16.808 cp_fm_diag_elpa_base 82 14.4 12.402 14.003 16.803 16.803 multiply_cannon_sync_h2d 9768 15.5 15.300 15.320 15.300 15.320 hybrid_alltoall_any 5069 16.4 1.276 2.972 10.795 12.636 make_images_data 4884 15.5 0.057 0.060 10.600 12.287 init_scf_run 11 5.9 0.000 0.001 12.249 12.249 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.248 12.249 pw_transfer 1511 11.6 0.091 0.092 11.124 11.134 fft_wrap_pw1pw2 1261 12.7 0.011 0.011 10.893 10.904 fft_wrap_pw1pw2_140 511 13.2 3.011 3.068 9.689 9.709 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.373 9.459 wfi_extrapolate 11 7.9 0.001 0.001 9.121 9.121 mp_alltoall_d11v 2385 14.1 8.123 8.896 8.123 8.896 cp_fm_cholesky_decompose 22 10.9 8.759 8.770 8.759 8.770 dbcsr_mm_accdrv_process 20250 16.0 3.722 5.367 6.985 8.727 grid_integrate_task_list 125 12.3 8.396 8.588 8.396 8.588 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.029 8.032 density_rs2pw 125 9.7 0.005 0.006 7.743 7.878 calculate_dm_sparse 125 9.5 0.001 0.001 6.585 6.656 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.296 6.344 grid_collocate_task_list 125 9.7 6.218 6.257 6.218 6.257 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.080 6.170 rs_scatter_matrices 136 9.7 3.609 4.365 5.911 6.123 fft3d_ps 1261 14.7 2.782 2.801 6.075 6.115 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=296.125000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2824.909091, yerr=164.271092 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 1.255895E+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 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.041 84.762 84.763 qs_energies 1 2.0 0.000 0.000 84.232 84.245 ls_scf 1 3.0 0.000 0.000 83.322 83.335 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.438 72.589 multiply_cannon 111 7.7 0.017 0.020 55.794 56.978 multiply_cannon_loop 111 8.7 0.208 0.224 52.425 53.750 ls_scf_main 1 4.0 0.000 0.000 52.197 52.197 density_matrix_trs4 2 5.0 0.002 0.003 46.690 46.783 ls_scf_init_scf 1 4.0 0.000 0.000 28.086 28.087 ls_scf_init_matrix_S 1 5.0 0.000 0.000 26.997 27.057 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.926 24.940 mp_waitall_1 11031 10.9 22.339 24.744 22.339 24.744 multiply_cannon_multrec 2664 9.7 8.178 8.936 15.467 17.281 multiply_cannon_sync_h2d 2664 9.7 13.680 15.709 13.680 15.709 make_m2s 222 7.7 0.008 0.011 12.994 13.371 make_images 222 8.7 0.099 0.109 12.972 13.349 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.668 12.216 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.442 8.936 make_images_data 222 9.7 0.004 0.005 7.558 8.053 dbcsr_mm_accdrv_process 4760 10.4 0.507 0.614 6.908 7.914 hybrid_alltoall_any 227 10.6 0.216 1.831 6.497 7.568 dbcsr_mm_accdrv_process_sort 4760 11.4 6.202 7.132 6.202 7.132 calculate_norms 4752 9.8 5.513 6.199 5.513 6.199 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.016 5.146 mp_sum_l 807 5.4 3.109 4.788 3.109 4.788 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.051 3.942 mp_irecv_dv 6231 10.9 2.035 3.908 2.035 3.908 make_images_sizes 222 9.7 0.000 0.000 0.674 3.888 mp_alltoall_i44 222 10.7 0.674 3.888 0.674 3.888 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.312 3.451 arnoldi_extremal 4 6.8 0.000 0.000 3.212 3.239 arnoldi_normal_ev 4 7.8 0.001 0.003 3.212 3.239 build_subspace 16 8.4 0.009 0.012 3.122 3.123 ls_scf_post 1 4.0 0.000 0.000 3.039 3.052 ls_scf_store_result 1 5.0 0.000 0.000 2.858 2.897 dbcsr_special_finalize 555 9.7 0.005 0.006 2.409 2.773 dbcsr_merge_single_wm 555 10.7 0.456 0.590 2.400 2.765 make_images_pack 222 9.7 2.207 2.640 2.209 2.641 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.315 2.577 dbcsr_sort_data 658 11.4 2.202 2.526 2.202 2.526 dbcsr_matrix_vector_mult_local 304 10.0 2.066 2.475 2.068 2.477 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.276 2.350 buffer_matrices_ensure_size 222 8.7 1.761 2.086 1.761 2.086 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.774 1.775 rebuild_ks_matrix 3 7.3 0.000 0.000 1.765 1.766 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.765 1.766 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.763000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1127.