=== 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: 1edf9b76a3d3b174dfad07f04161737566d716f7 ################# 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 2.6.2, 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 (03.02.2023) # # \ 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 ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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_prefixed_scalapack.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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/01 job id: 44893326 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/02 job id: 44893327 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/03 job id: 44893328 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/04 job id: 44893330 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/05 job id: 44893331 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/06 job id: 44893332 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/07 job id: 44893333 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/08 job id: 44893334 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/09 job id: 44893335 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/10 job id: 44893336 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/11 job id: 44893337 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/12 job id: 44893338 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/13 job id: 44893339 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/14 job id: 44893340 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/15 job id: 44893341 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/16 job id: 44893342 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/17 job id: 44893343 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/18 job id: 44893344 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/19 job id: 44893345 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/20 job id: 44893346 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/21 job id: 44893347 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/22 job id: 44893348 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/23 job id: 44893349 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/24 job id: 44893351 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/25 job id: 44893352 --- 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/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/26 job id: 44893353 --- 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 ~~~~~~~ === 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]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 0.023 0.035 133.981 133.982 farming_run 1 2.0 133.313 133.315 133.950 133.954 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.450983E+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 229 1108280. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 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 0.010 0.024 115.358 115.359 qs_energies 1 2.0 0.000 0.000 115.074 115.076 mp2_main 1 3.0 0.000 0.000 113.070 113.072 mp2_gpw_main 1 4.0 0.019 0.026 112.166 112.167 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 93.500 93.580 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.349 55.430 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.168 41.657 46.561 get_2c_integrals 1 6.0 0.000 0.000 37.355 37.979 integrate_v_rspace 273 8.0 0.435 0.448 25.041 29.638 pw_transfer 6555 10.6 0.369 0.377 27.428 27.744 fft_wrap_pw1pw2 5465 11.4 0.044 0.045 26.100 26.245 grid_integrate_task_list 273 9.0 20.855 25.914 20.855 25.914 fft_wrap_pw1pw2_100 2178 12.4 1.206 1.445 23.622 23.794 compute_2c_integrals 1 7.0 0.003 0.003 19.631 19.633 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.929 19.354 mp2_eri_2c_integrate_gpw 1 9.0 2.381 2.434 18.925 19.350 rpa_ri_compute_en 1 5.0 0.001 0.002 18.560 18.641 cp_fm_cholesky_decompose 12 8.2 17.681 18.291 17.681 18.291 cholesky_decomp 1 7.0 0.000 0.000 16.571 17.178 fft3d_s 5443 13.4 16.152 16.388 16.174 16.411 ao_to_mo_and_store_B_mult_1 272 7.0 10.848 15.556 10.848 15.556 calculate_wavefunction 272 8.0 5.418 5.501 12.537 13.180 rpa_num_int 1 6.0 0.000 0.001 10.507 10.516 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.493 10.508 calc_mat_Q 8 8.0 0.000 0.000 9.339 9.432 contract_S_to_Q 8 9.0 0.000 0.000 8.759 8.854 calc_potential_gpw 544 9.5 0.005 0.005 8.254 8.713 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.246 8.551 parallel_gemm_fm 14 9.1 0.000 0.000 8.341 8.418 parallel_gemm_fm_cosma 14 10.1 8.341 8.418 8.341 8.418 potential_pw2rs 545 10.0 0.108 0.110 7.718 8.367 create_integ_mat 1 6.0 0.014 0.027 7.875 7.884 collocate_single_gaussian 272 10.0 0.039 0.042 7.485 7.703 array2fm 1 7.0 0.000 0.000 6.738 7.346 pw_scatter_s 2720 13.7 4.443 4.642 4.443 4.642 pw_gather_s 2722 13.2 3.890 4.334 3.890 4.334 array2fm_buffer_send 1 8.0 2.991 3.167 2.991 3.167 pw_poisson_solve 545 10.5 1.143 1.203 2.183 2.455 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.166213, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2723.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 0.079 0.101 398.193 398.194 farming_run 1 2.0 397.412 397.421 398.097 398.164 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.223053E+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 704 407793. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.034 210.837 210.837 qs_energies 1 2.0 0.000 0.000 210.532 210.546 scf_env_do_scf 1 3.0 0.000 0.000 107.372 107.372 qs_ks_update_qs_env 5 5.0 0.000 0.000 106.471 106.479 rebuild_ks_matrix 4 6.0 0.000 0.000 106.470 106.477 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.062 106.470 106.477 hfx_ks_matrix 4 8.0 0.001 0.001 106.068 106.071 integrate_four_center 4 9.0 0.144 0.465 106.067 106.071 mp2_main 1 3.0 0.000 0.000 102.861 102.874 mp2_gpw_main 1 4.0 0.033 0.050 101.965 101.983 integrate_four_center_main 4 10.0 0.098 0.548 97.271 99.981 integrate_four_center_bin 266 11.0 97.173 99.833 97.173 99.833 init_scf_loop 1 4.0 0.000 0.000 92.882 92.882 mp2_ri_gpw_compute_in 1 5.0 0.066 0.072 75.036 76.100 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.498 55.566 mp2_eri_3c_integrate_gpw 91 7.0 0.146 0.166 42.224 47.314 integrate_v_rspace 95 8.0 0.398 0.565 28.580 33.485 pw_transfer 2240 10.6 0.146 0.162 29.911 30.485 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.928 29.421 grid_integrate_task_list 95 9.0 23.846 28.954 23.846 28.954 ao_to_mo_and_store_B_mult_1 91 7.0 10.591 28.541 10.591 28.541 mp2_ri_gpw_compute_en 1 5.0 0.062 0.083 26.779 28.430 fft_wrap_pw1pw2_100 730 12.4 1.255 1.471 26.634 26.994 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.843 1.909 25.021 25.032 get_2c_integrals 1 6.0 0.000 0.000 20.438 20.474 compute_2c_integrals 1 7.0 0.030 0.115 19.421 19.427 compute_2c_integrals_loop_lm 1 8.0 0.002 0.006 18.792 19.155 mp2_eri_2c_integrate_gpw 1 9.0 1.740 1.865 18.790 19.154 fft3d_s 1823 13.4 18.432 18.734 18.445 18.747 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.489 14.489 calculate_wavefunction 91 8.0 2.041 2.072 9.768 10.058 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.577 8.801 9.523 potential_pw2rs 186 10.0 0.033 0.035 8.634 9.192 local_gemm 172 8.0 8.246 8.959 8.246 8.959 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.516 7.986 8.712 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.193 8.538 calc_potential_gpw 182 9.5 0.002 0.002 7.896 8.159 collocate_single_gaussian 91 10.0 0.016 0.019 7.856 8.073 mp_sendrecv_dm3 2068 8.0 6.022 6.750 6.022 6.750 mp_sync 38 10.4 3.461 6.629 3.461 6.629 mp2_ri_gpw_compute_en_ener 172 7.0 6.347 6.457 6.347 6.457 pw_gather_s 912 13.2 4.924 5.392 4.924 5.392 pw_scatter_s 910 13.7 3.917 4.257 3.917 4.257 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.956269, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1510.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.014656E+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 271. MP_Sync 580 MP_Alltoall 2083 260533. 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 0.012 0.029 52.064 52.065 qs_mol_dyn_low 1 2.0 0.003 0.004 51.831 51.839 qs_forces 11 3.9 0.002 0.002 51.771 51.772 qs_energies 11 4.9 0.002 0.002 50.302 50.317 scf_env_do_scf 11 5.9 0.000 0.001 44.219 44.219 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 42.138 42.138 dbcsr_multiply_generic 2286 12.5 0.093 0.096 33.444 33.909 qs_scf_new_mos 108 7.5 0.000 0.001 31.955 32.255 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.954 32.255 ot_scf_mini 108 9.5 0.002 0.002 30.289 30.501 multiply_cannon 2286 13.5 0.184 0.198 26.037 27.628 multiply_cannon_loop 2286 14.5 1.474 1.547 25.349 26.962 velocity_verlet 10 3.0 0.001 0.002 25.811 25.811 ot_mini 108 10.5 0.001 0.001 19.216 19.476 qs_ot_get_derivative 108 11.5 0.012 0.031 16.293 16.474 mp_waitall_1 245248 16.5 8.420 14.773 8.420 14.773 multiply_cannon_metrocomm3 54864 15.5 0.068 0.075 6.010 13.084 multiply_cannon_multrec 54864 15.5 4.225 6.593 7.732 11.261 rebuild_ks_matrix 119 8.3 0.000 0.000 7.902 8.044 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.902 8.044 multiply_cannon_sync_h2d 54864 15.5 5.910 7.326 5.910 7.326 mp_sum_l 7207 12.9 5.461 7.133 5.461 7.133 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.974 7.103 qs_ot_get_p 119 10.4 0.001 0.001 6.476 6.755 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.591 6.064 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.319 5.423 init_scf_run 11 5.9 0.000 0.001 4.872 4.872 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.872 4.872 dbcsr_mm_accdrv_process 76910 16.1 1.163 1.810 3.429 4.752 sum_up_and_integrate 119 10.3 0.012 0.014 4.494 4.500 integrate_v_rspace 119 11.3 0.002 0.002 4.481 4.489 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.231 4.341 calculate_rho_elec 119 8.7 0.011 0.017 4.231 4.340 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.641 3.675 calculate_dm_sparse 119 9.5 0.000 0.001 3.032 3.196 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.880 3.060 apply_single 119 13.6 0.000 0.000 2.879 3.060 multiply_cannon_metrocomm1 54864 15.5 0.051 0.056 1.765 2.855 jit_kernel_multiply 13 15.8 2.206 2.855 2.206 2.855 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.854 2.855 rs_pw_transfer 974 11.9 0.011 0.012 2.721 2.827 calculate_first_density_matrix 1 7.0 0.001 0.001 2.667 2.668 ot_diis_step 108 11.5 0.006 0.006 2.662 2.663 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.647 2.647 cp_fm_redistribute_end 50 14.0 2.409 2.625 2.415 2.627 cp_fm_diag_elpa_base 50 14.0 0.210 2.537 0.211 2.546 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.426 2.483 density_rs2pw 119 9.7 0.004 0.005 2.268 2.407 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.342 2.344 acc_transpose_blocks 54864 15.5 0.232 0.250 1.756 2.182 wfi_extrapolate 11 7.9 0.001 0.001 2.151 2.151 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.067 2.126 grid_integrate_task_list 119 12.3 2.014 2.126 2.014 2.126 init_scf_loop 11 6.9 0.000 0.000 2.065 2.065 pw_transfer 1439 11.6 0.052 0.058 1.834 1.911 potential_pw2rs 119 12.3 0.004 0.004 1.867 1.879 mp_sum_d 4127 12.0 1.286 1.870 1.286 1.870 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.758 1.839 make_m2s 4572 13.5 0.054 0.056 1.634 1.676 fft3d_ps 1201 14.6 0.369 0.471 1.529 1.601 make_images 4572 14.5 0.132 0.138 1.552 1.592 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.530 1.556 fft_wrap_pw1pw2_140 487 13.2 0.082 0.097 1.358 1.441 mp_alltoall_d11v 2130 13.8 1.236 1.414 1.236 1.414 mp_waitany 12084 13.8 1.211 1.357 1.211 1.357 grid_collocate_task_list 119 9.7 1.292 1.351 1.292 1.351 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.705 1.086 acc_transpose_blocks_kernels 54864 16.5 0.251 0.381 0.772 1.046 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.065000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.363636, yerr=0.881396 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 492.072960E+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 104 MP_Alltoall 2060 1617141. 