=== 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: 3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9 ################# 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 7.3.2, 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 (12.01.2023) # # \ 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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/01 job id: 44501808 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/02 job id: 44501811 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/03 job id: 44501815 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/04 job id: 44501819 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/05 job id: 44501820 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/06 job id: 44501822 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/07 job id: 44501823 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/08 job id: 44501824 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/09 job id: 44501825 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/10 job id: 44501826 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/11 job id: 44501827 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/12 job id: 44501828 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/13 job id: 44501829 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/14 job id: 44501830 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/15 job id: 44501831 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/16 job id: 44501832 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/17 job id: 44501833 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/18 job id: 44501834 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/19 job id: 44501835 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/20 job id: 44501836 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/21 job id: 44501837 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/22 job id: 44501838 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/23 job id: 44501839 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/24 job id: 44501840 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/25 job id: 44501841 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/26 job id: 44501842 --- 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_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 0.027 0.037 139.089 139.090 farming_run 1 2.0 138.403 138.404 139.053 139.055 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.457910E+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 228 1113141. 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 0.032 117.007 117.008 qs_energies 1 2.0 0.000 0.000 116.743 116.746 mp2_main 1 3.0 0.000 0.000 113.943 113.945 mp2_gpw_main 1 4.0 0.020 0.028 112.904 112.906 mp2_ri_gpw_compute_in 1 5.0 0.174 0.192 93.861 94.437 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.406 55.983 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.167 41.716 47.245 get_2c_integrals 1 6.0 0.001 0.010 37.720 38.283 integrate_v_rspace 273 8.0 0.436 0.451 25.140 30.365 pw_transfer 6555 10.6 0.371 0.381 27.418 28.013 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 26.070 26.657 grid_integrate_task_list 273 9.0 20.931 26.629 20.931 26.629 fft_wrap_pw1pw2_100 2178 12.4 1.205 1.397 23.580 24.154 compute_2c_integrals 1 7.0 0.016 0.025 19.631 19.633 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.928 19.310 mp2_eri_2c_integrate_gpw 1 9.0 2.382 2.428 18.925 19.308 rpa_ri_compute_en 1 5.0 0.018 0.021 18.936 19.087 cp_fm_cholesky_decompose 12 8.2 18.055 18.634 18.055 18.634 cholesky_decomp 1 7.0 0.000 0.000 16.934 17.508 fft3d_s 5443 13.4 16.170 16.601 16.191 16.621 ao_to_mo_and_store_B_mult_1 272 7.0 10.853 15.587 10.853 15.587 calculate_wavefunction 272 8.0 5.408 5.563 12.525 13.162 rpa_num_int 1 6.0 0.000 0.005 10.489 10.490 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.463 10.468 calc_mat_Q 8 8.0 0.000 0.000 9.321 9.408 contract_S_to_Q 8 9.0 0.000 0.000 8.742 8.830 calc_potential_gpw 544 9.5 0.005 0.006 8.194 8.692 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.217 8.534 potential_pw2rs 545 10.0 0.107 0.109 7.775 8.415 parallel_gemm_fm 14 9.1 0.000 0.000 8.314 8.412 parallel_gemm_fm_cosma 14 10.1 8.314 8.412 8.314 8.412 create_integ_mat 1 6.0 0.022 0.027 7.755 7.755 collocate_single_gaussian 272 10.0 0.039 0.042 7.498 7.748 array2fm 1 7.0 0.000 0.000 6.712 7.167 pw_scatter_s 2720 13.7 4.407 4.564 4.407 4.564 pw_gather_s 2722 13.2 3.866 4.123 3.866 4.123 array2fm_buffer_send 1 8.0 2.979 3.095 2.979 3.095 scf_env_do_scf 1 3.0 0.000 0.000 2.506 2.506 pw_poisson_solve 545 10.5 1.129 1.176 2.187 2.447 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.906045, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2730.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_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 0.033 0.041 401.652 401.653 farming_run 1 2.0 400.801 400.809 401.600 401.604 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.224999E+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 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 2368424. 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 0.030 211.934 211.935 qs_energies 1 2.0 0.000 0.000 211.686 211.692 scf_env_do_scf 1 3.0 0.000 0.000 107.121 107.121 qs_ks_update_qs_env 5 5.0 0.000 0.000 106.000 106.007 rebuild_ks_matrix 4 6.0 0.000 0.000 105.998 106.006 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.065 105.998 106.006 hfx_ks_matrix 4 8.0 0.001 0.001 105.589 105.593 integrate_four_center 4 9.0 0.144 0.460 105.588 105.592 mp2_main 1 3.0 0.000 0.000 104.277 104.283 mp2_gpw_main 1 4.0 0.036 0.054 103.204 103.213 integrate_four_center_main 4 10.0 0.079 0.516 96.883 99.773 integrate_four_center_bin 264 11.0 96.804 99.771 96.804 99.771 init_scf_loop 1 4.0 0.000 0.000 92.676 92.676 mp2_ri_gpw_compute_in 1 5.0 0.065 0.072 74.936 76.023 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.539 55.625 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.161 42.269 47.271 integrate_v_rspace 95 8.0 0.397 0.566 28.652 33.544 pw_transfer 2240 10.6 0.145 0.161 29.915 30.312 mp2_ri_gpw_compute_en 1 5.0 0.056 0.065 28.113 29.776 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.948 29.328 grid_integrate_task_list 95 9.0 23.919 28.978 23.919 28.978 fft_wrap_pw1pw2_100 730 12.4 1.240 1.397 26.613 27.003 ao_to_mo_and_store_B_mult_1 91 7.0 10.579 26.791 10.579 26.791 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.852 1.901 26.380 26.393 get_2c_integrals 1 6.0 