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 2.088657E+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 266696. MP_Allreduce 3058 10339. 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 0.029 0.047 91.136 91.137 qs_energies 1 2.0 0.000 0.000 90.658 90.663 ls_scf 1 3.0 0.000 0.000 89.351 89.356 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.468 75.848 multiply_cannon 111 7.7 0.029 0.038 53.201 57.502 ls_scf_main 1 4.0 0.000 0.000 55.034 55.042 multiply_cannon_loop 111 8.7 0.115 0.121 49.974 53.393 density_matrix_trs4 2 5.0 0.002 0.003 49.265 49.520 ls_scf_init_scf 1 4.0 0.000 0.000 30.791 30.792 mp_waitall_1 9105 10.9 21.287 29.968 21.287 29.968 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.614 29.713 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.211 27.220 multiply_cannon_multrec 1332 9.7 13.127 16.521 22.293 26.790 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.899 20.673 make_m2s 222 7.7 0.006 0.007 15.379 16.048 make_images 222 8.7 1.569 1.965 15.348 16.019 dbcsr_mm_accdrv_process 4041 10.4 0.313 0.452 8.766 10.305 dbcsr_mm_accdrv_process_sort 4041 11.4 8.359 9.853 8.359 9.853 make_images_data 222 9.7 0.004 0.004 8.863 9.805 hybrid_alltoall_any 227 10.6 0.520 2.476 8.255 9.168 mp_sum_l 807 5.4 5.592 8.578 5.592 8.578 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.208 7.592 mp_irecv_dv 3311 11.0 3.188 7.545 3.188 7.545 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.329 6.714 calculate_norms 2376 9.8 5.998 6.667 5.998 6.667 multiply_cannon_sync_h2d 1332 9.7 4.836 6.408 4.836 6.408 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.108 5.321 arnoldi_extremal 4 6.8 0.000 0.000 4.596 4.613 arnoldi_normal_ev 4 7.8 0.001 0.005 4.596 4.613 build_subspace 16 8.4 0.014 0.021 4.346 4.349 ls_scf_post 1 4.0 0.000 0.000 3.525 3.531 ls_scf_store_result 1 5.0 0.000 0.000 3.223 3.349 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.124 3.346 dbcsr_matrix_vector_mult_local 304 10.0 2.728 3.209 2.730 3.211 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.579 2.676 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.160 2.601 mp_allgather_i34 111 8.7 0.927 2.459 0.927 2.459 make_images_pack 222 9.7 2.023 2.436 2.025 2.438 dbcsr_sort_data 436 11.2 1.825 2.033 1.825 2.033 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.859 1.862 rebuild_ks_matrix 3 7.3 0.000 0.000 1.847 1.849 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.847 1.849 dbcsr_data_new 4174 10.1 1.609 1.833 1.609 1.833 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.137000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1708.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 2.762654E+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 265470. MP_Allreduce 3058 11181. 