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 0.013 0.031 39.274 39.275 qs_mol_dyn_low 1 2.0 0.003 0.003 39.078 39.086 qs_forces 11 3.9 0.002 0.002 39.022 39.022 qs_energies 11 4.9 0.001 0.002 37.321 37.323 scf_env_do_scf 11 5.9 0.000 0.001 32.093 32.094 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.516 29.517 dbcsr_multiply_generic 2286 12.5 0.100 0.102 21.315 21.677 qs_scf_new_mos 108 7.5 0.001 0.001 20.119 20.356 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.119 20.356 ot_scf_mini 108 9.5 0.002 0.003 19.231 19.403 velocity_verlet 10 3.0 0.001 0.002 18.552 18.553 multiply_cannon 2286 13.5 0.207 0.215 16.283 18.021 multiply_cannon_loop 2286 14.5 0.894 0.969 15.169 16.559 ot_mini 108 10.5 0.001 0.001 11.817 12.054 mp_waitall_1 200699 16.5 5.683 11.091 5.683 11.091 multiply_cannon_metrocomm3 27432 15.5 0.068 0.072 4.166 9.538 qs_ot_get_derivative 108 11.5 0.001 0.001 9.346 9.523 multiply_cannon_multrec 27432 15.5 1.968 4.290 5.887 8.677 rebuild_ks_matrix 119 8.3 0.000 0.000 7.301 7.448 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.300 7.447 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.452 6.585 dbcsr_mm_accdrv_process 47894 16.0 3.015 5.451 3.851 5.702 qs_ot_get_p 119 10.4 0.001 0.001 4.525 4.757 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.589 4.434 sum_up_and_integrate 119 10.3 0.024 0.026 4.278 4.285 integrate_v_rspace 119 11.3 0.002 0.002 4.254 4.261 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.046 4.122 apply_single 119 13.6 0.000 0.000 3.045 4.122 init_scf_run 11 5.9 0.000 0.001 4.003 4.003 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.003 4.003 mp_sum_l 7207 12.9 2.065 3.963 2.065 3.963 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.932 3.960 calculate_rho_elec 119 8.7 0.021 0.024 3.931 3.959 rs_pw_transfer 974 11.9 0.010 0.011 2.720 3.154 qs_ot_p2m_diag 50 11.0 0.008 0.013 3.056 3.075 multiply_cannon_sync_h2d 27432 15.5 2.191 2.783 2.191 2.783 make_m2s 4572 13.5 0.052 0.054 2.516 2.738 density_rs2pw 119 9.7 0.004 0.004 2.260 2.707 make_images 4572 14.5 0.199 0.236 2.427 2.648 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.615 2.615 init_scf_loop 11 6.9 0.000 0.000 2.553 2.553 calculate_first_density_matrix 1 7.0 0.000 0.001 2.420 2.423 ot_diis_step 108 11.5 0.011 0.011 2.419 2.420 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.167 2.260 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.248 2.248 cp_fm_redistribute_end 50 14.0 1.858 2.219 1.862 2.220 cp_fm_diag_elpa_base 50 14.0 0.343 2.103 0.357 2.158 calculate_dm_sparse 119 9.5 0.000 0.000 2.038 2.114 pw_transfer 1439 11.6 0.065 0.072 2.040 2.073 potential_pw2rs 119 12.3 0.006 0.006 2.007 2.016 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.948 1.985 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.969 1.971 grid_integrate_task_list 119 12.3 1.837 1.945 1.837 1.945 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.700 1.744 fft3d_ps 1201 14.6 0.511 0.565 1.648 1.679 prepare_preconditioner 11 7.9 0.000 0.000 1.636 1.663 make_preconditioner 11 8.9 0.000 0.000 1.636 1.663 jit_kernel_multiply 8 16.1 0.783 1.615 0.783 1.615 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.527 1.585 fft_wrap_pw1pw2_140 487 13.2 0.079 0.085 1.493 1.532 wfi_extrapolate 11 7.9 0.001 0.001 1.531 1.531 make_images_data 4572 15.5 0.044 0.050 1.127 1.520 acc_transpose_blocks 27432 15.5 0.109 0.114 1.178 1.471 hybrid_alltoall_any 4725 16.4 0.050 0.110 0.986 1.437 mp_alltoall_d11v 2130 13.8 1.216 1.400 1.216 1.400 grid_collocate_task_list 119 9.7 1.226 1.367 1.226 1.367 mp_allgather_i34 2286 14.5 0.559 1.329 0.559 1.329 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.261 1.268 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.166 1.215 mp_sum_d 4127 12.0 0.579 1.030 0.579 1.030 rs_pw_transfer_RS2PW_140 130 11.5 0.138 0.147 0.531 0.960 mp_waitany 5720 13.7 0.511 0.943 0.511 0.943 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.933 0.934 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.908 0.922 acc_transpose_blocks_kernels 27432 16.5 0.182 0.273 0.648 0.853 mp_alltoall_z22v 1201 16.6 0.718 0.823 0.718 0.823 make_images_sizes 4572 15.5 0.005 0.005 0.576 0.810 mp_alltoall_i44 4572 16.5 0.572 0.806 0.572 0.806 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.606 0.751 0.787 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.275000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=468.727273, yerr=2.135880 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 525.246464E+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 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 720463. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 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 0.015 0.031 32.571 32.572 qs_mol_dyn_low 1 2.0 0.003 0.004 32.333 32.340 qs_forces 11 3.9 0.013 0.096 32.277 32.278 qs_energies 11 4.9 0.002 0.006 30.625 30.627 scf_env_do_scf 11 5.9 0.001 0.001 25.913 25.913 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.278 23.279 dbcsr_multiply_generic 2286 12.5 0.094 0.095 16.267 16.359 velocity_verlet 10 3.0 0.002 0.003 15.434 15.435 qs_scf_new_mos 108 7.5 0.001 0.001 14.879 14.901 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.879 14.901 ot_scf_mini 108 9.5 0.002 0.003 14.153 14.165 multiply_cannon 2286 13.5 0.194 0.200 12.964 13.787 multiply_cannon_loop 2286 14.5 0.632 0.671 12.143 12.973 ot_mini 108 10.5 0.001 0.001 8.665 8.684 qs_ot_get_derivative 108 11.5 0.001 0.001 7.172 7.182 multiply_cannon_multrec 18288 15.5 1.934 2.853 6.717 6.974 rebuild_ks_matrix 119 8.3 0.000 0.000 6.541 6.559 qs_ks_build_kohn_sham_matrix 119 9.3 0.015 0.036 6.541 6.559 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.794 5.809 dbcsr_mm_accdrv_process 38222 16.0 3.980 5.296 4.700 5.457 sum_up_and_integrate 119 10.3 0.029 0.030 4.078 4.082 integrate_v_rspace 119 11.3 0.002 0.003 4.048 4.056 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.567 3.572 calculate_rho_elec 119 8.7 0.030 0.031 3.566 3.572 init_scf_run 11 5.9 0.000 0.001 3.524 3.524 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.524 3.524 mp_waitall_1 158411 16.6 2.612 3.431 2.612 3.431 qs_ot_get_p 119 10.4 0.001 0.001 3.296 3.313 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.681 3.290 rs_pw_transfer 974 11.9 0.009 0.010 2.394 2.628 init_scf_loop 11 6.9 0.001 0.010 2.612 2.613 density_rs2pw 119 9.7 0.004 0.004 2.086 2.340 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.016 2.312 apply_single 119 13.6 0.000 0.000 2.015 2.312 calculate_first_density_matrix 1 7.0 0.002 0.013 2.277 2.278 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.202 2.206 jit_kernel_multiply 11 16.2 0.669 2.048 0.669 2.048 pw_transfer 1439 11.6 0.065 0.070 1.997 2.008 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.366 1.999 make_m2s 4572 13.5 0.044 0.045 1.805 1.965 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.937 1.937 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.904 1.918 grid_integrate_task_list 119 12.3 1.812 1.897 1.812 1.897 make_images 4572 14.5 0.188 0.199 1.720 1.878 calculate_dm_sparse 119 9.5 0.000 0.000 1.824 1.840 potential_pw2rs 119 12.3 0.007 0.008 1.803 1.809 prepare_preconditioner 11 7.9 0.000 0.001 1.749 1.752 make_preconditioner 11 8.9 0.000 0.002 1.749 1.752 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.605 1.687 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.684 1.685 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.666 1.668 cp_fm_redistribute_end 50 14.0 1.253 1.659 1.254 1.659 cp_fm_diag_elpa_base 50 14.0 0.389 1.567 0.403 1.621 fft3d_ps 1201 14.6 0.522 0.539 1.584 1.598 multiply_cannon_sync_h2d 18288 15.5 1.404 1.590 1.404 1.590 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.558 1.566 fft_wrap_pw1pw2_140 487 13.2 0.090 0.094 1.506 1.520 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.493 1.500 mp_sum_l 7207 12.9 1.119 1.477 1.119 1.477 ot_diis_step 108 11.5 0.011 0.011 1.470 1.470 grid_collocate_task_list 119 9.7 1.211 1.372 1.211 1.372 acc_transpose_blocks 18288 15.5 0.076 0.078 1.212 1.249 wfi_extrapolate 11 7.9 0.001 0.001 1.199 1.199 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.189 1.197 make_images_data 4572 15.5 0.044 0.048 0.800 0.973 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.958 0.959 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.887 0.907 hybrid_alltoall_any 4725 16.4 0.054 0.112 0.694 0.896 mp_alltoall_d11v 2130 13.8 0.709 0.836 0.709 0.836 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.820 0.822 acc_transpose_blocks_kernels 18288 16.5 0.209 0.222 0.774 0.791 mp_waitany 9880 13.7 0.535 0.788 0.535 0.788 mp_alltoall_z22v 1201 16.6 0.662 0.766 0.662 0.766 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.124 0.518 0.754 cp_fm_cholesky_invert 11 10.9 0.743 0.747 0.743 0.747 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.652 0.708 mp_allgather_i34 2286 14.5 0.286 0.653 0.286 0.653 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.572000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=498.000000, yerr=2.662876 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 544.997376E+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 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 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 0.020 0.039 35.573 35.575 qs_mol_dyn_low 1 2.0 0.003 0.004 35.343 35.350 qs_forces 11 3.9 0.002 0.005 35.180 35.180 qs_energies 11 4.9 0.002 0.006 33.437 33.445 scf_env_do_scf 11 5.9 0.001 0.002 28.365 28.366 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 24.894 24.896 velocity_verlet 10 3.0 0.001 0.001 18.160 18.166 dbcsr_multiply_generic 2286 12.5 0.099 0.101 17.999 18.143 qs_scf_new_mos 108 7.5 0.001 0.001 16.195 16.249 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.194 16.249 ot_scf_mini 108 9.5 0.002 0.003 15.261 15.314 multiply_cannon 2286 13.5 0.225 0.262 14.453 15.260 multiply_cannon_loop 2286 14.5 0.929 0.965 13.485 14.141 ot_mini 108 10.5 0.001 0.001 9.221 9.290 multiply_cannon_multrec 27432 15.5 2.347 3.021 8.611 9.003 qs_ot_get_derivative 108 11.5 0.001 0.001 7.414 7.468 dbcsr_mm_accdrv_process 47916 15.9 5.181 7.191 6.171 7.418 rebuild_ks_matrix 119 8.3 0.000 0.000 6.731 6.785 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 6.730 6.785 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.976 6.024 sum_up_and_integrate 119 10.3 0.035 0.038 3.925 3.934 integrate_v_rspace 119 11.3 0.003 0.003 3.889 3.899 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.674 3.707 calculate_rho_elec 119 8.7 0.040 0.046 3.674 3.706 init_scf_run 11 5.9 0.000 0.001 3.699 3.699 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.699 3.699 qs_ot_get_p 119 10.4 0.001 0.001 3.428 3.505 init_scf_loop 11 6.9 0.001 0.003 3.447 3.451 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.838 3.163 prepare_preconditioner 11 7.9 0.000 0.000 2.573 2.582 make_preconditioner 11 8.9 0.000 0.002 2.573 2.582 make_full_inverse_cholesky 11 9.9 0.000 0.001 2.185 2.508 mp_waitall_1 137007 16.6 1.897 2.343 1.897 2.343 make_m2s 4572 13.5 0.054 0.056 2.218 2.335 rs_pw_transfer 974 11.9 0.009 0.010 2.139 2.330 density_rs2pw 119 9.7 0.004 0.004 2.090 2.275 calculate_first_density_matrix 1 7.0 0.002 0.012 2.263 2.266 make_images 4572 14.5 0.269 0.331 2.110 2.227 pw_transfer 1439 11.6 0.066 0.071 2.163 2.206 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.961 2.152 apply_single 119 13.6 0.000 0.000 1.961 2.152 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.129 2.138 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.070 2.118 calculate_dm_sparse 119 9.5 0.000 0.000 2.018 2.083 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.923 1.952 grid_integrate_task_list 119 12.3 1.825 1.943 1.825 1.943 jit_kernel_multiply 10 15.8 0.931 1.837 0.931 1.837 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.827 1.831 cp_dbcsr_syevd 50 12.0 0.003 0.005 1.815 1.815 fft3d_ps 1201 14.6 0.558 0.611 1.745 1.787 ot_diis_step 108 11.5 0.012 0.012 1.765 1.765 fft_wrap_pw1pw2_140 487 13.2 0.088 0.095 1.691 1.742 potential_pw2rs 119 12.3 0.008 0.010 1.669 1.675 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.625 1.641 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.538 1.539 cp_fm_redistribute_end 50 14.0 1.016 1.509 1.017 1.509 cp_fm_diag_elpa_base 50 14.0 0.471 1.442 0.490 1.480 acc_transpose_blocks 27432 15.5 0.113 0.115 1.438 1.464 multiply_cannon_metrocomm3 27432 15.5 0.037 0.038 0.827 1.399 wfi_extrapolate 11 7.9 0.003 0.016 1.382 1.382 grid_collocate_task_list 119 9.7 1.220 1.333 1.220 1.333 mp_sum_l 7207 12.9 0.912 1.259 0.912 1.259 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.206 1.218 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.107 1.128 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.102 1.103 cp_fm_upper_to_full 72 13.5 0.793 1.098 0.793 1.098 dbcsr_complete_redistribute 329 12.2 0.123 0.147 0.817 1.095 multiply_cannon_sync_h2d 27432 15.5 0.995 1.046 0.995 1.046 make_images_data 4572 15.5 0.044 0.048 0.859 0.988 hybrid_alltoall_any 4725 16.4 0.061 0.152 0.728 0.916 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.871 mp_alltoall_d11v 2130 13.8 0.752 0.865 0.752 0.865 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.595 0.860 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.846 0.851 acc_transpose_blocks_kernels 27432 16.5 0.267 0.275 0.816 0.826 mp_alltoall_z22v 1201 16.6 0.772 0.811 0.772 0.811 cp_fm_cholesky_invert 11 10.9 0.769 0.772 0.769 0.772 mp_alltoall_i22 627 13.8 0.456 0.743 0.456 0.743 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.575000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=517.636364, yerr=2.869043 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 608.350208E+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 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 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 0.031 28.053 28.054 qs_mol_dyn_low 1 2.0 0.003 0.003 27.885 27.892 qs_forces 11 3.9 0.002 0.002 27.827 27.828 qs_energies 11 4.9 0.001 0.001 26.114 26.117 scf_env_do_scf 11 5.9 0.000 0.001 21.335 21.335 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 18.859 18.859 velocity_verlet 10 3.0 0.002 0.002 14.314 14.316 dbcsr_multiply_generic 2286 12.5 0.092 0.094 12.040 12.147 qs_scf_new_mos 108 7.5 0.001 0.001 10.811 10.834 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.810 10.834 ot_scf_mini 108 9.5 0.002 0.002 10.137 