0.000 0.000 20.313 20.336 compute_2c_integrals 1 7.0 0.003 0.004 19.297 19.304 compute_2c_integrals_loop_lm 1 8.0 0.003 0.008 18.740 19.105 mp2_eri_2c_integrate_gpw 1 9.0 1.743 1.848 18.737 19.104 fft3d_s 1823 13.4 18.415 18.761 18.428 18.774 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.443 14.443 calculate_wavefunction 91 8.0 2.025 2.071 9.753 9.956 mp2_ri_gpw_compute_en_comm 22 7.0 0.484 0.502 9.401 9.850 mp2_ri_gpw_compute_en_expansio 172 7.0 0.553 0.579 8.740 9.118 potential_pw2rs 186 10.0 0.033 0.035 8.599 9.063 local_gemm 172 8.0 8.186 8.561 8.186 8.561 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.168 8.494 calc_potential_gpw 182 9.5 0.002 0.002 7.901 8.103 collocate_single_gaussian 91 10.0 0.016 0.021 7.826 8.059 mp_sendrecv_dm3 2068 8.0 7.478 7.917 7.478 7.917 mp2_ri_gpw_compute_en_ener 172 7.0 6.343 6.423 6.343 6.423 mp_sync 38 10.4 3.340 5.903 3.340 5.903 pw_gather_s 912 13.2 4.918 5.304 4.918 5.304 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=103.202479, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1512.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.124672E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 589622. 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 0.049 0.102 63.862 63.863 qs_mol_dyn_low 1 2.0 0.006 0.022 63.379 63.399 qs_forces 11 3.9 0.002 0.005 63.265 63.267 qs_energies 11 4.9 0.004 0.024 61.465 61.481 scf_env_do_scf 11 5.9 0.002 0.007 54.000 54.001 scf_env_do_scf_inner_loop 108 6.5 0.003 0.014 51.304 51.306 dbcsr_multiply_generic 2286 12.5 0.100 0.108 39.351 39.921 qs_scf_new_mos 108 7.5 0.000 0.001 38.597 38.905 qs_scf_loop_do_ot 108 8.5 0.001 0.007 38.596 38.905 ot_scf_mini 108 9.5 0.003 0.013 36.751 36.942 velocity_verlet 10 3.0 0.001 0.001 31.143 31.144 multiply_cannon 2286 13.5 0.182 0.192 28.663 30.798 multiply_cannon_loop 2286 14.5 1.504 1.567 27.376 29.164 ot_mini 108 10.5 0.001 0.001 23.367 23.632 qs_ot_get_derivative 108 11.5 0.001 0.005 19.974 20.161 mp_waitall_1 245248 16.5 11.454 17.615 11.454 17.615 multiply_cannon_metrocomm3 54864 15.5 0.069 0.074 6.799 15.127 multiply_cannon_multrec 54864 15.5 4.053 6.216 7.789 11.291 rebuild_ks_matrix 119 8.3 0.000 0.000 9.495 9.652 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.022 9.495 9.652 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.340 8.482 mp_sum_l 7207 12.9 6.579 8.457 6.579 8.457 qs_ot_get_p 119 10.4 0.001 0.001 8.045 8.300 multiply_cannon_sync_h2d 54864 15.5 5.630 7.445 5.630 7.445 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 7.041 7.164 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.518 6.976 init_scf_run 11 5.9 0.000 0.001 5.817 5.818 scf_env_initial_rho_setup 11 6.9 0.001 0.006 5.817 5.818 qs_rho_update_rho_low 119 7.7 0.001 0.009 5.659 5.762 calculate_rho_elec 119 8.7 0.012 0.024 5.657 5.755 sum_up_and_integrate 119 10.3 0.013 0.016 5.683 5.690 integrate_v_rspace 119 11.3 0.002 0.005 5.670 5.679 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 3.120 5.205 dbcsr_mm_accdrv_process 76910 16.1 1.204 1.853 3.659 5.092 qs_ot_p2m_diag 50 11.0 0.004 0.006 4.217 4.292 make_m2s 4572 13.5 0.054 0.056 3.855 4.007 make_images 4572 14.5 0.133 0.139 3.772 3.924 rs_pw_transfer 974 11.9 0.011 0.012 3.741 3.833 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.196 3.392 apply_single 119 13.6 0.000 0.000 3.195 3.391 calculate_dm_sparse 119 9.5 0.000 0.000 3.213 3.365 density_rs2pw 119 9.7 0.004 0.004 3.230 3.336 cp_dbcsr_syevd 50 12.0 0.002 0.003 3.291 3.291 jit_kernel_multiply 13 15.8 2.392 3.156 2.392 3.156 ot_diis_step 108 11.5 0.006 0.006 3.058 3.060 calculate_first_density_matrix 1 7.0 0.001 0.003 3.003 3.007 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.918 2.927 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.828 2.902 mp_alltoall_d11v 2130 13.8 2.689 2.854 2.689 2.854 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.812 2.821 cp_fm_redistribute_end 50 14.0 2.554 2.779 2.562 2.785 pw_transfer 1439 11.6 0.053 0.058 2.681 2.755 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 2.605 2.682 wfi_extrapolate 11 7.9 0.001 0.001 2.680 2.681 init_scf_loop 11 6.9 0.001 0.010 2.673 2.676 make_images_sizes 4572 15.5 0.004 0.005 2.147 2.673 mp_alltoall_i44 4572 16.5 2.143 2.669 2.143 2.669 potential_pw2rs 119 12.3 0.004 0.004 2.581 2.598 cp_fm_diag_elpa_base 50 14.0 0.212 2.555 0.213 2.567 fft3d_ps 1201 14.6 0.376 0.481 2.375 2.450 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.372 2.444 mp_sum_d 4129 12.0 1.720 2.385 1.720 2.385 acc_transpose_blocks 54864 15.5 0.232 0.251 1.832 2.280 grid_integrate_task_list 119 12.3 2.012 2.112 2.012 2.112 fft_wrap_pw1pw2_140 487 13.2 0.082 0.095 2.021 2.103 mp_waitany 12084 13.8 1.547 1.677 1.547 1.677 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.604 1.623 make_images_data 4572 15.5 0.039 0.043 1.039 1.601 prepare_preconditioner 11 7.9 0.000 0.001 1.426 1.458 make_preconditioner 11 8.9 0.000 0.002 1.426 1.458 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.967 1.441 hybrid_alltoall_any 4725 16.4 0.040 0.107 0.919 1.409 mp_alltoall_z22v 1201 16.6 1.340 1.399 1.340 1.399 grid_collocate_task_list 119 9.7 1.296 1.378 1.296 1.378 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.324 1.363 mp_allgather_i34 2286 14.5 0.817 1.282 0.817 1.282 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=63.863000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.272727, yerr=1.052349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.944192E+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 1031054. 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 0.029 0.035 38.894 38.895 qs_mol_dyn_low 1 2.0 0.003 0.003 38.703 38.712 qs_forces 11 3.9 0.002 0.002 38.640 38.640 qs_energies 11 4.9 0.002 0.002 36.960 36.963 scf_env_do_scf 11 5.9 0.000 0.001 31.883 31.884 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.454 29.456 dbcsr_multiply_generic 2286 12.5 0.100 0.103 21.353 21.714 qs_scf_new_mos 108 7.5 0.001 0.001 20.030 20.279 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.030 20.278 ot_scf_mini 108 9.5 0.002 0.003 19.146 19.319 velocity_verlet 10 3.0 0.001 0.001 18.009 18.010 multiply_cannon 2286 13.5 0.208 0.219 16.307 17.913 multiply_cannon_loop 2286 14.5 0.906 0.984 15.205 16.558 ot_mini 108 10.5 0.001 0.001 11.753 11.989 mp_waitall_1 200699 16.5 5.715 10.847 5.715 10.847 multiply_cannon_metrocomm3 27432 15.5 0.069 0.072 4.130 9.647 qs_ot_get_derivative 108 11.5 0.001 0.001 9.278 9.458 multiply_cannon_multrec 27432 15.5 1.974 4.554 5.862 8.652 rebuild_ks_matrix 119 8.3 