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 0.043 0.052 94.193 94.194 qs_energies 1 2.0 0.000 0.000 93.630 93.635 ls_scf 1 3.0 0.000 0.000 92.224 92.228 dbcsr_multiply_generic 111 6.7 0.016 0.016 76.815 77.162 ls_scf_main 1 4.0 0.000 0.000 57.785 57.791 multiply_cannon 111 7.7 0.040 0.087 52.752 57.444 multiply_cannon_loop 111 8.7 0.100 0.108 49.156 52.813 density_matrix_trs4 2 5.0 0.002 0.003 51.778 51.912 mp_waitall_1 7281 11.0 23.858 33.242 23.858 33.242 ls_scf_init_scf 1 4.0 0.000 0.000 30.845 30.847 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.628 29.744 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.240 27.253 multiply_cannon_multrec 888 9.7 12.590 15.088 21.066 23.924 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.060 22.154 make_m2s 222 7.7 0.006 0.007 17.231 18.465 make_images 222 8.7 1.973 2.297 17.193 18.428 hybrid_alltoall_any 227 10.6 0.618 2.880 9.449 10.886 make_images_data 222 9.7 0.003 0.004 9.827 10.821 dbcsr_mm_accdrv_process 3754 10.4 0.238 0.413 8.015 9.248 dbcsr_mm_accdrv_process_sort 3754 11.4 7.646 8.835 7.646 8.835 mp_sum_l 807 5.4 5.262 8.621 5.262 8.621 multiply_cannon_sync_h2d 888 9.7 6.085 7.578 6.085 7.578 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.110 6.894 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.470 6.877 mp_irecv_dv 2335 11.1 2.455 6.838 2.455 6.838 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.589 6.433 arnoldi_extremal 4 6.8 0.000 0.000 5.155 5.174 arnoldi_normal_ev 4 7.8 0.001 0.005 5.155 5.174 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.898 5.091 build_subspace 16 8.4 0.014 0.020 4.854 4.860 calculate_norms 1584 9.8 4.267 4.643 4.267 4.643 mp_allgather_i34 111 8.7 1.434 3.837 1.434 3.837 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.457 3.806 ls_scf_post 1 4.0 0.000 0.000 3.595 3.601 dbcsr_matrix_vector_mult_local 304 10.0 3.000 3.569 3.002 3.571 ls_scf_store_result 1 5.0 0.000 0.000 3.318 3.421 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.874 2.998 make_images_sizes 222 9.7 0.000 0.000 1.070 2.207 mp_alltoall_i44 222 10.7 1.070 2.207 1.070 2.207 dbcsr_sort_data 325 11.1 1.902 2.165 1.902 2.165 make_images_pack 222 9.7 1.819 2.128 1.821 2.131 dbcsr_data_release 9322 10.9 1.299 1.968 1.299 1.968 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.906 1.908 rebuild_ks_matrix 3 7.3 0.000 0.000 1.888 1.891 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.888 1.890 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.194000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2218.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 3.385758E+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 266696. MP_Allreduce 3058 13371. 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 0.070 0.137 97.321 97.322 qs_energies 1 2.0 0.000 0.000 96.508 96.514 ls_scf 1 3.0 0.000 0.001 94.860 94.867 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.657 78.928 ls_scf_main 1 4.0 0.000 0.001 58.952 58.953 multiply_cannon 111 7.7 0.068 0.123 51.619 56.335 density_matrix_trs4 2 5.0 0.004 0.020 52.822 52.956 multiply_cannon_loop 111 8.7 0.114 0.125 46.619 50.076 ls_scf_init_scf 1 4.0 0.000 0.001 32.641 32.643 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.449 31.509 mp_waitall_1 6369 11.0 22.891 29.496 22.891 29.496 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.990 29.003 multiply_cannon_multrec 1332 9.7 14.263 17.253 22.081 24.966 make_m2s 222 7.7 0.007 0.008 21.290 22.683 make_images 222 8.7 3.133 3.602 21.239 22.634 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.237 17.618 make_images_data 222 9.7 0.004 0.004 11.912 13.653 hybrid_alltoall_any 227 10.6 0.796 3.765 11.234 12.929 dbcsr_mm_accdrv_process 3641 10.4 0.225 0.399 7.447 8.961 dbcsr_mm_accdrv_process_sort 3641 11.4 7.077 8.552 7.077 8.552 mp_sum_l 807 5.4 4.173 7.684 4.173 7.684 multiply_cannon_sync_h2d 1332 9.7 5.494 6.223 5.494 6.223 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.206 6.060 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.073 6.047 mp_irecv_dv 3229 10.9 2.051 5.968 2.051 5.968 arnoldi_extremal 4 6.8 0.000 0.000 5.204 5.218 arnoldi_normal_ev 4 7.8 0.001 0.005 5.204 5.218 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.568 