10.161 multiply_cannon 2286 13.5 0.229 0.235 9.594 10.092 multiply_cannon_loop 2286 14.5 0.327 0.336 8.684 8.899 multiply_cannon_multrec 9144 15.5 1.597 2.050 5.754 6.125 rebuild_ks_matrix 119 8.3 0.000 0.000 6.025 6.046 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.024 6.045 ot_mini 108 10.5 0.001 0.001 5.732 5.763 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.369 5.388 qs_ot_get_derivative 108 11.5 0.001 0.001 4.450 4.475 dbcsr_mm_accdrv_process 12550 15.8 3.190 3.984 4.056 4.149 sum_up_and_integrate 119 10.3 0.038 0.040 3.688 3.694 integrate_v_rspace 119 11.3 0.003 0.003 3.650 3.656 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.564 3.571 calculate_rho_elec 119 8.7 0.059 0.061 3.564 3.571 init_scf_run 11 5.9 0.000 0.001 3.345 3.345 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.345 3.345 qs_ot_get_p 119 10.4 0.001 0.001 2.659 2.697 init_scf_loop 11 6.9 0.000 0.000 2.457 2.458 calculate_first_density_matrix 1 7.0 0.001 0.001 2.217 2.217 pw_transfer 1439 11.6 0.066 0.070 2.161 2.173 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.067 2.082 density_rs2pw 119 9.7 0.004 0.004 1.892 2.031 mp_waitall_1 115863 16.7 1.503 1.993 1.503 1.993 grid_integrate_task_list 119 12.3 1.846 1.915 1.846 1.915 make_m2s 4572 13.5 0.034 0.035 1.689 1.863 rs_pw_transfer 974 11.9 0.008 0.008 1.668 1.815 make_images 4572 14.5 0.265 0.302 1.600 1.773 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.767 1.770 jit_kernel_multiply 10 15.7 0.829 1.752 0.829 1.752 fft3d_ps 1201 14.6 0.562 0.573 1.726 1.735 calculate_dm_sparse 119 9.5 0.000 0.000 1.713 1.730 fft_wrap_pw1pw2_140 487 13.2 0.088 0.090 1.716 1.730 prepare_preconditioner 11 7.9 0.000 0.000 1.702 1.706 make_preconditioner 11 8.9 0.000 0.000 1.702 1.706 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.591 1.619 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.576 1.577 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.523 1.524 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.489 1.510 potential_pw2rs 119 12.3 0.010 0.011 1.422 1.425 grid_collocate_task_list 119 9.7 1.270 1.391 1.270 1.391 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.370 1.377 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.313 1.314 cp_fm_redistribute_end 50 14.0 0.656 1.295 0.657 1.296 cp_fm_diag_elpa_base 50 14.0 0.594 1.226 0.637 1.283 ot_diis_step 108 11.5 0.013 0.013 1.270 1.270 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.227 1.228 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.153 1.182 apply_single 119 13.6 0.000 0.000 1.153 1.182 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.131 1.135 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.111 1.126 wfi_extrapolate 11 7.9 0.001 0.001 1.079 1.079 hybrid_alltoall_any 4725 16.4 0.062 0.174 0.758 1.024 make_images_data 4572 15.5 0.038 0.042 0.787 1.001 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.866 0.921 acc_transpose_blocks 9144 15.5 0.038 0.039 0.880 0.904 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.401 0.857 mp_alltoall_d11v 2130 13.8 0.751 0.848 0.751 0.848 cp_fm_cholesky_invert 11 10.9 0.846 0.848 0.846 0.848 multiply_cannon_sync_h2d 9144 15.5 0.710 0.793 0.710 0.793 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.777 0.780 qs_env_update_s_mstruct 11 6.9 0.003 0.005 0.682 0.730 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.707 0.722 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.641 0.664 mp_alltoall_z22v 1201 16.6 0.612 0.663 0.612 0.663 mp_allgather_i34 2286 14.5 0.239 0.642 0.239 0.642 yz_to_x 606 15.1 0.266 0.277 0.583 0.600 mp_waitany 5200 13.7 0.452 0.590 0.452 0.590 x_to_yz 595 16.2 0.276 0.289 0.570 0.576 qs_create_task_list 11 7.9 0.000 0.001 0.542 0.569 generate_qs_task_list 11 8.9 0.189 0.211 0.542 0.568 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.054000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=575.181818, yerr=5.490789 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 772.272128E+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 0.016 0.035 41.804 41.805 qs_mol_dyn_low 1 2.0 0.003 0.003 41.603 41.610 qs_forces 11 3.9 0.002 0.002 41.547 41.547 qs_energies 11 4.9 0.001 0.001 39.555 39.558 scf_env_do_scf 11 5.9 0.001 0.001 33.662 33.662 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.677 25.678 velocity_verlet 10 3.0 0.006 0.007 23.541 23.546 dbcsr_multiply_generic 2286 12.5 0.099 0.100 17.530 17.763 qs_scf_new_mos 108 7.5 0.001 0.001 15.827 15.930 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.826 15.929 ot_scf_mini 108 9.5 0.002 0.002 14.742 14.843 multiply_cannon 2286 13.5 0.303 0.309 13.705 14.632 multiply_cannon_loop 2286 14.5 0.342 0.350 12.456 13.457 ot_mini 108 10.5 0.001 0.001 8.765 8.882 multiply_cannon_multrec 9144 15.5 3.373 4.718 8.584 8.675 init_scf_loop 11 6.9 0.000 0.000 7.956 7.959 rebuild_ks_matrix 119 8.3 0.000 0.000 7.353 7.491 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.352 7.491 prepare_preconditioner 11 7.9 0.000 0.000 6.970 6.984 make_preconditioner 11 8.9 0.000 0.000 6.970 6.984 qs_ot_get_derivative 108 11.5 0.001 0.001 6.772 6.871 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.490 6.860 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.653 6.780 dbcsr_mm_accdrv_process 12550 15.8 4.299 6.308 5.087 6.369 cp_fm_upper_to_full 72 14.2 3.229 4.651 3.229 4.651 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.279 4.299 calculate_rho_elec 119 8.7 0.118 0.121 4.279 4.298 sum_up_and_integrate 119 10.3 0.065 0.067 4.050 4.057 integrate_v_rspace 119 11.3 0.003 0.003 3.985 3.993 init_scf_run 11 5.9 0.000 0.001 3.826 3.826 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.826 3.826 qs_ot_get_p 119 10.4 0.001 0.001 3.272 3.413 mp_waitall_1 94719 16.7 2.290 3.397 2.290 3.397 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.478 2.895 dbcsr_complete_redistribute 329 12.2 0.287 0.292 2.029 2.882 pw_transfer 1439 11.6 0.068 0.069 2.823 2.829 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.725 2.732 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.717 2.563 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.146 2.446 apply_single 119 13.6 0.000 0.000 2.145 2.446 make_m2s 4572 13.5 0.038 0.038 2.253 2.408 calculate_dm_sparse 119 9.5 0.000 0.000 2.377 2.394 fft3d_ps 1201 14.6 0.595 0.604 2.349 2.358 calculate_first_density_matrix 1 7.0 0.001 0.001 2.344 2.350 mp_alltoall_i22 627 13.8 1.457 2.319 1.457 2.319 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.476 2.315 fft_wrap_pw1pw2_140 487 13.2 0.094 0.098 2.292 2.299 make_images 4572 14.5 0.351 0.385 2.132 2.286 density_rs2pw 119 9.7 0.004 0.004 2.206 2.232 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.252 2.139 grid_integrate_task_list 119 12.3 2.054 2.066 2.054 2.066 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.000 2.002 ot_diis_step 108 11.5 0.014 0.014 1.966 1.966 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.844 1.845 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.792 1.792 mp_sum_l 7207 12.9 1.033 1.770 1.033 1.770 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.726 1.726 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.670 1.723 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.579 1.620 grid_collocate_task_list 119 9.7 1.486 1.505 1.486 1.505 cp_fm_cholesky_invert 11 10.9 1.489 1.492 1.489 1.492 rs_pw_transfer 974 11.9 0.009 0.009 1.444 1.477 potential_pw2rs 119 12.3 0.014 0.014 1.454 1.457 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.448 1.456 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.433 1.433 cp_fm_diag_elpa_base 50 14.0 1.285 1.339 1.431 1.431 wfi_extrapolate 11 7.9 0.001 0.001 1.408 1.408 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.114 1.339 make_images_data 4572 15.5 0.042 0.045 1.054 1.246 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.137 1.157 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.128 1.146 mp_alltoall_d11v 2130 13.8 1.082 1.108 1.082 1.108 acc_transpose_blocks 9144 15.5 0.038 0.039 1.075 1.092 jit_kernel_multiply 6 15.7 0.761 1.072 0.761 1.072 multiply_cannon_sync_h2d 9144 15.5 1.043 1.050 1.043 1.050 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.990 1.031 qs_create_task_list 11 7.9 0.016 0.036 0.972 0.984 generate_qs_task_list 11 8.9 0.370 0.388 0.956 0.984 yz_to_x 606 15.1 0.461 0.475 0.940 0.950 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.911 0.924 mp_alltoall_z22v 1201 16.6 0.802 0.840 0.802 0.840 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.805000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=724.818182, yerr=13.583534 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 503.783424E+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 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 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 0.032 0.087 84.270 84.271 qs_mol_dyn_low 1 2.0 0.008 0.022 83.702 83.710 qs_forces 11 3.9 0.010 0.015 83.620 83.628 qs_energies 11 4.9 0.004 0.013 80.763 80.790 scf_env_do_scf 11 5.9 0.001 0.002 71.694 71.698 scf_env_do_scf_inner_loop 99 6.5 0.002 0.009 65.919 65.919 dbcsr_multiply_generic 2055 12.4 0.107 0.110 52.464 52.717 qs_scf_new_mos 99 7.5 0.001 0.001 48.468 48.599 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.468 48.598 ot_scf_mini 99 9.5 0.002 0.003 46.071 46.158 multiply_cannon 2055 13.4 0.174 0.183 42.756 43.761 multiply_cannon_loop 2055 14.4 1.534 1.572 41.821 42.863 velocity_verlet 10 3.0 0.001 0.002 42.493 42.494 ot_mini 99 10.5 0.001 0.001 27.780 27.873 qs_ot_get_derivative 99 11.5 0.001 0.001 20.948 21.056 multiply_cannon_multrec 49320 15.4 12.122 13.002 17.194 17.950 rebuild_ks_matrix 110 8.3 0.000 0.000 14.645 14.782 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 14.645 14.782 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.923 13.060 mp_waitall_1 220248 16.4 11.804 12.669 11.804 12.669 multiply_cannon_sync_h2d 49320 15.4 9.908 10.580 9.908 10.580 qs_ot_get_p 110 10.4 0.001 0.001 9.684 9.790 multiply_cannon_metrocomm3 49320 15.4 0.078 0.083 7.122 8.476 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.201 7.742 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.180 7.638 apply_single 110 13.6 0.000 0.001 7.180 7.638 sum_up_and_integrate 110 10.3 0.036 0.043 7.150 7.166 integrate_v_rspace 110 11.3 0.003 0.004 7.113 7.137 init_scf_run 11 5.9 0.000 0.001 6.970 6.970 scf_env_initial_rho_setup 11 6.9 0.000 0.002 6.969 6.970 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.627 6.757 calculate_rho_elec 110 8.6 0.021 0.026 6.626 6.756 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.677 6.745 ot_diis_step 99 11.5 0.006 0.006 6.523 6.523 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.432 6.463 mp_sum_l 6514 12.8 5.416 6.279 5.416 6.279 init_scf_loop 11 6.9 0.025 0.112 5.746 5.748 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.402 5.402 dbcsr_mm_accdrv_process 87628 16.1 2.066 2.169 4.952 5.236 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.923 4.924 cp_fm_redistribute_end 48 14.0 4.296 4.897 4.300 4.898 cp_fm_diag_elpa_base 48 14.0 0.592 4.749 0.595 4.783 rs_pw_transfer 902 11.9 0.011 0.013 3.719 4.231 wfi_extrapolate 11 7.9 0.001 0.001 4.027 4.027 multiply_cannon_metrocomm1 49320 15.4 0.058 0.062 2.809 3.985 density_rs2pw 110 9.6 0.004 0.005 3.412 3.967 make_m2s 4110 13.4 0.060 0.065 3.860 3.966 calculate_dm_sparse 110 9.5 0.000 0.001 3.869 3.962 make_images 4110 14.4 0.176 0.190 3.765 3.874 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.675 3.679 prepare_preconditioner 11 7.9 0.000 0.000 3.437 3.457 make_preconditioner 11 8.9 0.000 0.001 3.437 3.457 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.344 3.428 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.356 3.421 grid_integrate_task_list 110 12.3 3.224 3.409 3.224 3.409 pw_transfer 1331 11.6 0.054 0.068 3.271 3.334 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.217 3.268 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.183 3.250 calculate_first_density_matrix 1 7.0 0.001 0.003 2.854 2.860 fft_wrap_pw1pw2_140 451 13.1 0.172 0.192 2.715 2.793 fft3d_ps 1111 14.6 0.777 0.884 2.711 2.765 potential_pw2rs 110 12.3 0.006 0.007 2.691 2.711 jit_kernel_multiply 13 15.9 2.605 2.700 2.605 2.700 mp_waitany 14300 13.8 1.868 2.530 1.868 2.530 mp_alltoall_d11v 2046 13.8 2.087 2.514 2.087 2.514 grid_collocate_task_list 110 9.6 2.084 2.297 2.084 2.297 acc_transpose_blocks 49320 15.4 0.227 0.239 2.187 2.259 mp_sum_d 3881 11.9 1.565 2.072 1.565 2.072 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.936 1.953 make_images_data 4110 15.4 0.042 0.045 1.783 1.931 cp_fm_cholesky_invert 11 10.9 1.893 1.897 1.893 1.897 hybrid_alltoall_any 4261 16.3 0.081 0.482 1.539 1.826 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.746 1.771 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.271000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.818182, yerr=3.128118 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 586.924032E+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 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 1325457. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 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 0.018 0.034 69.046 69.048 qs_mol_dyn_low 1 2.0 0.003 0.004 68.775 68.785 qs_forces 11 3.9 0.002 0.002 68.701 68.703 qs_energies 11 4.9 0.002 0.008 65.351 65.354 scf_env_do_scf 11 5.9 0.005 0.034 56.741 56.745 scf_env_do_scf_inner_loop 99 6.5 0.005 0.022 49.021 49.026 dbcsr_multiply_generic 2055 12.4 0.113 0.117 37.655 37.868 velocity_verlet 10 3.0 0.003 0.010 36.280 36.281 qs_scf_new_mos 99 7.5 0.003 0.023 32.546 32.716 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.543 32.715 multiply_cannon 2055 13.4 0.220 0.244 30.996 32.409 ot_scf_mini 99 9.5 0.003 0.004 30.888 31.045 multiply_cannon_loop 2055 14.4 0.922 0.949 29.680 30.703 ot_mini 99 10.5 0.001 0.001 18.123 18.294 multiply_cannon_multrec 24660 15.4 7.616 9.321 13.899 15.650 rebuild_ks_matrix 110 8.3 0.000 0.000 13.967 14.140 qs_ks_build_kohn_sham_matrix 110 9.3 0.016 0.042 13.967 14.140 qs_ot_get_derivative 99 11.5 0.001 0.001 12.357 12.514 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.313 12.470 mp_waitall_1 176588 16.5 7.550 10.009 7.550 10.009 multiply_cannon_sync_h2d 24660 15.4 6.986 8.185 6.986 8.185 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 5.132 7.807 init_scf_loop 11 6.9 0.001 0.009 7.677 7.678 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.468 7.255 apply_single 110 13.6 0.000 0.001 6.467 7.255 sum_up_and_integrate 110 10.3 0.051 0.058 6.688 6.714 integrate_v_rspace 110 11.3 0.002 0.003 6.636 6.664 dbcsr_mm_accdrv_process 52282 16.1 4.785 5.709 6.125 6.431 qs_ot_get_p 110 10.4 0.001 0.001 6.167 6.347 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.239 6.252 calculate_rho_elec 110 8.6 0.039 0.047 6.238 6.252 init_scf_run 11 5.9 0.000 0.001 6.217 6.218 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.217 6.218 ot_diis_step 99 11.5 0.010 0.010 5.718 5.718 prepare_preconditioner 11 7.9 0.000 0.000 5.535 5.554 make_preconditioner 11 8.9 0.000 0.001 5.535 5.554 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.679 5.362 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.124 5.279 make_m2s 4110 13.4 0.057 0.060 4.147 4.640 make_images 4110 14.4 0.396 0.441 4.038 4.528 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.237 4.258 pw_transfer 1331 11.6 0.066 0.075 3.802 3.956 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.695 3.852 density_rs2pw 110 9.6 0.004 0.004 3.367 3.828 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.807 3.808 wfi_extrapolate 11 7.9 0.001 0.001 3.529 3.530 rs_pw_transfer 902 11.9 0.012 0.013 2.991 3.497 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.354 3.355 grid_integrate_task_list 110 12.3 3.147 3.348 3.147 3.348 fft_wrap_pw1pw2_140 451 13.1 0.203 0.220 3.133 3.296 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.264 3.265 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.173 3.248 cp_fm_redistribute_end 48 14.0 2.440 3.237 2.442 3.237 fft3d_ps 1111 14.6 1.099 1.311 3.044 3.190 cp_fm_diag_elpa_base 48 14.0 0.761 3.096 0.792 3.177 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.007 3.064 calculate_dm_sparse 110 9.5 0.001 0.001 2.953 2.983 make_images_data 4110 15.4 0.046 0.049 2.312 2.812 hybrid_alltoall_any 4261 16.3 0.102 0.440 1.968 2.784 calculate_first_density_matrix 1 7.0 0.001 0.002 2.591 2.594 mp_sum_l 6514 12.8 1.891 2.579 1.891 2.579 cp_fm_cholesky_invert 11 10.9 2.570 2.577 2.570 2.577 potential_pw2rs 110 12.3 0.008 0.009 2.488 2.517 grid_collocate_task_list 110 9.6 2.055 2.478 2.055 2.478 mp_alltoall_d11v 2046 13.8 1.698 2.091 1.698 2.091 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.945 1.967 jit_kernel_multiply 9 16.4 0.989 1.899 0.989 1.899 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.871 1.871 multiply_cannon_metrocomm4 22605 15.4 0.075 0.079 0.800 1.842 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.823 1.840 mp_waitany 10164 13.8 1.219 1.753 1.219 1.753 mp_irecv_dv 57340 16.2 0.675 1.668 0.675 1.668 mp_allgather_i34 2055 14.4 0.647 1.657 0.647 1.657 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.532 1.544 acc_transpose_blocks 24660 15.4 0.112 0.115 1.512 1.531 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.370 1.477 rs_pw_transfer_RS2PW_140 121 11.5 0.205 0.215 0.967 1.460 dbcsr_complete_redistribute 325 12.2 0.244 0.306 1.143 1.406 mp_alltoall_z22v 1111 16.6 1.262 1.394 1.262 1.394 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.048000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.000000, yerr=7.931525 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 660.975616E+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 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 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 0.035 0.094 60.781 60.787 qs_mol_dyn_low 1 2.0 0.005 0.012 60.218 60.227 qs_forces 11 3.9 0.002 0.002 60.152 60.158 qs_energies 11 4.9 0.003 0.008 56.944 56.952 scf_env_do_scf 11 5.9 0.001 0.002 49.084 49.084 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.350 40.350 velocity_verlet 10 3.0 0.001 0.002 33.040 33.041 dbcsr_multiply_generic 2055 12.4 0.108 0.110 28.534 28.772 qs_scf_new_mos 99 7.5 0.001 0.001 25.066 25.135 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.066 25.134 ot_scf_mini 99 9.5 0.003 0.003 23.844 23.937 multiply_cannon 2055 13.4 0.211 0.224 22.160 23.411 multiply_cannon_loop 2055 14.4 0.611 0.625 20.913 21.880 ot_mini 99 10.5 0.001 0.001 13.727 13.820 rebuild_ks_matrix 110 8.3 0.000 0.000 12.520 12.633 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 12.519 12.633 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.029 11.133 multiply_cannon_multrec 16440 15.4 3.947 4.881 9.686 10.593 mp_waitall_1 139946 16.5 6.956 9.683 6.956 9.683 qs_ot_get_derivative 99 11.5 0.001 0.001 9.219 9.311 init_scf_loop 11 6.9 0.001 0.004 8.697 8.699 prepare_preconditioner 11 7.9 0.000 0.000 6.909 6.922 make_preconditioner 11 8.9 0.000 0.002 6.908 6.922 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.244 6.909 sum_up_and_integrate 110 10.3 0.059 0.060 6.635 6.653 make_full_inverse_cholesky 11 9.9 0.000 0.001 6.259 6.611 integrate_v_rspace 110 11.3 0.003 0.003 6.576 6.593 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.099 6.110 calculate_rho_elec 110 8.6 0.058 0.059 6.098 6.109 dbcsr_mm_accdrv_process 34862 16.1 4.656 5.348 5.594 5.844 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.134 5.616 apply_single 110 13.6 0.000 0.000 5.134 5.616 init_scf_run 11 5.9 0.000 0.001 5.432 5.432 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.432 5.432 qs_ot_get_p 110 10.4 0.001 0.001 5.307 5.417 make_m2s 4110 13.4 0.050 0.051 4.187 4.599 make_images 4110 14.4 0.389 0.505 4.072 4.484 ot_diis_step 99 11.5 0.010 0.011 4.478 4.478 density_rs2pw 110 9.6 0.004 0.005 3.191 4.403 multiply_cannon_sync_h2d 16440 15.4 3.783 4.304 3.783 4.304 rs_pw_transfer 902 11.9 0.010 0.012 2.795 3.995 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.152 3.886 pw_transfer 1331 11.6 0.065 0.072 3.727 3.737 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.682 3.687 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.620 3.633 grid_integrate_task_list 110 12.3 3.189 3.375 3.189 3.375 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.335 3.336 fft_wrap_pw1pw2_140 451 13.1 0.213 0.217 3.120 3.137 wfi_extrapolate 11 7.9 0.001 0.002 2.985 2.985 hybrid_alltoall_any 4261 16.3 0.105 0.374 2.236 2.969 make_images_data 4110 15.4 0.043 0.048 2.453 2.949 fft3d_ps 1111 14.6 1.087 1.097 2.920 2.931 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.788 2.791 cp_fm_redistribute_end 48 14.0 1.741 2.763 1.744 2.764 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.755 2.756 cp_fm_diag_elpa_base 48 14.0 0.957 2.610 1.015 2.713 cp_fm_cholesky_invert 11 10.9 2.690 2.696 2.690 2.696 calculate_dm_sparse 110 9.5 0.001 0.001 2.491 2.528 mp_waitany 17072 13.8 1.253 2.508 1.253 2.508 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.435 2.483 grid_collocate_task_list 110 9.6 2.077 2.468 2.077 2.468 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.339 2.392 potential_pw2rs 110 12.3 0.010 0.010 2.372 2.382 calculate_first_density_matrix 1 7.0 0.001 0.003 2.360 2.361 mp_sum_l 6514 12.8 1.672 2.330 1.672 2.330 multiply_cannon_metrocomm4 14385 15.4 0.045 0.050 0.870 2.298 mp_alltoall_d11v 2046 13.8 1.815 2.284 1.815 2.284 mp_irecv_dv 48980 15.7 0.800 2.175 0.800 2.175 rs_pw_transfer_RS2PW_140 121 11.5 0.178 0.197 0.920 2.109 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.979 1.986 dbcsr_complete_redistribute 325 12.2 0.323 0.354 1.472 1.948 jit_kernel_multiply 10 16.6 0.552 1.895 0.552 1.895 cp_fm_upper_to_full 70 13.6 1.413 1.891 1.413 1.891 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.692 1.705 mp_allgather_i34 2055 14.4 0.573 1.676 0.573 1.676 cp_fm_cholesky_decompose 22 10.9 1.610 1.631 1.610 1.631 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.374 1.472 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.452 1.466 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.952 1.411 rs_gather_matrices 110 12.3 0.232 0.260 0.935 1.361 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.262 1.270 acc_transpose_blocks 16440 15.4 0.072 0.074 1.220 1.244 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.787000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=626.000000, yerr=9.467648 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 729.149440E+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 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 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 0.025 0.061 65.442 65.444 qs_mol_dyn_low 1 2.0 0.003 0.004 65.005 65.014 qs_forces 11 3.9 0.003 0.007 64.941 64.943 qs_energies 11 4.9 0.005 0.022 61.525 61.530 scf_env_do_scf 11 5.9 0.001 0.001 53.223 53.226 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 41.390 41.391 velocity_verlet 10 3.0 0.002 0.002 36.986 36.988 dbcsr_multiply_generic 2055 12.4 0.114 0.117 29.271 29.520 qs_scf_new_mos 99 7.5 0.001 0.001 26.307 26.403 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.306 26.402 ot_scf_mini 99 9.5 0.003 0.004 24.681 24.786 multiply_cannon 2055 13.4 0.241 0.261 22.344 23.278 multiply_cannon_loop 2055 14.4 0.878 0.904 20.984 21.645 ot_mini 99 10.5 0.001 0.001 14.046 14.159 multiply_cannon_multrec 24660 15.4 4.234 6.828 12.672 13.816 rebuild_ks_matrix 110 8.3 0.000 0.000 12.249 12.359 qs_ks_build_kohn_sham_matrix 110 9.3 0.020 0.074 12.249 12.359 init_scf_loop 11 6.9 0.009 0.040 11.791 11.792 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.831 10.927 prepare_preconditioner 11 7.9 0.002 0.013 10.002 10.030 make_preconditioner 11 8.9 0.000 0.002 10.000 10.029 qs_ot_get_derivative 99 11.5 0.001 0.001 9.889 9.992 make_full_inverse_cholesky 11 9.9 0.000 0.001 8.230 9.704 dbcsr_mm_accdrv_process 52304 16.0 6.934 8.715 8.292 9.344 sum_up_and_integrate 110 10.3 0.066 0.069 6.543 6.556 integrate_v_rspace 110 11.3 0.003 0.003 6.476 6.491 mp_waitall_1 121746 16.5 4.174 6.262 4.174 6.262 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.125 6.137 calculate_rho_elec 110 8.6 0.077 0.081 6.124 6.137 qs_ot_get_p 110 10.4 0.001 0.001 5.685 5.850 make_m2s 4110 13.4 0.059 0.061 5.293 5.581 init_scf_run 11 5.9 0.001 0.003 5.461 5.462 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.461 5.462 make_images 4110 14.4 0.571 0.692 5.152 5.436 cp_fm_upper_to_full 70 13.8 3.274 4.654 3.274 4.654 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.013 4.123 apply_single 110 13.6 0.000 0.000 4.013 4.123 ot_diis_step 99 11.5 0.011 0.012 4.118 4.118 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.914 3.928 pw_transfer 1331 11.6 0.065 0.073 3.820 3.847 dbcsr_complete_redistribute 325 12.2 0.420 0.465 2.660 3.791 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.714 3.745 density_rs2pw 110 9.6 0.004 0.004 3.073 3.583 grid_integrate_task_list 110 12.3 3.243 3.446 3.243 3.446 cp_dbcsr_syevd 48 12.0 0.003 0.005 3.436 3.436 multiply_cannon_sync_h2d 24660 15.4 3.194 3.406 3.194 3.406 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.262 3.312 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.125 3.246 multiply_cannon_metrocomm3 24660 15.4 0.035 0.037 1.409 3.225 fft_wrap_pw1pw2_140 451 13.1 0.203 0.212 3.172 3.207 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.022 3.077 fft3d_ps 1111 14.6 1.090 1.124 3.006 3.031 rs_pw_transfer 902 11.9 0.010 0.011 2.449 2.980 make_images_data 4110 15.4 0.046 0.049 2.639 2.968 wfi_extrapolate 11 7.9 0.001 0.002 2.966 2.966 calculate_dm_sparse 110 9.5 0.001 0.001 2.917 2.943 hybrid_alltoall_any 4261 16.3 0.119 0.458 2.338 2.926 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.760 2.864 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.819 2.821 mp_alltoall_i22 605 13.7 1.648 2.817 1.648 2.817 cp_fm_redistribute_end 48 14.0 1.407 2.785 1.409 2.787 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.775 2.777 cp_fm_cholesky_invert 11 10.9 2.767 2.774 2.767 2.774 cp_fm_diag_elpa_base 48 14.0 1.297 2.651 1.375 2.758 grid_collocate_task_list 110 9.6 2.181 2.460 2.181 2.460 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.420 2.458 calculate_first_density_matrix 1 7.0 0.001 0.004 2.402 2.406 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.293 2.295 potential_pw2rs 110 12.3 0.013 0.013 2.224 2.232 mp_alltoall_d11v 2046 13.8 1.770 2.121 1.770 2.121 jit_kernel_multiply 