0.000 0.000 7.331 7.469 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.331 7.469 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.492 6.617 dbcsr_mm_accdrv_process 47894 16.0 3.116 5.350 3.819 5.649 qs_ot_get_p 119 10.4 0.001 0.001 4.463 4.703 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.536 4.399 sum_up_and_integrate 119 10.3 0.024 0.026 4.269 4.276 integrate_v_rspace 119 11.3 0.002 0.002 4.245 4.253 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.055 4.130 apply_single 119 13.6 0.000 0.000 3.055 4.129 mp_sum_l 7207 12.9 2.043 4.011 2.043 4.011 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.912 3.948 calculate_rho_elec 119 8.7 0.021 0.024 3.911 3.948 init_scf_run 11 5.9 0.000 0.001 3.855 3.856 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.855 3.856 rs_pw_transfer 974 11.9 0.010 0.011 2.680 3.143 qs_ot_p2m_diag 50 11.0 0.009 0.012 2.981 3.001 multiply_cannon_sync_h2d 27432 15.5 2.196 2.847 2.196 2.847 make_m2s 4572 13.5 0.051 0.053 2.548 2.778 density_rs2pw 119 9.7 0.004 0.004 2.218 2.705 make_images 4572 14.5 0.200 0.239 2.461 2.688 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.530 2.531 ot_diis_step 108 11.5 0.010 0.011 2.425 2.427 init_scf_loop 11 6.9 0.000 0.000 2.409 2.411 calculate_first_density_matrix 1 7.0 0.000 0.000 2.334 2.336 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.163 2.254 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.168 2.169 cp_fm_redistribute_end 50 14.0 1.799 2.146 1.803 2.147 calculate_dm_sparse 119 9.5 0.000 0.001 2.022 2.098 cp_fm_diag_elpa_base 50 14.0 0.330 2.022 0.343 2.072 pw_transfer 1439 11.6 0.065 0.072 1.996 2.028 potential_pw2rs 119 12.3 0.006 0.006 1.953 1.961 grid_integrate_task_list 119 12.3 1.831 1.952 1.831 1.952 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.904 1.940 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.916 1.918 jit_kernel_multiply 9 15.9 0.650 1.750 0.650 1.750 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.691 1.733 fft3d_ps 1201 14.6 0.513 0.565 1.604 1.636 make_images_data 4572 15.5 0.044 0.050 1.163 1.578 fft_wrap_pw1pw2_140 487 13.2 0.079 0.085 1.513 1.551 prepare_preconditioner 11 7.9 0.000 0.000 1.509 1.536 make_preconditioner 11 8.9 0.000 0.000 1.509 1.536 hybrid_alltoall_any 4725 16.4 0.050 0.112 1.012 1.518 acc_transpose_blocks 27432 15.5 0.108 0.113 1.197 1.501 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.402 1.459 wfi_extrapolate 11 7.9 0.001 0.001 1.450 1.450 mp_alltoall_d11v 2130 13.8 1.286 1.426 1.286 1.426 grid_collocate_task_list 119 9.7 1.226 1.367 1.226 1.367 mp_allgather_i34 2286 14.5 0.545 1.329 0.545 1.329 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.246 1.254 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.187 1.237 mp_sum_d 4129 12.0 0.591 1.026 0.591 1.026 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.147 0.542 1.009 mp_waitany 5720 13.7 0.522 0.994 0.522 0.994 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.960 0.961 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.896 0.909 acc_transpose_blocks_kernels 27432 16.5 0.183 0.273 0.662 0.888 make_images_sizes 4572 15.5 0.005 0.005 0.574 0.827 mp_alltoall_i44 4572 16.5 0.570 0.823 0.570 0.823 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.608 0.762 0.813 mp_alltoall_z22v 1201 16.6 0.706 0.808 0.706 0.808 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.895000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.818182, yerr=1.898238 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 520.568832E+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 1607811. 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 0.055 0.113 33.201 33.202 qs_mol_dyn_low 1 2.0 0.004 0.008 32.798 32.807 qs_forces 11 3.9 0.012 0.050 32.474 32.477 qs_energies 11 4.9 0.002 0.005 30.873 30.883 scf_env_do_scf 11 5.9 0.001 0.002 25.909 25.909 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.312 23.313 dbcsr_multiply_generic 2286 12.5 0.094 0.096 16.304 16.395 velocity_verlet 10 3.0 0.001 0.002 15.480 15.489 qs_scf_new_mos 108 7.5 0.001 0.001 14.817 14.830 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.816 14.829 ot_scf_mini 108 9.5 0.002 0.003 14.102 14.109 multiply_cannon 2286 13.5 0.194 0.198 12.984 13.792 multiply_cannon_loop 2286 14.5 0.637 0.667 12.204 13.019 ot_mini 108 10.5 0.001 0.002 8.727 8.739 qs_ot_get_derivative 108 11.5 0.001 0.001 7.224 7.231 multiply_cannon_multrec 18288 15.5 1.930 2.891 6.720 7.037 rebuild_ks_matrix 119 8.3 0.000 0.000 6.518 6.534 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 6.517 6.534 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.770 5.784 dbcsr_mm_accdrv_process 38222 16.0 4.073 5.311 4.706 5.505 sum_up_and_integrate 119 10.3 0.029 0.031 4.055 4.061 integrate_v_rspace 119 11.3 0.002 0.003 4.025 4.035 init_scf_run 11 5.9 0.000 0.001 3.671 3.671 scf_env_initial_rho_setup 11 6.9 0.001 0.003 3.671 3.671 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.635 3.640 calculate_rho_elec 119 8.7 0.030 0.031 3.635 3.640 mp_waitall_1 158411 16.6 2.669 3.552 2.669 3.552 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.654 3.298 qs_ot_get_p 119 10.4 0.001 0.001 3.216 3.239 rs_pw_transfer 974 11.9 0.009 0.010 2.522 2.746 init_scf_loop 11 6.9 0.001 0.004 2.579 2.580 calculate_first_density_matrix 1 7.0 0.001 0.004 2.437 2.439 density_rs2pw 119 9.7 0.004 0.004 2.155 2.403 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.063 2.373 apply_single 119 13.6 0.000 0.000 2.062 2.373 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.164 2.169 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.368 2.101 pw_transfer 1439 11.6 0.066 0.072 2.012 2.021 make_m2s 4572 13.5 0.044 0.045 1.837 1.981 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.918 1.932 jit_kernel_multiply 10 16.3 0.582 1.906 0.582 1.906 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.898 1.899 make_images 4572 14.5 0.190 0.203 1.753 1.896 grid_integrate_task_list 119 12.3 1.802 1.881 1.802 1.881 calculate_dm_sparse 119 9.5 0.000 0.000 1.803 1.819 potential_pw2rs 119 12.3 0.007 0.008 1.808 1.815 prepare_preconditioner 11 7.9 0.000 0.001 1.779 1.781 make_preconditioner 11 8.9 0.000 0.002 1.778 1.781 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.627 1.713 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.674 1.676 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.642 1.644 cp_fm_redistribute_end 50 14.0 1.229 1.624 1.230 1.625 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.613 1.618 fft3d_ps 1201 14.6 0.524 0.539 1.597 