5.041 build_subspace 16 8.4 0.014 0.021 4.859 4.865 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.557 4.743 calculate_norms 2376 9.8 4.226 4.639 4.226 4.639 mp_allgather_i34 111 8.7 2.114 4.392 2.114 4.392 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.597 3.859 dbcsr_matrix_vector_mult_local 304 10.0 3.159 3.656 3.161 3.658 dbcsr_sort_data 658 11.4 3.012 3.363 3.012 3.363 ls_scf_post 1 4.0 0.000 0.001 3.267 3.272 dbcsr_special_finalize 555 9.7 0.006 0.007 2.750 3.196 dbcsr_merge_single_wm 555 10.7 0.529 0.650 2.742 3.188 ls_scf_store_result 1 5.0 0.000 0.000 2.979 3.069 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.995 3.066 dbcsr_data_release 10477 10.7 1.601 2.433 1.601 2.433 dbcsr_finalize 304 7.8 0.049 0.061 1.796 1.980 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.949 1.951 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.322000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2713.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 4.687634E+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 265558. MP_Allreduce 3049 15663. 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 0.043 0.062 98.914 98.915 qs_energies 1 2.0 0.000 0.000 98.156 98.161 ls_scf 1 3.0 0.000 0.000 96.247 96.257 dbcsr_multiply_generic 111 6.7 0.018 0.018 77.869 78.066 ls_scf_main 1 4.0 0.000 0.000 62.116 62.117 multiply_cannon 111 7.7 0.086 0.133 55.148 60.707 density_matrix_trs4 2 5.0 0.002 0.003 54.944 55.033 multiply_cannon_loop 111 8.7 0.069 0.078 50.618 52.408 mp_waitall_1 5436 11.0 26.576 32.483 26.576 32.483 ls_scf_init_scf 1 4.0 0.000 0.000 30.543 30.548 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.313 29.349 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.159 27.170 multiply_cannon_multrec 444 9.7 14.065 16.470 21.067 24.420 make_m2s 222 7.7 0.004 0.005 17.810 20.311 make_images 222 8.7 3.720 4.424 17.749 20.252 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.247 16.554 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.009 14.407 make_images_data 222 9.7 0.003 0.004 10.117 12.488 hybrid_alltoall_any 227 10.6 0.787 3.766 9.805 12.270 multiply_cannon_sync_h2d 444 9.7 6.505 8.882 6.505 8.882 dbcsr_mm_accdrv_process 3003 10.4 0.163 0.344 6.707 7.806 dbcsr_mm_accdrv_process_sort 3003 11.4 6.393 7.462 6.393 7.462 mp_allgather_i34 111 8.7 2.760 6.969 2.760 6.969 arnoldi_extremal 4 6.8 0.000 0.000 5.755 5.767 arnoldi_normal_ev 4 7.8 0.002 0.005 5.755 5.767 mp_sum_l 807 5.4 2.951 5.672 2.951 5.672 build_subspace 16 8.4 0.015 0.020 5.366 5.375 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.600 4.761 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.132 4.342 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.024 4.262 dbcsr_matrix_vector_mult_local 304 10.0 3.638 4.115 3.640 4.117 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.626 4.022 mp_irecv_dv 1241 11.2 1.608 4.000 1.608 4.000 calculate_norms 792 9.8 3.529 3.674 3.529 3.674 make_images_sizes 222 9.7 0.000 0.000 1.046 3.638 mp_alltoall_i44 222 10.7 1.045 3.638 1.045 3.638 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.532 3.627 ls_scf_post 1 4.0 0.000 0.000 3.588 3.592 ls_scf_store_result 1 5.0 0.000 0.000 3.355 3.414 dbcsr_finalize 304 7.8 0.062 0.078 2.199 2.287 dbcsr_merge_all 275 8.9 0.472 0.519 2.049 2.126 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.081 2.083 rebuild_ks_matrix 3 7.3 0.000 0.000 2.049 2.051 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.000 2.049 2.051 dbcsr_data_release 10123 10.8 1.326 1.996 1.326 1.996 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.915000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3650.