10 15.8 1.030 1.939 1.030 1.939 cp_fm_cholesky_decompose 22 10.9 1.695 1.735 1.695 1.735 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.697 1.725 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.601 1.701 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.668 1.680 mp_waitany 13376 13.8 1.106 1.591 1.106 1.591 multiply_cannon_metrocomm4 20550 15.4 0.058 0.061 0.839 1.584 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.529 1.546 acc_transpose_blocks 24660 15.4 0.106 0.108 1.508 1.541 mp_irecv_dv 62702 16.1 0.740 1.503 0.740 1.503 mp_sum_l 6514 12.8 0.953 1.459 0.953 1.459 rs_pw_transfer_RS2PW_140 121 11.5 0.170 0.178 0.836 1.370 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.240 1.338 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.444000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=690.090909, yerr=7.913792 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 831.881216E+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 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 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 0.016 0.032 56.077 56.079 qs_mol_dyn_low 1 2.0 0.003 0.003 55.792 55.801 qs_forces 11 3.9 0.002 0.002 55.728 55.729 qs_energies 11 4.9 0.001 0.001 51.996 52.003 scf_env_do_scf 11 5.9 0.001 0.001 43.748 43.748 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.872 35.873 velocity_verlet 10 3.0 0.002 0.002 31.520 31.523 dbcsr_multiply_generic 2055 12.4 0.104 0.106 22.656 22.779 qs_scf_new_mos 99 7.5 0.001 0.001 20.481 20.543 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.480 20.542 ot_scf_mini 99 9.5 0.002 0.002 19.240 19.267 multiply_cannon 2055 13.4 0.246 0.265 17.195 18.393 multiply_cannon_loop 2055 14.4 0.318 0.330 15.865 16.127 rebuild_ks_matrix 110 8.3 0.000 0.000 12.183 12.208 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.183 12.208 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.829 10.850 ot_mini 99 10.5 0.001 0.001 10.289 10.312 multiply_cannon_multrec 8220 15.4 3.216 4.455 7.425 8.306 init_scf_loop 11 6.9 0.000 0.000 7.826 7.827 mp_waitall_1 103326 16.6 6.033 7.633 6.033 7.633 sum_up_and_integrate 110 10.3 0.079 0.080 6.658 6.673 integrate_v_rspace 110 11.3 0.003 0.003 6.579 6.593 qs_ot_get_derivative 99 11.5 0.001 0.001 6.526 6.552 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.453 6.469 calculate_rho_elec 110 8.6 0.115 0.116 6.453 6.468 prepare_preconditioner 11 7.9 0.000 0.000 6.103 6.108 make_preconditioner 11 8.9 0.000 0.000 6.103 6.108 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.677 5.755 init_scf_run 11 5.9 0.000 0.001 5.092 5.092 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.092 5.092 qs_ot_get_p 110 10.4 0.001 0.001 5.032 5.057 dbcsr_mm_accdrv_process 17442 15.9 2.831 3.764 4.080 4.973 make_m2s 4110 13.4 0.038 0.040 4.226 4.487 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.962 4.410 make_images 4110 14.4 0.632 0.687 4.097 4.359 pw_transfer 1331 11.6 0.066 0.071 4.214 4.225 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.107 4.121 ot_diis_step 99 11.5 0.012 0.012 3.743 3.743 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.685 3.689 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.664 3.676 apply_single 110 13.6 0.000 0.000 3.664 3.676 fft_wrap_pw1pw2_140 451 13.1 0.216 0.220 3.553 3.575 density_rs2pw 110 9.6 0.004 0.004 3.197 3.571 grid_integrate_task_list 110 12.3 3.348 3.531 3.348 3.531 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.385 3.385 fft3d_ps 1111 14.6 1.139 1.150 3.351 3.361 multiply_cannon_sync_h2d 8220 15.4 2.925 2.995 2.925 2.995 cp_fm_cholesky_invert 11 10.9 2.946 2.950 2.946 2.950 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.856 2.857 cp_fm_redistribute_end 48 14.0 0.729 2.825 0.734 2.826 make_images_data 4110 15.4 0.038 0.043 2.411 2.805 hybrid_alltoall_any 4261 16.3 0.199 0.861 2.358 2.798 cp_fm_diag_elpa_base 48 14.0 1.909 2.627 2.084 2.788 wfi_extrapolate 11 7.9 0.001 0.001 2.775 2.775 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.701 2.702 rs_pw_transfer 902 11.9 0.010 0.010 2.251 2.641 grid_collocate_task_list 110 9.6 2.278 2.539 2.278 2.539 calculate_dm_sparse 110 9.5 0.001 0.001 2.464 2.509 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.498 2.499 calculate_first_density_matrix 1 7.0 0.001 0.001 2.221 2.222 potential_pw2rs 110 12.3 0.015 0.016 2.195 2.204 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.140 2.158 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.774 2.000 mp_alltoall_d11v 2046 13.8 1.672 1.994 1.672 1.994 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.906 1.918 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.783 1.798 cp_fm_cholesky_decompose 22 10.9 1.698 1.715 1.698 1.715 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.690 1.695 qs_env_update_s_mstruct 11 6.9 0.005 0.035 1.526 1.656 mp_allgather_i34 2055 14.4 0.516 1.616 0.516 1.616 dbcsr_complete_redistribute 325 12.2 0.559 0.582 1.448 1.543 mp_waitany 9240 13.8 1.066 1.493 1.066 1.493 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.427 1.438 qs_create_task_list 11 7.9 0.001 0.001 1.226 1.332 generate_qs_task_list 11 8.9 0.379 0.447 1.226 1.331 multiply_cannon_metrocomm1 8220 15.4 0.020 0.021 0.785 1.311 jit_kernel_multiply 7 15.6 0.944 1.266 0.944 1.266 rs_gather_matrices 110 12.3 0.324 0.365 0.944 1.237 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.166 0.786 1.197 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.154 1.181 mp_alltoall_z22v 1111 16.6 1.102 1.125 1.102 1.125 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.079000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=784.272727, yerr=12.388344 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 1.357263E+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 0.034 0.053 88.606 88.608 qs_mol_dyn_low 1 2.0 0.003 0.003 88.237 88.246 qs_forces 11 3.9 0.002 0.003 88.173 88.174 qs_energies 11 4.9 0.001 0.001 83.983 83.986 scf_env_do_scf 11 5.9 0.001 0.001 73.947 73.947 velocity_verlet 10 3.0 0.002 0.002 56.849 56.855 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 44.687 44.688 init_scf_loop 11 6.9 0.000 0.000 29.187 29.189 dbcsr_multiply_generic 2055 12.4 0.119 0.120 28.340 28.425 prepare_preconditioner 11 7.9 0.000 0.000 27.111 27.127 make_preconditioner 11 8.9 0.000 0.000 27.110 27.127 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.106 26.577 qs_scf_new_mos 99 7.5 0.001 0.001 26.198 26.256 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.197 26.255 ot_scf_mini 99 9.5 0.002 0.002 24.444 24.481 multiply_cannon 2055 13.4 0.349 0.386 21.441 22.201 multiply_cannon_loop 2055 14.4 0.340 0.343 19.630 19.931 cp_fm_upper_to_full 70 14.2 13.211 18.937 13.211 18.937 rebuild_ks_matrix 110 8.3 0.000 0.001 14.329 14.372 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.329 14.371 ot_mini 99 10.5 0.001 0.001 13.505 13.543 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.991 13.029 dbcsr_complete_redistribute 325 12.2 1.020 1.046 7.602 11.000 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.576 9.962 multiply_cannon_multrec 8220 15.4 4.369 4.547 9.514 9.643 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.990 9.346 mp_alltoall_i22 605 13.7 5.629 9.037 5.629 9.037 qs_ot_get_derivative 99 11.5 0.001 0.001 8.968 9.006 mp_waitall_1 84994 16.7 7.481 8.393 7.481 8.393 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.789 7.827 calculate_rho_elec 110 8.6 0.227 0.227 7.788 7.827 sum_up_and_integrate 110 10.3 0.150 0.152 7.448 7.461 integrate_v_rspace 110 11.3 0.003 0.004 7.298 7.312 qs_ot_get_p 110 10.4 0.001 0.001 5.773 5.824 init_scf_run 11 5.9 0.000 0.001 5.731 5.731 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.731 5.731 make_m2s 4110 13.4 0.043 0.043 5.168 5.662 make_images 4110 14.4 0.879 0.934 4.980 5.473 pw_transfer 1331 11.6 0.075 0.075 5.352 5.363 cp_fm_cholesky_invert 11 10.9 5.352 5.357 5.352 5.357 dbcsr_mm_accdrv_process 11614 15.7 3.281 3.663 5.004 5.280 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.236 5.248 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.649 5.124 apply_single 110 13.6 0.000 0.000 4.649 5.124 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.777 5.090 fft_wrap_pw1pw2_140 451 13.1 0.221 0.222 4.554 4.566 ot_diis_step 99 11.5 0.015 0.016 4.515 4.515 fft3d_ps 1111 14.6 1.294 1.304 4.432 4.444 qs_ot_p2m_diag 48 11.0 0.150 0.155 4.221 4.228 multiply_cannon_sync_h2d 8220 15.4 3.952 3.960 3.952 3.960 density_rs2pw 110 9.6 0.004 0.004 3.823 3.858 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.782 3.782 grid_integrate_task_list 110 12.3 3.662 3.717 3.662 3.717 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.706 3.707 hybrid_alltoall_any 4261 16.3 0.256 0.551 2.883 3.516 make_images_data 4110 15.4 0.041 0.045 2.796 3.409 wfi_extrapolate 11 7.9 0.001 0.001 3.333 3.333 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.877 3.328 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.167 3.168 cp_fm_diag_elpa_base 48 14.0 2.613 2.819 3.165 3.166 calculate_dm_sparse 110 9.5 0.001 0.001 3.115 3.135 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.863 2.866 grid_collocate_task_list 110 9.6 2.629 2.652 2.629 2.652 potential_pw2rs 110 12.3 0.021 0.022 2.581 2.584 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.384 2.407 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.360 2.375 rs_pw_transfer 902 11.9 0.011 0.011 2.312 2.356 qs_env_update_s_mstruct 11 6.9 0.002 0.002 2.225 2.286 calculate_first_density_matrix 1 7.0 0.001 0.001 2.277 2.278 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.101 2.190 mp_alltoall_d11v 2046 13.8 1.952 2.056 1.952 2.056 cp_fm_cholesky_decompose 22 10.9 2.042 2.054 2.042 2.054 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.943 1.947 qs_create_task_list 11 7.9 0.000 0.001 1.896 1.944 generate_qs_task_list 11 8.9 0.735 0.789 1.895 1.943 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.769 1.800 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.608000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1230.272727, yerr=46.892703 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 632.115200E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 56760. MP_Allreduce 11253 785. MP_Sync 170 MP_Alltoall 2226 2520936. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_comm_split 83 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 0.012 0.031 205.118 205.119 qs_mol_dyn_low 1 2.0 0.003 0.004 204.755 204.768 qs_forces 11 3.9 0.003 0.004 204.664 204.667 qs_energies 11 4.9 0.001 0.002 199.074 199.088 scf_env_do_scf 11 5.9 0.001 0.001 182.405 182.409 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 161.669 161.671 dbcsr_multiply_generic 2507 12.6 0.177 0.181 124.453 125.197 velocity_verlet 10 3.0 0.001 0.002 123.342 123.343 qs_scf_new_mos 117 7.6 0.001 0.001 121.111 121.370 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.111 121.369 ot_scf_mini 117 9.6 0.003 0.003 114.517 114.785 multiply_cannon 2507 13.6 0.237 0.246 100.933 103.608 multiply_cannon_loop 2507 14.6 2.094 2.154 98.703 101.058 ot_mini 117 10.6 0.001 0.001 65.960 66.240 multiply_cannon_multrec 60168 15.6 33.195 35.172 41.675 43.164 qs_ot_get_derivative 117 11.6 0.001 0.001 41.154 41.420 rebuild_ks_matrix 128 8.3 0.001 0.001 34.004 34.285 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 34.003 34.284 mp_waitall_1 267128 16.5 28.397 31.084 28.397 31.084 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.575 30.858 multiply_cannon_sync_h2d 60168 15.6 27.267 29.070 27.267 29.070 qs_ot_get_p 128 10.4 0.001 0.001 26.699 26.983 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.330 24.937 apply_single 128 13.6 0.001 0.001 24.330 24.936 ot_diis_step 117 11.6 0.008 0.008 24.473 24.474 init_scf_loop 11 6.9 0.000 0.001 20.658 20.659 qs_ot_p2m_diag 83 11.4 0.077 0.091 19.961 20.065 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.406 19.602 multiply_cannon_metrocomm3 60168 15.6 0.114 0.119 15.690 18.244 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.427 17.428 prepare_preconditioner 11 7.9 0.000 0.000 16.017 16.053 make_preconditioner 11 8.9 0.000 0.000 16.017 16.053 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.249 15.430 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.551 14.556 sum_up_and_integrate 128 10.3 0.089 0.107 14.460 14.474 cp_fm_redistribute_end 83 14.4 11.474 14.440 11.486 14.442 integrate_v_rspace 128 11.3 0.003 0.004 14.370 14.387 make_m2s 5014 13.6 0.104 0.112 13.841 14.184 cp_fm_diag_elpa_base 83 14.4 2.915 14.068 2.945 14.179 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.009 14.136 calculate_rho_elec 128 8.7 0.045 0.064 14.009 14.135 make_images 5014 14.6 0.398 0.416 13.660 14.016 init_scf_run 11 5.9 0.000 0.001 12.481 12.481 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.481 12.481 density_rs2pw 128 9.7 0.007 0.008 7.377 10.699 mp_sum_l 7870 13.0 8.411 9.705 8.411 9.705 cp_fm_cholesky_invert 11 10.9 9.177 9.185 9.177 9.185 rs_pw_transfer 1046 11.9 0.016 0.019 5.895 9.154 wfi_extrapolate 11 7.9 0.001 0.001 9.104 9.104 dbcsr_mm_accdrv_process 124484 16.2 3.290 3.434 8.040 8.616 calculate_dm_sparse 128 9.5 0.001 0.001 8.494 8.600 pw_transfer 1547 11.6 0.075 0.093 8.230 8.494 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 8.026 8.294 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.947 8.084 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.848 7.967 make_images_data 5014 15.6 0.066 0.071 6.842 7.766 multiply_cannon_metrocomm1 60168 15.6 0.086 0.091 6.055 7.610 grid_integrate_task_list 128 12.3 7.052 7.595 7.052 7.595 hybrid_alltoall_any 5200 16.5 0.290 2.256 5.994 7.299 fft_wrap_pw1pw2_140 523 13.2 0.444 0.517 6.922 7.213 fft3d_ps 1291 14.7 2.102 2.898 6.814 7.013 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.735 6.742 mp_waitany 16020 13.9 2.635 5.960 2.635 5.960 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.843 5.921 grid_collocate_task_list 128 9.7 4.571 5.800 4.571 5.800 mp_alltoall_d11v 2415 14.1 4.216 5.608 4.216 5.608 rs_pw_transfer_RS2PW_140 139 11.5 0.278 0.296 2.114 5.432 potential_pw2rs 128 12.3 0.009 0.011 4.842 4.873 cp_fm_cholesky_decompose 22 10.9 4.642 4.655 4.642 4.655 mp_sum_d 4462 12.2 3.528 4.391 3.528 4.391 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.119000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.636364, yerr=7.327070 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/16/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 2.183246E+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 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 827.269120E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57341. MP_Allreduce 11227 947. MP_Sync 170 MP_Alltoall 1969 6397111. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.033 193.209 193.211 qs_mol_dyn_low 1 2.0 0.003 0.004 192.803 192.815 qs_forces 11 3.9 0.010 0.056 192.721 192.722 qs_energies 11 4.9 0.003 0.013 185.814 185.830 scf_env_do_scf 11 5.9 0.001 0.003 169.123 169.133 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 136.062 136.066 velocity_verlet 10 3.0 0.001 0.002 121.304 121.305 dbcsr_multiply_generic 2507 12.6 0.186 0.191 97.630 98.739 qs_scf_new_mos 117 7.6 0.001 0.001 95.168 95.601 qs_scf_loop_do_ot 117 8.6 0.001 0.002 95.168 95.600 ot_scf_mini 117 9.6 0.004 0.005 90.367 90.934 multiply_cannon 2507 13.6 0.477 0.529 77.502 81.877 multiply_cannon_loop 2507 14.6 1.248 1.294 74.121 76.290 ot_mini 117 10.6 0.001 0.001 49.885 50.370 mp_waitall_1 214728 16.6 24.446 38.836 24.446 38.836 multiply_cannon_multrec 30084 15.6 22.132 26.573 31.753 36.629 rebuild_ks_matrix 128 8.3 0.001 0.001 33.867 34.385 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.022 33.867 34.384 init_scf_loop 11 6.9 0.001 0.006 32.965 32.971 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.433 30.908 qs_ot_get_derivative 117 11.6 0.001 0.002 27.997 28.540 prepare_preconditioner 11 7.9 0.000 0.000 28.455 28.526 make_preconditioner 11 8.9 0.001 0.010 28.455 28.526 multiply_cannon_metrocomm3 30084 15.6 0.095 0.102 15.567 28.009 make_full_inverse_cholesky 11 9.9 0.000 0.001 27.158 27.678 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.975 23.040 apply_single 128 13.6 0.001 0.001 21.975 23.040 qs_ot_get_p 128 10.4 0.001 0.002 22.202 22.767 multiply_cannon_sync_h2d 30084 15.6 19.385 21.950 19.385 21.950 ot_diis_step 117 11.6 0.014 0.015 21.711 21.712 qs_ot_p2m_diag 83 11.4 0.187 0.215 17.334 17.365 cp_fm_cholesky_invert 11 10.9 16.657 16.670 16.657 16.670 cp_dbcsr_syevd 83 12.4 0.006 0.007 16.191 16.192 make_m2s 5014 13.6 0.090 0.097 14.408 16.105 make_images 5014 14.6 1.140 1.323 14.198 15.891 sum_up_and_integrate 128 10.3 0.114 0.131 15.037 15.070 integrate_v_rspace 128 11.3 0.003 0.004 14.923 14.962 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.521 14.560 calculate_rho_elec 128 8.7 0.088 0.105 14.520 14.559 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.906 12.913 cp_fm_redistribute_end 83 14.4 7.530 12.830 7.542 12.832 cp_fm_diag_elpa_base 83 14.4 5.053 12.336 5.272 12.692 init_scf_run 11 5.9 0.000 0.001 11.793 11.795 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.793 11.795 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.022 11.429 make_images_data 5014 15.6 0.063 0.070 8.722 10.822 hybrid_alltoall_any 5200 16.5 0.341 1.509 7.603 10.415 multiply_cannon_metrocomm4 27577 15.6 0.098 0.111 3.728 10.292 density_rs2pw 128 9.7 0.006 0.007 7.742 10.125 mp_irecv_dv 69486 16.3 3.533 9.911 3.533 9.911 dbcsr_mm_accdrv_process 62242 16.2 4.549 5.343 9.078 9.681 pw_transfer 1547 11.6 0.087 0.107 9.278 9.360 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.052 9.138 wfi_extrapolate 11 7.9 0.001 0.002 8.525 8.526 rs_pw_transfer 1046 11.9 0.014 0.016 5.922 8.336 fft_wrap_pw1pw2_140 523 13.2 0.472 0.531 7.934 8.053 grid_integrate_task_list 128 12.3 7.170 7.498 7.170 7.498 fft3d_ps 1291 14.7 2.748 2.915 7.412 7.468 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.283 7.076 cp_fm_cholesky_decompose 22 10.9 6.950 7.018 6.950 7.018 calculate_dm_sparse 128 9.5 0.001 0.001 6.468 6.631 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.321 6.330 grid_collocate_task_list 128 9.7 4.705 5.877 4.705 5.877 mp_sum_l 7870 13.0 3.940 5.760 3.940 5.760 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.369 5.526 mp_alltoall_d11v 2415 14.1 4.658 5.385 4.658 5.385 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.316 5.378 potential_pw2rs 128 12.3 0.015 0.017 5.266 5.286 mp_allgather_i34 2507 14.6 1.950 5.196 1.950 5.196 mp_waitany 11748 13.9 2.515 4.852 2.515 4.852 rs_pw_transfer_RS2PW_140 139 11.5 0.352 0.375 2.114 4.509 dbcsr_complete_redistribute 395 12.7 0.809 0.885 3.148 3.993 mp_sum_d 4457 12.1 2.670 3.992 2.670 3.992 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.844 3.888 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=193.211000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=787.636364, yerr=4.006194 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+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 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 934.363136E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+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 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57197. MP_Allreduce 11251 985. MP_Sync 170 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.096 178.317 178.318 qs_mol_dyn_low 1 2.0 0.003 0.004 177.812 177.825 qs_forces 11 3.9 0.003 0.003 177.711 177.718 qs_energies 11 4.9 0.002 0.007 171.044 171.057 scf_env_do_scf 11 5.9 0.001 0.002 155.491 155.491 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 120.156 120.159 velocity_verlet 10 3.0 0.001 0.002 114.229 114.231 dbcsr_multiply_generic 2507 12.6 0.182 0.187 81.327 82.380 qs_scf_new_mos 117 7.6 0.001 0.001 81.772 82.083 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.771 82.082 ot_scf_mini 117 9.6 0.004 0.004 77.646 77.982 multiply_cannon 2507 13.6 0.500 0.524 61.533 66.258 multiply_cannon_loop 2507 14.6 0.850 0.878 58.367 60.741 ot_mini 117 10.6 0.001 0.001 42.560 42.889 init_scf_loop 11 6.9 0.001 0.003 35.237 35.239 mp_waitall_1 170520 16.6 24.725 34.100 24.725 34.100 prepare_preconditioner 11 7.9 0.000 0.000 31.191 31.248 make_preconditioner 11 8.9 0.000 0.000 31.191 31.248 rebuild_ks_matrix 128 8.3 0.001 0.001 30.820 31.232 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.020 30.819 31.231 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.892 30.234 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.759 28.127 multiply_cannon_multrec 20056 15.6 13.505 16.297 22.135 24.913 multiply_cannon_metrocomm3 20056 15.6 0.059 0.063 14.934 24.080 qs_ot_get_derivative 117 11.6 0.002 0.002 22.836 23.171 qs_ot_get_p 128 10.4 0.001 0.001 20.501 20.974 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.854 20.869 apply_single 128 13.6 0.001 0.001 19.853 20.869 ot_diis_step 117 11.6 0.018 0.018 19.625 19.626 qs_ot_p2m_diag 83 11.4 0.265 0.272 16.237 16.246 multiply_cannon_sync_h2d 20056 15.6 14.382 15.723 14.382 15.723 make_m2s 5014 13.6 0.082 0.088 14.740 15.712 make_images 5014 14.6 1.171 1.259 14.506 15.474 cp_dbcsr_syevd 83 12.4 0.006 0.012 15.222 15.223 sum_up_and_integrate 128 10.3 0.131 0.141 14.733 14.763 cp_fm_cholesky_invert 11 10.9 14.635 14.644 14.635 14.644 integrate_v_rspace 128 11.3 0.003 0.004 14.601 14.626 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.404 14.435 calculate_rho_elec 128 8.7 0.132 0.147 14.403 14.434 cp_fm_diag_elpa 83 13.4 0.001 0.001 11.959 11.961 cp_fm_redistribute_end 83 14.4 4.519 11.879 4.537 11.881 cp_fm_diag_elpa_base 83 14.4 6.915 11.280 7.320 11.739 init_scf_run 11 5.9 0.000 0.001 10.504 10.505 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.504 10.505 make_images_data 5014 15.6 0.060 0.067 9.103 10.443 hybrid_alltoall_any 5200 16.5 0.434 1.993 7.904 9.579 density_rs2pw 128 9.7 0.006 0.006 7.302 9.344 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.973 9.223 pw_transfer 1547 11.6 0.086 0.108 9.010 9.134 multiply_cannon_metrocomm4 17549 15.6 0.062 0.072 3.448 9.012 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.784 8.913 mp_irecv_dv 50230 16.2 3.326 8.767 3.326 8.767 dbcsr_mm_accdrv_process 41502 16.2 4.356 5.199 8.078 8.196 grid_integrate_task_list 128 12.3 7.304 7.906 7.304 7.906 fft_wrap_pw1pw2_140 523 13.2 0.478 0.521 7.723 7.858 cp_fm_cholesky_decompose 22 10.9 7.513 7.548 7.513 7.548 cp_fm_upper_to_full 105 14.5 5.727 7.489 5.727 7.489 wfi_extrapolate 11 7.9 0.001 0.002 7.468 7.468 rs_pw_transfer 1046 11.9 0.013 0.014 5.290 7.316 fft3d_ps 1291 14.7 2.667 2.893 7.045 7.128 dbcsr_complete_redistribute 395 12.7 1.175 1.210 4.541 6.317 calculate_dm_sparse 128 9.5 0.001 0.001 5.747 5.872 grid_collocate_task_list 128 9.7 4.901 5.788 4.901 5.788 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.423 5.432 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.569 5.269 copy_fm_to_dbcsr 209 11.7 0.002 0.003 3.351 5.123 potential_pw2rs 128 12.3 0.020 0.023 4.867 4.880 mp_alltoall_d11v 2415 14.1 4.208 4.765 4.208 4.765 mp_allgather_i34 2507 14.6 1.690 4.680 1.690 4.680 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.529 4.676 mp_sum_l 7870 13.0 3.241 4.495 3.241 4.495 mp_waitany 11748 13.9 2.355 4.442 2.355 4.442 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.055 4.087 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.280 4.003 rs_pw_transfer_RS2PW_140 139 11.5 0.330 0.349 1.936 3.962 mp_alltoall_i22 716 14.1 1.912 3.867 1.912 3.867 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.784 3.785 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.575 3.619 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.318000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=886.727273, yerr=10.287600 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 1.151082E+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 4103 56948. MP_Allreduce 11296 1066. MP_Sync 170 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_comm_split 83 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 0.024 0.069 189.376 189.377 qs_mol_dyn_low 1 2.0 0.003 0.004 188.915 188.928 qs_forces 11 3.9 0.003 0.009 188.805 188.809 qs_energies 11 4.9 0.002 0.006 181.548 181.560 scf_env_do_scf 11 5.9 0.001 0.003 164.509 164.517 velocity_verlet 10 3.0 0.005 0.006 125.234 125.236 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 117.915 117.916 qs_scf_new_mos 117 7.6 0.001 0.001 79.934 80.256 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.933 80.255 dbcsr_multiply_generic 2507 12.6 0.189 0.195 79.351 80.060 ot_scf_mini 117 9.6 0.003 0.004 75.433 75.709 multiply_cannon 2507 13.6 0.555 0.579 54.626 56.753 multiply_cannon_loop 2507 14.6 1.172 1.206 50.849 52.517 init_scf_loop 11 6.9 0.007 0.056 46.471 46.472 ot_mini 117 10.6 0.001 0.001 42.214 42.489 prepare_preconditioner 11 7.9 0.000 0.002 42.200 42.240 make_preconditioner 11 8.9 0.001 0.009 42.199 42.240 make_full_inverse_cholesky 11 9.9 0.000 0.001 35.998 40.831 rebuild_ks_matrix 128 8.3 0.001 0.001 30.391 30.657 qs_ks_build_kohn_sham_matrix 128 9.3 0.022 0.061 30.390 30.656 multiply_cannon_multrec 30084 15.6 14.154 19.301 26.000 30.556 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.359 27.592 mp_waitall_1 147882 16.7 17.240 26.680 17.240 26.680 qs_ot_get_derivative 117 11.6 0.002 0.002 22.607 22.877 make_m2s 5014 13.6 0.096 0.102 20.362 21.311 make_images 5014 14.6 1.934 2.277 20.057 21.007 qs_ot_get_p 128 10.4 0.001 0.001 19.447 19.750 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.984 19.606 apply_single 128 13.6 0.001 0.001 18.984 19.606 ot_diis_step 117 11.6 0.017 0.019 19.471 19.472 cp_fm_cholesky_invert 11 10.9 16.301 16.309 16.301 16.309 cp_fm_upper_to_full 105 14.7 10.848 15.988 10.848 15.988 qs_ot_p2m_diag 83 11.4 0.342 0.389 15.386 15.440 sum_up_and_integrate 128 10.3 0.141 0.154 15.060 15.085 integrate_v_rspace 128 11.3 0.004 0.009 14.919 14.950 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.798 14.831 calculate_rho_elec 128 8.7 0.175 0.190 14.798 14.831 multiply_cannon_metrocomm3 30084 15.6 0.046 0.049 6.349 14.522 cp_dbcsr_syevd 83 12.4 0.008 0.030 14.088 14.089 dbcsr_complete_redistribute 395 12.7 1.517 1.637 8.905 12.591 multiply_cannon_sync_h2d 30084 15.6 11.830 12.586 11.830 12.586 make_images_data 5014 15.6 0.063 0.068 10.924 12.407 dbcsr_mm_accdrv_process 62264 16.2 7.376 8.339 11.418 11.949 hybrid_alltoall_any 5200 16.5 0.524 2.224 9.851 11.199 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.477 11.138 init_scf_run 11 5.9 0.000 0.001 11.116 11.117 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.116 11.117 cp_fm_diag_elpa 83 13.4 0.001 0.002 10.910 10.915 cp_fm_redistribute_end 83 14.4 1.880 10.804 1.894 10.809 cp_fm_diag_elpa_base 83 14.4 8.310 10.186 8.883 10.673 pw_transfer 1547 11.6 0.088 0.106 9.652 9.739 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.177 9.731 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.373 9.578 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 9.423 9.520 mp_alltoall_i22 716 14.1 5.380 9.100 5.380 9.100 density_rs2pw 128 9.7 0.006 0.006 7.210 8.550 fft_wrap_pw1pw2_140 523 13.2 0.483 0.492 8.362 8.476 grid_integrate_task_list 128 12.3 7.539 7.872 7.539 7.872 cp_fm_cholesky_decompose 22 10.9 7.689 7.809 7.689 7.809 wfi_extrapolate 11 7.9 0.002 0.008 7.735 7.735 fft3d_ps 1291 14.7 2.772 2.881 7.647 7.714 multiply_cannon_metrocomm4 25070 15.6 0.077 0.087 2.771 7.180 mp_irecv_dv 76098 16.2 2.625 6.918 2.625 6.918 calculate_dm_sparse 128 9.5 0.001 0.001 6.169 6.243 rs_pw_transfer 1046 11.9 0.013 0.014 4.817 6.229 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.744 5.804 grid_collocate_task_list 128 9.7 5.082 5.740 5.082 5.740 mp_alltoall_d11v 2415 14.1 4.951 5.447 4.951 5.447 potential_pw2rs 128 12.3 0.023 0.023 4.848 4.864 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.631 4.688 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.473 4.475 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.364 4.443 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.236 4.292 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.377000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1082.818182, yerr=20.012806 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.512579E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532237E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58352. MP_Allreduce 10977 1175. MP_Sync 87 MP_Alltoall 1712 18838186. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.035 175.408 175.409 qs_mol_dyn_low 1 2.0 0.003 0.003 175.037 175.049 qs_forces 11 3.9 0.003 0.003 174.428 174.435 qs_energies 11 4.9 0.001 0.002 166.800 166.810 scf_env_do_scf 11 5.9 0.001 0.001 149.628 149.633 velocity_verlet 10 3.0 0.002 0.002 115.073 115.108 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 113.820 113.822 qs_scf_new_mos 117 7.6 0.001 0.001 74.993 75.050 qs_scf_loop_do_ot 117 8.6 0.001 0.001 74.992 75.050 dbcsr_multiply_generic 2507 12.6 0.182 0.186 73.813 74.262 ot_scf_mini 117 9.6 0.004 0.004 70.598 70.654 multiply_cannon 2507 13.6 0.582 0.615 54.301 59.233 multiply_cannon_loop 2507 14.6 0.441 0.451 49.398 50.158 ot_mini 117 10.6 0.001 0.001 39.247 39.310 init_scf_loop 11 6.9 0.000 0.000 35.662 35.665 mp_waitall_1 125778 16.7 25.503 33.091 25.503 33.091 prepare_preconditioner 11 7.9 0.000 0.000 31.670 31.692 make_preconditioner 11 8.9 0.000 0.000 31.670 31.692 rebuild_ks_matrix 128 8.3 0.001 0.001 30.048 30.106 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 30.047 30.106 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.538 29.807 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.256 27.305 multiply_cannon_multrec 10028 15.6 10.428 14.827 17.693 21.291 ot_diis_step 117 11.6 0.020 0.024 19.801 19.802 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.436 19.681 apply_single 128 13.6 0.001 0.001 19.436 19.680 qs_ot_get_derivative 117 11.6 0.001 0.002 19.376 19.432 multiply_cannon_metrocomm3 10028 15.6 0.023 0.024 12.373 19.010 qs_ot_get_p 128 10.4 0.001 0.001 18.437 18.521 make_m2s 5014 13.6 0.066 0.071 15.969 18.460 make_images 5014 14.6 2.270 2.787 15.663 18.156 cp_fm_cholesky_invert 11 10.9 18.108 18.115 18.108 18.115 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.502 15.535 calculate_rho_elec 128 8.7 0.259 0.268 15.502 15.535 sum_up_and_integrate 128 10.3 0.180 0.190 15.123 15.170 integrate_v_rspace 128 11.3 0.004 0.004 14.943 14.998 qs_ot_p2m_diag 83 11.4 0.495 0.501 14.699 14.717 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.569 13.572 make_images_data 5014 15.6 0.051 0.060 9.732 12.273 multiply_cannon_sync_h2d 10028 15.6 11.630 12.259 11.630 12.259 hybrid_alltoall_any 5200 16.5 0.818 3.595 9.579 11.666 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.475 10.486 cp_fm_diag_elpa_base 83 14.4 10.229 10.308 10.464 10.477 init_scf_run 11 5.9 0.000 0.001 10.349 10.349 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.349 10.349 pw_transfer 1547 11.6 0.085 0.094 9.847 9.888 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 9.624 9.671 fft_wrap_pw1pw2_140 523 13.2 0.495 0.515 8.485 8.532 grid_integrate_task_list 128 12.3 7.772 8.136 7.772 8.136 cp_fm_cholesky_decompose 22 10.9 7.910 8.016 7.910 8.016 density_rs2pw 128 9.7 0.005 0.006 7.048 7.918 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 7.796 7.837 fft3d_ps 1291 14.7 2.706 2.793 7.786 7.804 dbcsr_mm_accdrv_process 20762 16.1 2.650 3.618 6.894 7.520 mp_allgather_i34 2507 14.6 3.105 7.466 3.105 7.466 wfi_extrapolate 11 7.9 0.001 0.001 7.423 7.423 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 4.135 6.946 calculate_dm_sparse 128 9.5 0.001 0.001 6.045 6.134 grid_collocate_task_list 128 9.7 5.391 6.087 5.391 6.087 mp_alltoall_d11v 2415 14.1 4.958 5.845 4.958 5.845 dbcsr_complete_redistribute 395 12.7 2.129 2.247 5.148 5.472 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.311 5.312 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.139 5.148 rs_pw_transfer 1046 11.9 0.012 0.013 4.155 5.065 potential_pw2rs 128 12.3 0.027 0.028 4.606 4.621 multiply_cannon_metrocomm4 7521 15.6 0.024 0.027 1.846 4.431 mp_irecv_dv 28860 15.9 1.809 4.347 1.809 4.347 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.083 4.112 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.554 3.878 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.697 3.706 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.677 3.703 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.370 3.683 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.507 3.597 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=175.409000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1411.636364, yerr=51.928750 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 11.606412E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.143037E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108490000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3991 59299. MP_Allreduce 10975 1515. MP_Sync 86 MP_Alltoall 1700 36954339. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.046 0.074 300.336 300.336 qs_mol_dyn_low 1 2.0 0.003 0.004 299.782 299.792 qs_forces 11 3.9 0.003 0.003 299.686 299.689 qs_energies 11 4.9 0.049 0.056 290.218 290.225 scf_env_do_scf 11 5.9 0.001 0.001 267.782 267.790 velocity_verlet 10 3.0 0.002 0.002 215.466 215.473 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 140.682 140.684 init_scf_loop 11 6.9 0.000 0.000 126.842 126.846 prepare_preconditioner 11 7.9 0.000 0.000 121.753 121.788 make_preconditioner 11 8.9 0.000 0.000 121.753 121.788 make_full_inverse_cholesky 11 9.9 0.000 0.000 97.325 118.912 qs_scf_new_mos 116 7.6 0.001 0.001 90.397 90.483 qs_scf_loop_do_ot 116 8.6 0.001 0.001 90.396 90.482 ot_scf_mini 116 9.6 0.004 0.004 85.680 85.756 dbcsr_multiply_generic 2485 12.5 0.210 0.217 82.353 83.056 cp_fm_upper_to_full 104 14.8 54.782 78.266 54.782 78.266 multiply_cannon 2485 13.5 0.692 0.727 57.926 58.493 multiply_cannon_loop 2485 14.5 0.468 0.477 54.249 55.782 ot_mini 116 10.6 0.001 0.001 43.940 44.010 dbcsr_complete_redistribute 393 12.7 3.963 3.993 30.004 43.158 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.653 39.768 rebuild_ks_matrix 127 8.3 0.001 0.001 38.035 38.104 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 38.034 38.104 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.383 37.270 mp_alltoall_i22 712 14.1 22.233 35.481 22.233 35.481 qs_ks_update_qs_env 127 7.6 0.001 0.001 34.982 35.053 cp_fm_cholesky_invert 11 10.9 33.542 33.549 33.542 33.549 mp_waitall_1 102768 16.8 27.720 31.387 27.720 31.387 qs_ot_get_p 127 10.4 0.001 0.001 26.445 26.491 qs_ot_get_derivative 116 11.6 0.002 0.002 23.735 23.816 qs_ot_p2m_diag 82 11.4 0.867 0.874 22.386 22.417 make_m2s 4970 13.5 0.075 0.078 19.790 20.797 cp_dbcsr_syevd 82 12.4 0.006 0.006 20.679 20.681 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.482 20.492 calculate_rho_elec 127 8.7 0.479 0.480 20.482 20.492 make_images 4970 14.5 3.710 3.857 19.315 20.326 ot_diis_step 116 11.6 0.023 0.025 20.164 20.165 multiply_cannon_metrocomm3 9940 15.5 0.023 0.023 18.296 19.655 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.391 19.606 apply_single 127 13.6 0.001 0.001 19.390 19.606 sum_up_and_integrate 127 10.3 0.318 0.320 19.440 19.522 integrate_v_rspace 127 11.3 0.004 0.004 19.122 19.202 multiply_cannon_multrec 9940 15.5 10.457 12.247 17.833 17.907 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.447 17.447 cp_fm_diag_elpa_base 82 14.4 12.957 14.642 17.443 17.443 multiply_cannon_sync_h2d 9940 15.5 15.542 15.550 15.542 15.550 hybrid_alltoall_any 5155 16.4 1.293 3.011 10.799 12.643 make_images_data 4970 15.5 0.059 0.063 10.653 12.433 init_scf_run 11 5.9 0.000 0.001 12.281 12.281 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.281 12.281 pw_transfer 1535 11.6 0.092 0.092 12.232 12.248 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 11.999 12.014 fft_wrap_pw1pw2_140 519 13.2 0.543 0.578 10.661 10.679 fft3d_ps 1281 14.7 2.720 2.728 10.045 10.067 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.199 9.259 wfi_extrapolate 11 7.9 0.001 0.001 8.974 8.974 dbcsr_mm_accdrv_process 20590 16.0 3.841 5.591 7.138 8.912 cp_fm_cholesky_decompose 22 10.9 8.864 8.902 8.864 8.902 mp_alltoall_d11v 2401 14.1 7.827 8.770 7.827 8.770 grid_integrate_task_list 127 12.3 8.520 8.660 8.520 8.660 density_rs2pw 127 9.7 0.005 0.005 8.217 8.322 qs_energies_init_hamiltonians 11 5.9 0.021 0.041 8.134 8.134 calculate_dm_sparse 127 9.5 0.001 0.001 6.608 6.641 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.415 6.466 grid_collocate_task_list 127 9.7 6.306 6.333 6.306 6.333 rs_scatter_matrices 138 9.7 3.583 4.644 5.936 6.177 copy_dbcsr_to_fm 185 11.7 0.004 0.005 6.037 6.118 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=300.336000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2855.818182, yerr=156.903455 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 1.261724E+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 6805335. 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 0.063 0.086 85.209 85.210 qs_energies 1 2.0 0.000 0.000 84.520 84.526 ls_scf 1 3.0 0.000 0.002 83.602 83.610 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.470 72.677 multiply_cannon 111 7.7 0.017 0.020 55.783 57.179 multiply_cannon_loop 111 8.7 0.208 0.223 52.409 54.061 ls_scf_main 1 4.0 0.000 0.000 52.422 52.422 density_matrix_trs4 2 5.0 0.002 0.003 46.854 46.956 ls_scf_init_scf 1 4.0 0.000 0.002 28.136 28.138 ls_scf_init_matrix_S 1 5.0 0.000 0.001 26.962 27.029 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.848 24.866 mp_waitall_1 11031 10.9 22.369 24.385 22.369 24.385 multiply_cannon_multrec 2664 9.7 8.173 8.895 15.459 17.117 multiply_cannon_sync_h2d 2664 9.7 13.699 15.168 13.699 15.168 make_m2s 222 7.7 0.008 0.010 13.095 13.566 make_images 222 8.7 0.099 0.110 13.073 13.545 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.578 12.531 make_images_data 222 9.7 0.004 0.005 7.630 8.158 hybrid_alltoall_any 227 10.6 0.215 1.832 6.555 7.912 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.492 7.906 dbcsr_mm_accdrv_process 4760 10.4 0.508 0.605 6.908 7.879 dbcsr_mm_accdrv_process_sort 4760 11.4 6.198 7.095 6.198 7.095 calculate_norms 4752 9.8 5.527 6.157 5.527 6.157 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.055 5.149 mp_sum_l 807 5.4 3.061 4.571 3.061 4.571 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.075 3.736 mp_irecv_dv 6231 10.9 2.058 3.702 2.058 3.702 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.265 3.473 arnoldi_extremal 4 6.8 0.000 0.000 3.292 3.320 arnoldi_normal_ev 4 7.8 0.001 0.002 3.292 3.320 make_images_sizes 222 9.7 0.000 0.000 0.670 3.280 mp_alltoall_i44 222 10.7 0.670 3.280 0.670 3.280 build_subspace 16 8.4 0.009 0.012 3.198 3.201 ls_scf_post 1 4.0 0.000 0.002 3.044 3.051 ls_scf_store_result 1 5.0 0.000 0.000 2.853 2.893 dbcsr_special_finalize 555 9.7 0.005 0.006 2.443 2.824 dbcsr_merge_single_wm 555 10.7 0.464 0.594 2.434 2.816 make_images_pack 222 9.7 2.206 2.626 2.208 2.627 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.380 2.624 dbcsr_sort_data 658 11.4 2.227 2.550 2.227 2.550 dbcsr_matrix_vector_mult_local 304 10.0 2.075 2.488 2.077 2.490 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.302 2.374 buffer_matrices_ensure_size 222 8.7 1.758 2.078 1.758 2.078 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.826 1.827 rebuild_ks_matrix 3 7.3 0.000 0.000 1.816 1.818 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.009 1.816 1.818 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.210000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 2.135527E+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 0.030 0.047 91.336 91.336 qs_energies 1 2.0 0.000 0.000 90.882 90.887 ls_scf 