1.612 multiply_cannon_sync_h2d 18288 15.5 1.419 1.586 1.419 1.586 cp_fm_diag_elpa_base 50 14.0 0.378 1.536 0.392 1.574 fft_wrap_pw1pw2_140 487 13.2 0.090 0.093 1.531 1.544 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.522 1.535 mp_sum_l 7207 12.9 1.140 1.485 1.140 1.485 ot_diis_step 108 11.5 0.011 0.015 1.484 1.484 grid_collocate_task_list 119 9.7 1.215 1.358 1.215 1.358 acc_transpose_blocks 18288 15.5 0.075 0.077 1.237 1.259 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.181 1.189 wfi_extrapolate 11 7.9 0.001 0.001 1.185 1.185 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.037 1.046 make_images_data 4572 15.5 0.044 0.048 0.828 0.992 hybrid_alltoall_any 4725 16.4 0.054 0.113 0.711 0.902 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.859 0.881 mp_waitany 9880 13.7 0.587 0.830 0.587 0.830 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.819 0.821 rs_pw_transfer_RS2PW_140 130 11.5 0.120 0.125 0.586 0.820 acc_transpose_blocks_kernels 18288 16.5 0.211 0.222 0.796 0.817 mp_alltoall_z22v 1201 16.6 0.675 0.779 0.675 0.779 mp_alltoall_d11v 2130 13.8 0.672 0.771 0.672 0.771 cp_fm_cholesky_invert 11 10.9 0.731 0.736 0.731 0.736 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.659 0.730 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.613 0.664 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.202000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.454545, yerr=1.372697 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_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 551.890944E+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 0.017 0.039 36.480 36.481 qs_mol_dyn_low 1 2.0 0.014 0.023 36.174 36.181 qs_forces 11 3.9 0.025 0.050 36.085 36.090 qs_energies 11 4.9 0.002 0.007 34.307 34.314 scf_env_do_scf 11 5.9 0.001 0.006 28.703 28.704 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.290 25.291 dbcsr_multiply_generic 2286 12.5 0.101 0.102 18.672 18.817 velocity_verlet 10 3.0 0.001 0.001 17.662 17.665 qs_scf_new_mos 108 7.5 0.001 0.001 16.545 16.601 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.545 16.600 ot_scf_mini 108 9.5 0.002 0.003 15.605 15.657 multiply_cannon 2286 13.5 0.222 0.259 15.054 15.519 multiply_cannon_loop 2286 14.5 0.937 0.970 14.087 14.388 ot_mini 108 10.5 0.001 0.001 9.635 9.702 multiply_cannon_multrec 27432 15.5 2.451 3.193 8.980 9.389 qs_ot_get_derivative 108 11.5 0.001 0.001 7.822 7.877 dbcsr_mm_accdrv_process 47916 15.9 5.517 7.243 6.436 7.760 rebuild_ks_matrix 119 8.3 0.000 0.000 6.772 6.822 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 6.772 6.822 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.037 6.081 sum_up_and_integrate 119 10.3 0.035 0.038 3.943 3.950 init_scf_run 11 5.9 0.000 0.001 3.946 3.947 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.946 3.946 integrate_v_rspace 119 11.3 0.002 0.003 3.908 3.916 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.656 3.689 calculate_rho_elec 119 8.7 0.040 0.046 3.655 3.688 qs_ot_get_p 119 10.4 0.001 0.001 3.338 3.408 init_scf_loop 11 6.9 0.001 0.008 3.389 3.390 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.851 3.248 calculate_first_density_matrix 1 7.0 0.000 0.003 2.582 2.583 mp_waitall_1 137007 16.6 2.017 2.567 2.017 2.567 prepare_preconditioner 11 7.9 0.000 0.000 2.516 2.524 make_preconditioner 11 8.9 0.000 0.002 2.516 2.524 rs_pw_transfer 974 11.9 0.009 0.009 2.265 2.489 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.128 2.454 make_m2s 4572 13.5 0.054 0.055 2.221 2.350 density_rs2pw 119 9.7 0.004 0.004 2.065 2.263 calculate_dm_sparse 119 9.5 0.000 0.000 2.202 2.247 pw_transfer 1439 11.6 0.066 0.072 2.207 2.246 make_images 4572 14.5 0.271 0.331 2.114 2.242 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.114 2.158 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.029 2.154 apply_single 119 13.6 0.000 0.000 2.029 2.154 jit_kernel_multiply 10 16.1 0.858 2.136 0.858 2.136 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.012 2.040 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.992 2.003 grid_integrate_task_list 119 12.3 1.834 1.940 1.834 1.940 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.896 1.897 fft3d_ps 1201 14.6 0.570 0.621 1.787 1.829 fft_wrap_pw1pw2_140 487 13.2 0.089 0.097 1.732 1.779 ot_diis_step 108 11.5 0.012 0.012 1.772 1.772 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.708 1.721 potential_pw2rs 119 12.3 0.008 0.009 1.693 1.699 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.680 1.680 acc_transpose_blocks 27432 15.5 0.112 0.114 1.546 1.574 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.861 1.446 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.441 1.443 cp_fm_redistribute_end 50 14.0 0.957 1.419 0.958 1.420 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.378 1.387 cp_fm_diag_elpa_base 50 14.0 0.439 1.347 0.459 1.386 qs_energies_init_hamiltonians 11 5.9 0.003 0.016 1.379 1.381 grid_collocate_task_list 119 9.7 1.221 1.356 1.221 1.356 wfi_extrapolate 11 7.9 0.001 0.001 1.313 1.313 mp_sum_l 7207 12.9 0.972 1.263 0.972 1.263 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.105 1.125 cp_fm_upper_to_full 72 13.5 0.809 1.119 0.809 1.119 multiply_cannon_sync_h2d 27432 15.5 0.998 1.065 0.998 1.065 dbcsr_complete_redistribute 329 12.2 0.123 0.155 0.764 1.037 make_images_data 4572 15.5 0.045 0.049 0.867 1.008 acc_transpose_blocks_kernels 27432 16.5 0.269 0.278 0.919 0.939 hybrid_alltoall_any 4725 16.4 0.061 0.149 0.732 0.930 qs_env_update_s_mstruct 11 6.9 0.001 0.004 0.848 0.899 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.869 mp_alltoall_d11v 2130 13.8 0.720 0.863 0.720 0.863 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.160 0.638 0.850 mp_alltoall_z22v 1201 16.6 0.801 0.839 0.801 0.839 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.830 0.835 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.567 0.832 mp_waitany 7280 13.7 0.603 0.818 0.603 0.818 cp_fm_cholesky_invert 11 10.9 0.737 0.740 0.737 0.740 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.481000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=525.000000, yerr=2.132007 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_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 601.440256E+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 0.018 0.030 28.573 28.575 qs_mol_dyn_low 1 2.0 0.003 0.004 28.364 28.372 qs_forces 11 3.9 0.002 0.002 28.290 28.291 qs_energies 11 4.9 0.001 0.001 26.580 26.583 scf_env_do_scf 11 5.9 0.000 0.001 21.905 21.905 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.419 