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 8.708780E+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 284111. MP_Allreduce 3043 21950. 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 0.081 0.102 108.340 108.341 qs_energies 1 2.0 0.000 0.000 106.920 106.934 ls_scf 1 3.0 0.000 0.000 103.862 103.876 dbcsr_multiply_generic 111 6.7 0.024 0.028 77.581 77.726 ls_scf_main 1 4.0 0.000 0.000 65.989 65.991 density_matrix_trs4 2 5.0 0.002 0.003 56.776 56.849 multiply_cannon 111 7.7 0.174 0.237 50.063 51.821 multiply_cannon_loop 111 8.7 0.069 0.075 46.630 47.495 ls_scf_init_scf 1 4.0 0.000 0.000 34.189 34.190 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.658 32.669 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.815 29.828 mp_waitall_1 4527 11.1 22.446 26.423 22.446 26.423 make_m2s 222 7.7 0.005 0.005 23.882 24.832 make_images 222 8.7 4.579 4.963 23.775 24.724 multiply_cannon_multrec 444 9.7 17.849 18.572 22.441 23.196 hybrid_alltoall_any 227 10.6 1.657 3.614 12.891 15.778 make_images_data 222 9.7 0.003 0.004 13.119 15.513 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.719 11.554 multiply_cannon_sync_h2d 444 9.7 8.849 8.892 8.849 8.892 arnoldi_extremal 4 6.8 0.000 0.000 7.266 7.279 arnoldi_normal_ev 4 7.8 0.003 0.009 7.266 7.279 build_subspace 16 8.4 0.026 0.036 6.711 6.722 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.348 5.511 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.050 5.297 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.151 5.243 dbcsr_matrix_vector_mult_local 304 10.0 4.907 5.222 4.910 5.224 dbcsr_mm_accdrv_process 1814 10.4 0.211 0.322 4.415 4.537 dbcsr_mm_accdrv_process_sort 1814 11.4 4.113 4.242 4.113 4.242 ls_scf_post 1 4.0 0.000 0.000 3.684 3.697 make_images_sizes 222 9.7 0.000 0.000 1.439 3.557 mp_alltoall_i44 222 10.7 1.439 3.556 1.439 3.556 ls_scf_store_result 1 5.0 0.000 0.000 3.434 3.445 mp_allgather_i34 111 8.7 1.063 3.382 1.063 3.382 calculate_norms 792 9.8 3.239 3.277 3.239 3.277 dbcsr_finalize 304 7.8 0.082 0.089 3.089 3.180 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 3.027 3.027 dbcsr_merge_all 275 8.9 0.890 0.914 2.864 2.943 dbcsr_data_release 12724 10.6 2.344 2.882 2.344 2.882 dbcsr_complete_redistribute 5 7.6 1.430 1.463 2.762 2.867 matrix_ls_to_qs 2 6.0 0.000 0.000 2.417 2.540 dbcsr_sort_data 325 11.1 2.438 2.501 2.438 2.501 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.463 2.465 rebuild_ks_matrix 3 7.3 0.000 0.000 2.398 2.399 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.398 2.399 dbcsr_new_transposed 4 7.5 0.244 0.254 2.307 2.325 dbcsr_frobenius_norm 74 6.6 2.057 2.135 2.189 2.231 dbcsr_add_d 103 6.2 0.000 0.000 2.163 2.226 dbcsr_add_anytype 103 7.2 0.859 0.892 2.163 2.226 mp_sum_l 807 5.4 1.309 2.175 1.309 2.175 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.341000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6856.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e162f23508ef73105089f58ab0f087a87f283da0_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 582.221824E+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 76559. MP_Allreduce 13152 2094. MP_Sync 1064 MP_Alltoall 2588 640525995. 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 0.021 0.118 226.288 226.289 qs_mol_dyn_low 1 2.0 0.004 0.067 225.286 225.305 qs_forces 5 3.8 0.006 0.091 225.136 225.140 qs_energies 5 4.8 0.001 0.024 222.113 222.133 scf_env_do_scf 5 5.8 0.000 0.002 204.968 204.971 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 178.925 178.926 qs_scf_new_mos 105 7.6 0.000 0.001 138.356 138.539 qs_scf_loop_do_ot 105 8.6 0.001 0.001 138.356 138.539 dbcsr_multiply_generic 1445 12.2 0.126 0.135 132.835 133.258 ot_scf_mini 105 9.6 0.003 0.003 128.748 128.886 multiply_cannon 1445 13.2 0.276 0.288 