1 3.0 0.000 0.000 89.574 89.578 dbcsr_multiply_generic 111 6.7 0.015 0.015 75.599 75.967 multiply_cannon 111 7.7 0.028 0.040 53.427 57.431 ls_scf_main 1 4.0 0.000 0.000 55.293 55.299 multiply_cannon_loop 111 8.7 0.116 0.122 50.146 53.708 density_matrix_trs4 2 5.0 0.002 0.003 49.544 49.708 ls_scf_init_scf 1 4.0 0.000 0.000 30.755 30.756 mp_waitall_1 9105 10.9 21.443 30.224 21.443 30.224 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.505 29.626 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.091 27.107 multiply_cannon_multrec 1332 9.7 13.178 16.684 22.363 26.998 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.965 21.266 make_m2s 222 7.7 0.006 0.007 15.448 16.180 make_images 222 8.7 1.570 1.961 15.418 16.152 dbcsr_mm_accdrv_process 4041 10.4 0.300 0.450 8.784 10.364 dbcsr_mm_accdrv_process_sort 4041 11.4 8.377 9.914 8.377 9.914 make_images_data 222 9.7 0.004 0.004 8.866 9.876 hybrid_alltoall_any 227 10.6 0.523 2.482 8.242 9.438 mp_sum_l 807 5.4 5.426 8.642 5.426 8.642 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.198 7.505 mp_irecv_dv 3311 11.0 3.179 7.459 3.179 7.459 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.216 7.008 calculate_norms 2376 9.8 5.999 6.770 5.999 6.770 multiply_cannon_sync_h2d 1332 9.7 4.806 6.162 4.806 6.162 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.092 5.258 arnoldi_extremal 4 6.8 0.000 0.000 4.633 4.656 arnoldi_normal_ev 4 7.8 0.001 0.005 4.633 4.656 build_subspace 16 8.4 0.014 0.021 4.370 4.373 ls_scf_post 1 4.0 0.000 0.000 3.526 3.531 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.137 3.358 ls_scf_store_result 1 5.0 0.000 0.000 3.239 3.353 dbcsr_matrix_vector_mult_local 304 10.0 2.733 3.213 2.735 3.215 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.242 2.698 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.610 2.690 make_images_pack 222 9.7 2.021 2.426 2.023 2.428 mp_allgather_i34 111 8.7 0.970 2.400 0.970 2.400 dbcsr_sort_data 436 11.2 1.826 2.134 1.826 2.134 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.915 1.917 rebuild_ks_matrix 3 7.3 0.000 0.000 1.903 1.905 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.903 1.905 dbcsr_data_new 4174 10.1 1.608 1.859 1.608 1.859 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.336000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1737.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 2.701083E+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 0.031 0.047 94.368 94.369 qs_energies 1 2.0 0.000 0.000 93.791 93.812 ls_scf 1 3.0 0.000 0.000 92.315 92.336 dbcsr_multiply_generic 111 6.7 0.015 0.016 76.972 77.271 ls_scf_main 1 4.0 0.000 0.000 58.046 58.051 multiply_cannon 111 7.7 0.039 0.072 53.311 56.672 multiply_cannon_loop 111 8.7 0.100 0.105 49.639 53.834 density_matrix_trs4 2 5.0 0.002 0.003 52.051 52.199 mp_waitall_1 7281 11.0 23.940 34.476 23.940 34.476 ls_scf_init_scf 1 4.0 0.000 0.000 30.682 30.684 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.477 29.552 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.063 27.076 multiply_cannon_multrec 888 9.7 12.583 15.449 21.095 24.543 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.443 23.856 make_m2s 222 7.7 0.006 0.007 16.783 18.148 make_images 222 8.7 1.965 2.311 16.744 18.107 make_images_data 222 9.7 0.004 0.004 9.432 10.560 hybrid_alltoall_any 227 10.6 0.621 2.851 9.073 10.503 dbcsr_mm_accdrv_process 3754 10.4 0.256 0.411 8.012 9.220 mp_sum_l 807 5.4 5.327 8.927 5.327 8.927 dbcsr_mm_accdrv_process_sort 3754 11.4 7.644 8.810 7.644 8.810 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.183 7.109 multiply_cannon_sync_h2d 888 9.7 6.117 7.083 6.117 7.083 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.409 6.906 mp_irecv_dv 2335 11.1 2.395 6.861 2.395 6.861 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.690 6.349 arnoldi_extremal 4 6.8 0.000 0.000 5.119 5.136 arnoldi_normal_ev 4 7.8 0.001 0.005 5.119 5.136 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.927 5.122 build_subspace 16 8.4 0.014 0.020 4.815 4.823 calculate_norms 1584 9.8 4.274 4.625 4.274 4.625 mp_allgather_i34 111 8.7 1.513 3.817 1.513 3.817 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.449 3.786 ls_scf_post 1 4.0 0.000 0.000 3.588 3.610 dbcsr_matrix_vector_mult_local 304 10.0 3.008 3.586 3.011 3.588 ls_scf_store_result 1 5.0 0.000 0.000 3.321 3.413 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.865 2.972 dbcsr_sort_data 325 11.1 1.851 2.190 1.851 2.190 make_images_pack 222 9.7 1.808 2.108 1.810 2.111 make_images_sizes 222 9.7 0.000 0.000 1.068 2.106 mp_alltoall_i44 222 10.7 1.068 2.105 1.068 2.105 dbcsr_data_release 9322 10.9 1.337 1.936 1.337 1.936 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.923 1.925 rebuild_ks_matrix 3 7.3 0.000 0.000 1.905 1.907 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.905 1.907 dbcsr_finalize 304 7.8 0.027 0.032 1.603 1.891 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.369000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2253.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 3.331412E+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 0.032 0.049 96.635 96.636 qs_energies 1 2.0 0.000 0.000 96.036 96.043 ls_scf 1 3.0 0.000 0.000 94.303 94.308 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.039 78.335 ls_scf_main 1 4.0 0.000 0.000 58.532 58.533 multiply_cannon 111 7.7 0.053 0.086 51.185 55.961 density_matrix_trs4 2 5.0 0.002 0.003 52.375 52.511 multiply_cannon_loop 111 8.7 0.115 0.127 46.064 49.552 ls_scf_init_scf 1 4.0 0.000 0.000 32.550 32.552 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.314 31.393 mp_waitall_1 6369 11.0 22.137 30.502 22.137 30.502 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.846 28.859 multiply_cannon_multrec 1332 9.7 14.202 17.399 22.016 25.630 make_m2s 222 7.7 0.007 0.008 20.889 22.378 make_images 222 8.7 3.137 3.608 20.838 22.330 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 8.824 17.028 make_images_data 222 9.7 0.004 0.004 11.645 13.263 hybrid_alltoall_any 227 10.6 0.798 3.762 10.974 12.932 dbcsr_mm_accdrv_process 3641 10.4 0.194 0.396 7.458 8.947 dbcsr_mm_accdrv_process_sort 3641 11.4 7.087 8.541 7.087 8.541 mp_sum_l 807 5.4 4.389 7.980 4.389 7.980 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.451 6.360 multiply_cannon_sync_h2d 1332 9.7 5.509 6.267 5.509 6.267 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.066 5.924 mp_irecv_dv 3229 10.9 2.043 5.856 2.043 5.856 arnoldi_extremal 4 6.8 0.000 0.000 5.264 5.279 arnoldi_normal_ev 4 7.8 0.001 0.005 5.264 5.279 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.499 5.048 build_subspace 16 8.4 0.014 0.021 4.933 4.939 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.561 4.699 mp_allgather_i34 111 8.7 2.245 4.695 2.245 4.695 calculate_norms 2376 9.8 4.213 4.560 4.213 4.560 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.592 3.907 dbcsr_matrix_vector_mult_local 304 10.0 3.166 3.667 3.168 3.669 dbcsr_sort_data 658 11.4 3.064 3.392 3.064 3.392 ls_scf_post 1 4.0 0.000 0.000 3.220 3.227 dbcsr_special_finalize 555 9.7 0.006 0.006 2.812 3.180 dbcsr_merge_single_wm 555 10.7 0.536 0.660 2.804 3.172 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.045 3.084 ls_scf_store_result 1 5.0 0.000 0.000 2.967 3.021 dbcsr_data_release 10477 10.7 1.577 2.387 1.577 2.387 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.975 1.977 dbcsr_finalize 304 7.8 0.049 0.061 1.796 1.968 rebuild_ks_matrix 3 7.3 0.000 0.000 1.952 1.953 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.952 1.953 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.636000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2711.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 4.578099E+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 0.044 0.079 99.279 99.280 qs_energies 1 2.0 0.000 0.000 98.441 98.447 ls_scf 1 3.0 0.000 0.000 96.490 96.502 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.752 77.944 ls_scf_main 1 4.0 0.000 0.000 62.268 62.269 multiply_cannon 111 7.7 0.092 0.177 55.487 60.731 density_matrix_trs4 2 5.0 0.002 0.002 54.890 54.979 multiply_cannon_loop 111 8.7 0.070 0.079 50.910 52.364 mp_waitall_1 5436 11.0 26.363 32.126 26.363 32.126 ls_scf_init_scf 1 4.0 0.000 0.000 30.630 30.636 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.316 29.351 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.125 27.137 multiply_cannon_multrec 444 9.7 14.099 16.711 21.095 24.409 make_m2s 222 7.7 0.004 0.005 17.473 20.027 make_images 222 8.7 3.724 4.423 17.411 19.967 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.152 15.869 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.280 15.814 make_images_data 222 9.7 0.003 0.004 9.734 12.149 hybrid_alltoall_any 227 10.6 0.791 3.781 9.539 11.983 multiply_cannon_sync_h2d 444 9.7 6.529 8.994 6.529 8.994 dbcsr_mm_accdrv_process 3003 10.4 0.176 0.346 6.704 7.829 dbcsr_mm_accdrv_process_sort 3003 11.4 6.389 7.482 6.389 7.482 mp_allgather_i34 111 8.7 2.803 7.068 2.803 7.068 arnoldi_extremal 4 6.8 0.000 0.000 5.815 5.828 arnoldi_normal_ev 4 7.8 0.001 0.004 5.814 5.828 mp_sum_l 807 5.4 2.847 5.805 2.847 5.805 build_subspace 16 8.4 0.015 0.020 5.423 5.432 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.614 4.768 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.956 4.534 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.155 4.351 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.670 4.197 mp_irecv_dv 1241 11.2 1.649 4.172 1.649 4.172 dbcsr_matrix_vector_mult_local 304 10.0 3.660 4.138 3.662 4.140 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.775 3.870 calculate_norms 792 9.8 3.549 3.690 3.549 3.690 ls_scf_post 1 4.0 0.000 0.000 3.592 3.598 ls_scf_store_result 1 5.0 0.000 0.000 3.369 3.412 make_images_sizes 222 9.7 0.000 0.000 1.068 3.262 mp_alltoall_i44 222 10.7 1.068 3.262 1.068 3.262 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.270 2.271 dbcsr_finalize 304 7.8 0.062 0.078 2.201 2.254 rebuild_ks_matrix 3 7.3 0.000 0.000 2.237 2.238 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.237 2.238 dbcsr_merge_all 275 8.9 0.473 0.522 2.052 2.089 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.280000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3588.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1edf9b76a3d3b174dfad07f04161737566d716f7_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 8.714699E+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 0.080 0.099 108.646 108.646 qs_energies 1 2.0 0.000 0.000 107.248 107.261 ls_scf 1 3.0 0.000 0.000 104.309 104.323 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.722 77.860 ls_scf_main 1 4.0 0.000 0.000 66.191 66.191 density_matrix_trs4 2 5.0 0.002 0.003 56.890 56.962 multiply_cannon 111 7.7 0.150 0.294 49.975 51.716 multiply_cannon_loop 111 8.7 0.068 0.071 46.531 47.593 ls_scf_init_scf 1 4.0 0.000 0.000 34.449 34.450 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.780 32.796 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.944 29.957 mp_waitall_1 4527 11.1 22.319 26.207 22.319 26.207 make_m2s 222 7.7 0.005 0.005 23.988 24.971 make_images 222 8.7 4.585 4.974 23.882 24.864 multiply_cannon_multrec 444 9.7 17.852 18.625 22.435 23.023 hybrid_alltoall_any 227 10.6 1.661 3.645 12.948 16.008 make_images_data 222 9.7 0.003 0.003 13.221 15.682 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.479 11.276 multiply_cannon_sync_h2d 444 9.7 8.820 8.872 8.820 8.872 arnoldi_extremal 4 6.8 0.000 0.000 7.349 7.362 arnoldi_normal_ev 4 7.8 0.003 0.009 7.349 7.362 build_subspace 16 8.4 0.026 0.036 6.788 6.799 dbcsr_matrix_vector_mult 304 9.0 0.016 0.032 5.403 5.554 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.279 5.373 dbcsr_matrix_vector_mult_local 304 10.0 4.954 5.266 4.957 5.269 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.994 5.256 dbcsr_mm_accdrv_process 1814 10.4 0.253 0.316 4.409 4.532 dbcsr_mm_accdrv_process_sort 1814 11.4 4.109 4.240 4.109 4.240 ls_scf_post 1 4.0 0.000 0.000 3.670 3.683 make_images_sizes 222 9.7 0.000 0.000 1.449 3.601 mp_alltoall_i44 222 10.7 1.449 3.600 1.449 3.600 ls_scf_store_result 1 5.0 0.000 0.000 3.414 3.432 mp_allgather_i34 111 8.7 1.096 3.429 1.096 3.429 calculate_norms 792 9.8 3.272 3.330 3.272 3.330 dbcsr_finalize 304 7.8 0.082 0.089 3.089 3.168 dbcsr_merge_all 275 8.9 0.894 0.923 2.874 2.950 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.908 2.908 dbcsr_complete_redistribute 5 7.6 1.441 1.501 2.790 2.900 dbcsr_data_release 12724 10.6 2.338 2.872 2.338 2.872 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.597 2.599 matrix_ls_to_qs 2 6.0 0.000 0.000 2.453 2.563 rebuild_ks_matrix 3 7.3 0.000 0.000 2.531 2.533 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.531 2.533 dbcsr_sort_data 325 11.1 2.442 2.498 2.442 2.498 dbcsr_new_transposed 4 7.5 0.244 0.252 2.292 2.304 dbcsr_frobenius_norm 74 6.6 2.057 2.133 2.185 2.225 mp_sum_l 807 5.4 1.424 2.220 1.424 2.220 dbcsr_add_d 103 6.2 0.000 0.000 2.136 2.204 dbcsr_add_anytype 103 7.2 0.860 0.891 2.136 2.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.646000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6812.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 1edf9b76a3d3b174dfad07f04161737566d716f7 Summary: empty Status: OK