19.420 velocity_verlet 10 3.0 0.001 0.001 14.292 14.296 dbcsr_multiply_generic 2286 12.5 0.093 0.096 12.330 12.466 qs_scf_new_mos 108 7.5 0.001 0.001 11.185 11.213 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.184 11.212 ot_scf_mini 108 9.5 0.002 0.002 10.497 10.527 multiply_cannon 2286 13.5 0.231 0.236 9.733 10.305 multiply_cannon_loop 2286 14.5 0.330 0.342 8.820 9.104 rebuild_ks_matrix 119 8.3 0.000 0.000 6.130 6.159 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 6.129 6.158 multiply_cannon_multrec 9144 15.5 1.597 1.868 5.756 5.968 ot_mini 108 10.5 0.001 0.001 5.921 5.956 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.480 5.505 qs_ot_get_derivative 108 11.5 0.001 0.001 4.642 4.672 dbcsr_mm_accdrv_process 12550 15.8 3.101 4.124 4.058 4.164 sum_up_and_integrate 119 10.3 0.038 0.041 3.771 3.774 integrate_v_rspace 119 11.3 0.003 0.003 3.733 3.737 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.632 3.639 calculate_rho_elec 119 8.7 0.059 0.061 3.632 3.639 init_scf_run 11 5.9 0.000 0.001 3.245 3.245 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.245 3.245 qs_ot_get_p 119 10.4 0.001 0.001 2.782 2.821 init_scf_loop 11 6.9 0.000 0.000 2.466 2.467 pw_transfer 1439 11.6 0.066 0.070 2.210 2.223 mp_waitall_1 115863 16.7 1.688 2.214 1.688 2.214 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.116 2.130 density_rs2pw 119 9.7 0.004 0.004 1.979 2.123 calculate_first_density_matrix 1 7.0 0.000 0.000 2.104 2.105 make_m2s 4572 13.5 0.034 0.035 1.780 1.950 grid_integrate_task_list 119 12.3 1.859 1.930 1.859 1.930 rs_pw_transfer 974 11.9 0.008 0.008 1.763 1.884 make_images 4572 14.5 0.269 0.307 1.691 1.859 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.819 1.822 fft3d_ps 1201 14.6 0.561 0.572 1.774 1.787 fft_wrap_pw1pw2_140 487 13.2 0.087 0.091 1.748 1.762 jit_kernel_multiply 10 15.4 0.919 1.752 0.919 1.752 calculate_dm_sparse 119 9.5 0.000 0.000 1.712 1.727 prepare_preconditioner 11 7.9 0.000 0.000 1.715 1.720 make_preconditioner 11 8.9 0.000 0.000 1.715 1.720 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.589 1.621 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.605 1.606 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.540 1.542 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.496 1.514 potential_pw2rs 119 12.3 0.010 0.011 1.476 1.478 grid_collocate_task_list 119 9.7 1.276 1.393 1.276 1.393 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.378 1.388 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.338 1.339 cp_fm_redistribute_end 50 14.0 0.668 1.309 0.669 1.309 cp_fm_diag_elpa_base 50 14.0 0.594 1.226 0.638 1.283 ot_diis_step 108 11.5 0.013 0.013 1.266 1.266 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 1.223 1.223 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.203 1.220 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.140 1.168 apply_single 119 13.6 0.000 0.000 1.140 1.167 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.117 1.121 wfi_extrapolate 11 7.9 0.001 0.001 1.092 1.092 make_images_data 4572 15.5 0.038 0.042 0.839 1.055 hybrid_alltoall_any 4725 16.4 0.062 0.174 0.781 1.046 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.921 acc_transpose_blocks 9144 15.5 0.038 0.039 0.904 0.912 cp_fm_cholesky_invert 11 10.9 0.838 0.840 0.838 0.840 mp_alltoall_d11v 2130 13.8 0.754 0.831 0.754 0.831 multiply_cannon_sync_h2d 9144 15.5 0.710 0.790 0.710 0.790 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.773 0.776 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.727 0.737 qs_env_update_s_mstruct 11 6.9 0.002 0.003 0.677 0.728 mp_alltoall_z22v 1201 16.6 0.660 0.719 0.660 0.719 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.398 0.712 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 0.436 0.671 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.664 0.668 mp_allgather_i34 2286 14.5 0.246 0.668 0.246 0.668 yz_to_x 606 15.1 0.265 0.276 0.614 0.632 x_to_yz 595 16.2 0.275 0.289 0.587 0.593 mp_waitany 5200 13.7 0.466 0.578 0.466 0.578 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.575000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=570.000000, yerr=4.264014 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_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 753.643520E+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 0.026 0.066 41.970 41.979 qs_mol_dyn_low 1 2.0 0.003 0.005 41.742 41.749 qs_forces 11 3.9 0.002 0.002 40.587 40.589 qs_energies 11 4.9 0.001 0.001 38.605 38.608 scf_env_do_scf 11 5.9 0.001 0.001 32.919 32.920 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.199 25.199 velocity_verlet 10 3.0 0.001 0.001 24.024 24.166 dbcsr_multiply_generic 2286 12.5 0.100 0.101 17.112 17.277 qs_scf_new_mos 108 7.5 0.001 0.001 15.408 15.500 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.407 15.500 ot_scf_mini 108 9.5 0.002 0.002 14.347 14.442 multiply_cannon 2286 13.5 0.301 0.308 13.505 14.367 multiply_cannon_loop 2286 14.5 0.343 0.351 12.290 13.191 ot_mini 108 10.5 0.001 0.001 8.598 8.710 multiply_cannon_multrec 9144 15.5 3.350 4.814 8.535 8.612 init_scf_loop 11 6.9 0.000 0.000 7.694 7.695 rebuild_ks_matrix 119 8.3 0.000 0.001 7.313 7.449 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 7.312 7.449 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.613 6.737 prepare_preconditioner 11 7.9 0.000 0.000 6.720 6.734 make_preconditioner 11 8.9 0.000 0.000 6.720 6.733 qs_ot_get_derivative 108 11.5 0.001 0.001 6.625 6.721 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.276 6.616 dbcsr_mm_accdrv_process 12550 15.8 4.067 5.588 5.063 6.309 cp_fm_upper_to_full 72 14.2 3.171 4.563 3.171 4.563 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.314 4.367 calculate_rho_elec 119 8.7 0.118 0.121 4.313 4.367 sum_up_and_integrate 119 10.3 0.065 0.066 4.019 4.026 integrate_v_rspace 119 11.3 0.003 0.003 3.954 3.961 init_scf_run 11 5.9 0.000 0.001 3.650 3.650 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.649 3.650 qs_ot_get_p 119 10.4 0.001 0.001 3.075 3.205 mp_waitall_1 94719 16.7 2.199 3.123 2.199 3.123 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.437 2.835 pw_transfer 1439 11.6 0.069 0.069 2.812 2.817 dbcsr_complete_redistribute 329 12.2 0.291 0.296 1.956 2.768 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.713 2.718 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.661 2.475 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.120 2.427 apply_single 119 13.6 0.000 0.000 2.120 2.427 fft3d_ps 1201 14.6 0.601 0.610 2.335 2.339 fft_wrap_pw1pw2_140 