114.710 116.709 multiply_cannon_loop 1445 14.2 2.376 2.498 113.030 114.090 velocity_verlet 4 3.0 0.001 0.009 105.014 105.017 ot_mini 105 10.6 0.001 0.002 58.347 58.475 multiply_cannon_multrec 69360 15.2 31.901 36.765 39.489 44.801 mp_waitall_1 488190 16.1 32.601 41.231 32.601 41.231 qs_ot_get_p 112 10.4 0.001 0.001 40.838 41.108 qs_ot_get_derivative 55 11.6 0.001 0.005 36.726 36.865 multiply_cannon_sync_h2d 69360 15.2 30.920 35.000 30.920 35.000 multiply_cannon_metrocomm3 69360 15.2 0.194 0.204 23.996 32.814 qs_ot_p2m_diag 40 11.0 0.020 0.030 30.268 30.358 rebuild_ks_matrix 110 8.4 0.000 0.000 29.198 29.365 qs_ks_build_kohn_sham_matrix 110 9.4 0.013 0.027 29.198 29.364 cp_dbcsr_syevd 40 12.0 0.002 0.002 27.230 27.230 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.804 26.952 init_scf_loop 7 6.6 0.000 0.000 26.013 26.014 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 22.995 23.235 apply_single 62 13.6 0.000 0.000 22.995 23.235 cp_fm_syevd 40 13.0 0.000 0.001 22.054 22.234 prepare_preconditioner 7 7.6 0.000 0.000 21.184 21.216 make_preconditioner 7 8.6 0.000 0.000 21.184 21.216 ot_new_cg_direction 55 11.6 0.001 0.002 20.962 20.962 qs_rho_update_rho_low 110 7.6 0.001 0.002 17.283 17.636 calculate_rho_elec 110 8.6 0.030 0.033 17.283 17.636 cp_fm_redistribute_end 40 14.0 8.711 17.384 8.715 17.386 cp_fm_syevd_base 40 14.0 8.663 17.337 8.663 17.337 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.385 14.571 init_scf_run 5 5.8 0.000 0.001 14.443 14.444 scf_env_initial_rho_setup 5 6.8 0.001 0.003 14.443 14.444 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.274 14.333 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.396 13.539 rs_pw_transfer 690 11.5 0.011 0.013 11.500 12.725 density_rs2pw 110 9.6 0.006 0.007 11.330 12.578 pw_transfer 1645 12.4 0.084 0.109 12.095 12.345 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 11.952 12.206 mp_sum_l 4684 12.4 11.475 12.093 11.475 12.093 calculate_dm_sparse 110 9.5 0.000 0.001 11.139 11.321 qs_vxc_create 110 10.4 0.002 0.003 10.481 10.531 fft_wrap_pw1pw2_240 915 15.0 1.191 1.304 10.185 10.391 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 10.114 10.132 cp_fm_cholesky_invert 7 10.6 9.985 9.993 9.985 9.993 acc_transpose_blocks 69360 15.2 0.360 0.374 9.399 9.928 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 9.704 9.763 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 9.651 9.731 calculate_first_density_matrix 1 7.0 0.000 0.002 9.636 9.649 check_diag 80 13.5 8.591 8.907 9.474 9.631 sum_up_and_integrate 60 10.3 0.028 0.031 8.734 8.746 integrate_v_rspace 60 11.3 0.002 0.002 8.706 8.718 fft3d_pb 915 16.0 2.390 2.666 8.248 8.520 dbcsr_mm_accdrv_process 154766 15.8 3.990 4.144 7.455 8.217 acc_transpose_blocks_kernels 69360 16.2 0.849 0.895 7.502 7.957 multiply_cannon_metrocomm1 69360 15.2 0.090 0.096 4.062 7.752 xc_rho_set_and_dset_create 110 12.4 0.075 0.096 7.361 7.616 jit_kernel_transpose 5 15.0 6.653 7.083 6.653 7.083 xc_vxc_pw_create 60 11.3 0.039 0.050 7.012 7.061 make_m2s 2890 13.2 0.078 0.087 6.215 6.841 make_images 2890 14.2 0.240 0.259 6.109 6.736 make_full_single_inverse 7 9.6 0.001 0.001 6.646 6.679 xc_pw_derive 510 13.4 0.005 0.007 6.257 6.344 mp_waitany 7680 13.5 4.564 5.720 4.564 5.720 mp_alltoall_z22v 2340 17.7 5.318 5.594 5.318 5.594 potential_pw2rs 60 12.3 0.003 0.003 4.897 4.916 multiply_cannon_metrocomm4 67915 15.2 0.181 0.199 2.051 4.743 grid_collocate_task_list 110 9.6 4.094 4.534 4.094 4.534 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=226.289000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=554.400000, yerr=2.059126 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: e162f23508ef73105089f58ab0f087a87f283da0 Summary: empty Status: OK