487 13.2 0.095 0.097 2.298 2.306 mp_alltoall_i22 627 13.8 1.430 2.264 1.430 2.264 make_m2s 4572 13.5 0.038 0.038 2.121 2.261 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.439 2.246 calculate_dm_sparse 119 9.5 0.000 0.000 2.181 2.234 calculate_first_density_matrix 1 7.0 0.000 0.000 2.175 2.231 density_rs2pw 119 9.7 0.004 0.004 2.189 2.210 make_images 4572 14.5 0.355 0.388 2.000 2.139 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.291 2.119 grid_integrate_task_list 119 12.3 2.048 2.068 2.048 2.068 ot_diis_step 108 11.5 0.014 0.014 1.950 1.950 qs_ot_p2m_diag 50 11.0 0.043 0.043 1.867 1.868 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.824 1.825 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 1.774 1.776 jit_kernel_multiply 10 15.4 0.969 1.725 0.969 1.725 mp_sum_l 7207 12.9 0.958 1.712 0.958 1.712 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.636 1.685 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.598 1.598 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.566 1.589 grid_collocate_task_list 119 9.7 1.482 1.494 1.482 1.494 rs_pw_transfer 974 11.9 0.009 0.010 1.413 1.434 potential_pw2rs 119 12.3 0.014 0.014 1.421 1.424 cp_fm_cholesky_invert 11 10.9 1.388 1.391 1.388 1.391 wfi_extrapolate 11 7.9 0.001 0.001 1.357 1.357 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.322 1.323 cp_fm_diag_elpa_base 50 14.0 1.179 1.230 1.320 1.321 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.267 1.320 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.039 1.229 mp_sum_dm 438 4.9 1.019 1.173 1.019 1.173 mp_alltoall_d11v 2130 13.8 1.110 1.167 1.110 1.167 md_write_output 11 3.9 0.083 0.663 0.144 1.141 md_output 10 3.0 0.000 0.000 0.145 1.139 update_particle_set 20 4.0 0.000 0.000 0.994 1.137 make_images_data 4572 15.5 0.043 0.046 0.962 1.134 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.109 1.127 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.110 1.126 multiply_cannon_sync_h2d 9144 15.5 1.043 1.046 1.043 1.046 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.982 1.022 qs_create_task_list 11 7.9 0.001 0.001 0.942 0.953 generate_qs_task_list 11 8.9 0.369 0.389 0.941 0.952 yz_to_x 606 15.1 0.460 0.471 0.933 0.942 acc_transpose_blocks 9144 15.5 0.038 0.038 0.929 0.934 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.911 0.923 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.979000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=706.181818, yerr=18.009180 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_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 504.053760E+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 0.014 0.036 86.710 86.711 qs_mol_dyn_low 1 2.0 0.003 0.005 86.434 86.444 qs_forces 11 3.9 0.002 0.002 85.393 85.394 qs_energies 11 4.9 0.001 0.002 82.388 82.410 scf_env_do_scf 11 5.9 0.000 0.001 73.192 73.195 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 67.336 67.338 dbcsr_multiply_generic 2055 12.4 0.108 0.111 53.285 53.532 qs_scf_new_mos 99 7.5 0.000 0.001 49.386 49.508 qs_scf_loop_do_ot 99 8.5 0.001 0.001 49.385 49.507 ot_scf_mini 99 9.5 0.002 0.002 46.936 47.033 velocity_verlet 10 3.0 0.001 0.001 45.684 45.695 multiply_cannon 2055 13.4 0.175 0.181 43.157 44.103 multiply_cannon_loop 2055 14.4 1.550 1.591 42.154 43.103 ot_mini 99 10.5 0.001 0.001 28.351 28.452 qs_ot_get_derivative 99 11.5 0.001 0.001 21.382 21.480 multiply_cannon_multrec 49320 15.4 12.046 12.884 17.097 17.746 rebuild_ks_matrix 110 8.3 0.000 0.001 15.013 15.107 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 15.012 15.106 mp_waitall_1 220248 16.4 12.405 13.611 12.405 13.611 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.085 13.168 multiply_cannon_sync_h2d 49320 15.4 9.902 10.657 9.902 10.657 qs_ot_get_p 110 10.4 0.001 0.001 9.754 9.873 multiply_cannon_metrocomm3 49320 15.4 0.079 0.085 7.063 8.367 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.496 8.053 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.261 7.729 apply_single 110 13.6 0.000 0.001 7.261 7.729 sum_up_and_integrate 110 10.3 0.036 0.043 7.429 7.443 integrate_v_rspace 110 11.3 0.002 0.003 7.392 7.414 init_scf_run 11 5.9 0.000 0.001 7.101 7.101 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.100 7.101 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.927 7.075 calculate_rho_elec 110 8.6 0.021 0.026 6.927 7.074 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.790 6.845 ot_diis_step 99 11.5 0.006 0.006 6.659 6.660 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.379 6.420 mp_sum_l 6514 12.8 5.374 6.112 5.374 6.112 init_scf_loop 11 6.9 0.000 0.000 5.829 5.831 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.381 5.381 dbcsr_mm_accdrv_process 87628 16.1 2.100 2.205 4.930 5.234 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.927 4.929 cp_fm_redistribute_end 48 14.0 4.303 4.906 4.306 4.907 cp_fm_diag_elpa_base 48 14.0 0.594 4.771 0.597 4.802 rs_pw_transfer 902 11.9 0.011 0.013 3.854 4.520 make_m2s 4110 13.4 0.060 0.066 4.290 4.404 wfi_extrapolate 11 7.9 0.001 0.001 4.369 4.369 multiply_cannon_metrocomm1 49320 15.4 0.060 0.064 3.255 4.340 density_rs2pw 110 9.6 0.004 0.005 3.665 4.329 make_images 4110 14.4 0.177 0.190 4.195 4.314 calculate_dm_sparse 110 9.5 0.000 0.001 3.812 3.903 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.733 3.739 prepare_preconditioner 11 7.9 0.000 0.000 3.571 3.588 make_preconditioner 11 8.9 0.000 0.000 3.571 3.588 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.493 3.577 pw_transfer 1331 11.6 0.054 0.064 3.494 3.567 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.406 3.480 grid_integrate_task_list 110 12.3 3.239 3.401 3.239 3.401 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.344 3.390 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.297 3.340 fft3d_ps 1111 14.6 0.781 0.883 2.934 3.002 fft_wrap_pw1pw2_140 451 13.1 0.171 0.190 2.845 2.921 potential_pw2rs 110 12.3 0.006 0.007 2.799 2.817 mp_alltoall_d11v 2046 13.8 2.415 2.814 2.415 2.814 calculate_first_density_matrix 1 7.0 0.000 0.000 2.637 2.646 mp_waitany 14300 13.8 1.929 2.633 1.929 2.633 jit_kernel_multiply 13 15.9 2.547 2.563 2.547 2.563 acc_transpose_blocks 49320 15.4 0.225 0.236 2.220 2.300 grid_collocate_task_list 110 9.6 2.083 2.286 2.083 2.286 make_images_data 4110 15.4 0.042 0.045 1.974 2.129 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.070 2.089 mp_sum_d 3883 11.9 1.563 2.086 1.563 2.086 hybrid_alltoall_any 4261 16.3 0.081 0.479 1.687 1.998 cp_fm_cholesky_invert 11 10.9 1.915 1.918 1.915 1.918 mp_alltoall_z22v 1111 16.6 1.524 1.796 1.524 1.796 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=86.711000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.818182, yerr=2.917601 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_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 589.271040E+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 1220486. 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 0.015 0.029 68.176 68.177 qs_mol_dyn_low 1 2.0 0.004 0.006 67.913 67.947 qs_forces 11 3.9 0.002 0.005 67.839 67.843 qs_energies 11 4.9 0.003 0.013 64.516 64.521 scf_env_do_scf 11 5.9 0.001 0.001 55.986 55.989 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.501 48.501 dbcsr_multiply_generic 2055 12.4 0.114 0.118 37.298 37.478 velocity_verlet 10 3.0 0.001 0.001 35.915 35.916 qs_scf_new_mos 99 7.5 0.001 0.001 32.206 32.337 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.206 32.337 multiply_cannon 2055 13.4 0.223 0.245 30.894 32.302 multiply_cannon_loop 2055 14.4 0.935 0.956 29.633 30.718 ot_scf_mini 99 9.5 0.003 0.004 30.560 30.684 ot_mini 99 10.5 0.001 0.001 17.965 18.096 multiply_cannon_multrec 24660 15.4 7.642 9.166 13.824 15.598 rebuild_ks_matrix 110 8.3 0.000 0.000 13.778 13.907 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.777 13.907 qs_ot_get_derivative 99 11.5 0.001 0.001 12.182 12.306 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.128 12.241 mp_waitall_1 176588 16.5 7.439 9.877 7.439 9.877 multiply_cannon_sync_h2d 24660 15.4 7.015 8.077 7.015 8.077 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.092 7.687 init_scf_loop 11 6.9 0.000 0.001 7.451 7.451 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.500 7.250 apply_single 110 13.6 0.000 0.001 6.500 7.250 sum_up_and_integrate 110 10.3 0.051 0.058 6.605 6.618 integrate_v_rspace 110 11.3 0.002 0.002 6.554 6.567 dbcsr_mm_accdrv_process 52282 16.1 4.679 5.650 6.022 6.345 qs_ot_get_p 110 10.4 0.001 0.001 6.006 6.174 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.154 6.167 calculate_rho_elec 110 8.6 0.039 0.047 6.154 6.167 init_scf_run 11 5.9 0.000 0.001 6.078 6.079 scf_env_initial_rho_setup 11 6.9 0.002 0.003 6.078 6.079 ot_diis_step 99 11.5 0.010 0.010 5.737 5.737 prepare_preconditioner 11 7.9 0.000 0.000 5.436 5.460 make_preconditioner 11 8.9 0.000 0.001 5.436 5.460 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.605 5.319 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.038 5.200 make_m2s 4110 13.4 0.057 0.060 4.050 4.523 make_images 4110 14.4 0.398 0.444 3.941 4.410 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.188 4.207 pw_transfer 1331 11.6 0.066 0.073 3.733 3.873 density_rs2pw 110 9.6 0.004 0.004 3.281 3.769 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.626 3.768 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.766 3.767 wfi_extrapolate 11 7.9 0.001 0.001 3.496 3.496 rs_pw_transfer 902 11.9 0.012 0.014 2.942 3.449 grid_integrate_task_list 110 12.3 3.142 3.348 3.142 3.348 fft_wrap_pw1pw2_140 451 13.1 0.204 0.221 3.118 3.263 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.253 3.255 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.244 3.245 cp_fm_redistribute_end 48 14.0 2.435 3.232 2.436 3.232 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.109 3.184 cp_fm_diag_elpa_base 48 14.0 0.761 3.094 0.792 3.178 fft3d_ps 1111 14.6 1.106 1.321 2.973 3.109 calculate_dm_sparse 110 9.5 0.001 0.001 2.945 2.976 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.904 2.960 hybrid_alltoall_any 4261 16.3 0.101 0.442 2.021 2.755 make_images_data 4110 15.4 0.046 0.050 2.276 2.753 cp_fm_cholesky_invert 11 10.9 2.505 2.511 2.505 2.511 calculate_first_density_matrix 1 7.0 0.000 0.000 2.489 2.494 grid_collocate_task_list 110 9.6 2.051 2.478 2.051 2.478 potential_pw2rs 110 12.3 0.008 0.009 2.401 2.409 mp_sum_l 6514 12.8 1.727 2.328 1.727 2.328 mp_alltoall_d11v 2046 13.8 1.745 2.109 1.745 2.109 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.924 1.947 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 1.940 1.944 jit_kernel_multiply 10 16.4 0.993 1.909 0.993 1.909 multiply_cannon_metrocomm4 22605 15.4 0.072 0.077 0.784 1.885 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.813 1.828 mp_irecv_dv 57340 16.2 0.662 1.768 0.662 1.768 mp_waitany 10164 13.8 1.253 1.746 1.253 1.746 acc_transpose_blocks 24660 15.4 0.111 0.116 1.554 1.580 mp_allgather_i34 2055 14.4 0.590 1.573 0.590 1.573 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.530 1.542 rs_pw_transfer_RS2PW_140 121 11.5 0.217 0.239 1.010 1.514 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.365 1.470 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.177000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.545455, yerr=7.785343 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_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 657.981440E+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 0.115 0.229 60.405 60.406 qs_mol_dyn_low 1 2.0 0.006 0.012 59.522 59.531 qs_forces 11 3.9 0.007 0.024 59.449 59.453 qs_energies 11 4.9 0.002 0.011 56.179 56.188 scf_env_do_scf 11 5.9 0.001 0.002 47.973 47.973 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 39.612 39.613 velocity_verlet 10 3.0 0.001 0.001 32.313 32.314 dbcsr_multiply_generic 2055 12.4 0.108 0.113 28.239 28.513 qs_scf_new_mos 99 7.5 0.001 0.001 24.501 24.595 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.501 24.594 ot_scf_mini 99 9.5 0.002 0.003 23.293 23.406 multiply_cannon 2055 13.4 0.212 0.223 22.089 23.306 multiply_cannon_loop 2055 14.4 0.616 0.634 20.891 21.972 ot_mini 99 10.5 0.001 0.001 13.317 13.431 rebuild_ks_matrix 110 8.3 0.000 0.000 12.451 12.571 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.450 12.571 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.966 11.075 multiply_cannon_multrec 16440 15.4 3.881 5.102 9.681 10.839 mp_waitall_1 139946 16.5 6.873 10.301 6.873 10.301 qs_ot_get_derivative 99 11.5 0.001 0.001 8.905 9.019 init_scf_loop 11 6.9 0.001 0.005 8.327 8.327 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.325 7.507 sum_up_and_integrate 110 10.3 0.060 0.060 6.550 6.566 prepare_preconditioner 11 7.9 0.000 0.001 6.541 6.555 make_preconditioner 11 8.9 0.000 0.002 6.541 6.555 integrate_v_rspace 110 11.3 0.003 0.005 6.490 6.506 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.899 6.244 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.079 6.088 calculate_rho_elec 110 8.6 0.063 0.094 6.079 6.087 dbcsr_mm_accdrv_process 34862 16.1 4.660 5.271 5.655 5.758 init_scf_run 11 5.9 0.000 0.001 5.699 5.699 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.698 5.699 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.941 5.377 apply_single 110 13.6 0.000 0.000 4.941 5.377 qs_ot_get_p 110 10.4 0.001 0.001 5.163 5.294 multiply_cannon_sync_h2d 16440 15.4 3.688 4.507 3.688 4.507 make_m2s 4110 13.4 0.050 0.052 4.083 4.493 ot_diis_step 99 11.5 0.010 0.011 4.385 4.385 make_images 4110 14.4 0.392 0.509 3.969 4.378 density_rs2pw 110 9.6 0.004 0.005 3.135 4.370 rs_pw_transfer 902 11.9 0.010 0.011 2.709 3.934 pw_transfer 1331 11.6 0.066 0.073 3.718 3.726 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.045 3.694 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.624 3.628 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.611 3.622 grid_integrate_task_list 110 12.3 3.164 3.416 3.164 3.416 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.294 3.294 fft_wrap_pw1pw2_140 451 13.1 0.213 0.218 3.133 3.143 wfi_extrapolate 11 7.9 0.001 0.001 2.953 2.953 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.922 2.923 fft3d_ps 1111 14.6 1.087 1.097 2.910 2.920 make_images_data 4110 15.4 0.043 0.047 2.366 2.878 hybrid_alltoall_any 4261 16.3 0.105 0.373 2.126 2.851 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.797 2.798 cp_fm_redistribute_end 48 14.0 1.747 2.776 1.749 2.776 cp_fm_diag_elpa_base 48 14.0 0.966 2.631 1.024 2.736 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.525 2.576 calculate_first_density_matrix 1 7.0 0.001 0.005 2.571 2.573 cp_fm_cholesky_invert 11 10.9 2.563 2.569 2.563 2.569 mp_waitany 17072 13.8 1.271 2.549 1.271 2.549 calculate_dm_sparse 110 9.5 0.001 0.001 2.479 2.520 grid_collocate_task_list 110 9.6 2.082 2.480 2.082 2.480 multiply_cannon_metrocomm4 14385 15.4 0.044 0.048 0.892 2.459 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.268 2.334 mp_irecv_dv 48980 15.7 0.823 2.332 0.823 2.332 potential_pw2rs 110 12.3 0.011 0.011 2.278 2.285 mp_alltoall_d11v 2046 13.8 1.764 2.216 1.764 2.216 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.180 0.994 2.212 mp_sum_l 6514 12.8 1.468 2.097 1.468 2.097 qs_energies_init_hamiltonians 11 5.9 0.002 0.007 2.035 2.042 dbcsr_complete_redistribute 325 12.2 0.348 0.402 1.425 1.901 cp_fm_upper_to_full 70 13.6 1.386 1.864 1.386 1.864 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.685 1.701 mp_allgather_i34 2055 14.4 0.524 1.570 0.524 1.570 cp_fm_cholesky_decompose 22 10.9 1.515 1.537 1.515 1.537 jit_kernel_multiply 8 16.7 0.607 1.535 0.607 1.535 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.365 1.532 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.450 1.464 copy_fm_to_dbcsr 174 11.2 0.001 0.002 1.001 1.457 rs_gather_matrices 110 12.3 0.233 0.263 0.971 1.355 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.233 1.243 acc_transpose_blocks 16440 15.4 0.073 0.074 1.214 1.235 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.128 1.219 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.406000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=622.090909, yerr=7.936733 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 730.050560E+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 0.024 0.054 65.168 65.169 qs_mol_dyn_low 1 2.0 0.005 0.015 64.817 64.845 qs_forces 11 3.9 0.003 0.008 64.673 64.675 qs_energies 11 4.9 0.003 0.017 61.254 61.260 scf_env_do_scf 11 5.9 0.001 0.001 52.943 52.946 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.339 41.340 velocity_verlet 10 3.0 0.001 0.002 36.762 36.766 dbcsr_multiply_generic 2055 12.4 0.116 0.118 29.376 29.634 qs_scf_new_mos 99 7.5 0.001 0.001 26.165 26.269 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.164 26.268 ot_scf_mini 99 9.5 0.003 0.004 24.539 24.646 multiply_cannon 2055 13.4 0.246 0.264 22.636 23.682 multiply_cannon_loop 2055 14.4 0.886 0.907 21.092 21.719 ot_mini 99 10.5 0.001 0.001 14.075 14.207 multiply_cannon_multrec 24660 15.4 4.246 6.828 12.678 13.879 rebuild_ks_matrix 110 8.3 0.000 0.000 12.303 12.397 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.303 12.396 init_scf_loop 11 6.9 0.001 0.006 11.564 11.565 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.866 10.946 qs_ot_get_derivative 99 11.5 0.001 0.001 9.954 10.067 prepare_preconditioner 11 7.9 0.000 0.001 9.790 9.805 make_preconditioner 11 8.9 0.000 0.003 9.790 9.805 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.011 9.489 dbcsr_mm_accdrv_process 52304 16.0 6.858 8.410 8.286 9.235 sum_up_and_integrate 110 10.3 0.067 0.070 6.560 6.573 integrate_v_rspace 110 11.3 0.003 0.004 6.493 6.508 mp_waitall_1 121746 16.5 4.204 6.414 4.204 6.414 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.188 6.199 calculate_rho_elec 110 8.6 0.077 0.081 6.188 6.198 make_m2s 4110 13.4 0.059 0.061 5.179 5.666 qs_ot_get_p 110 10.4 0.001 0.001 5.468 5.626 make_images 4110 14.4 0.577 0.703 5.039 5.523 init_scf_run 11 5.9 0.000 0.001 5.504 5.504 scf_env_initial_rho_setup 11 6.9 0.003 0.003 5.503 5.504 cp_fm_upper_to_full 70 13.8 3.300 4.700 3.300 4.700 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.009 4.102 apply_single 110 13.6 0.000 0.000 4.009 4.102 ot_diis_step 99 11.5 0.011 0.011 4.086 4.087 pw_transfer 1331 11.6 0.065 0.074 3.946 3.974 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.840 3.871 dbcsr_complete_redistribute 325 12.2 0.417 0.467 2.621 3.775 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.748 3.762 density_rs2pw 110 9.6 0.004 0.004 3.105 3.613 grid_integrate_task_list 110 12.3 3.246 3.430 3.246 3.430 fft_wrap_pw1pw2_140 451 13.1 0.204 0.213 3.300 3.336 multiply_cannon_sync_h2d 24660 15.4 3.166 3.286 3.166 3.286 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.281 3.281 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.223 3.276 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.120 3.259 fft3d_ps 1111 14.6 1.092 1.132 3.128 3.154 hybrid_alltoall_any 4261 16.3 0.119 0.459 2.297 3.139 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.477 3.099 make_images_data 4110 15.4 0.046 0.050 2.572 3.054 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.984 3.048 wfi_extrapolate 11 7.9 0.001 0.001 2.963 2.963 calculate_dm_sparse 110 9.5 0.001 0.001 2.913 2.947 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.769 2.898 rs_pw_transfer 902 11.9 0.010 0.011 2.371 2.889 mp_alltoall_i22 605 13.7 1.647 2.810 1.647 2.810 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.794 2.797 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.707 2.708 cp_fm_redistribute_end 48 14.0 1.355 2.684 1.357 2.685 cp_fm_diag_elpa_base 48 14.0 1.247 2.552 1.326 2.662 cp_fm_cholesky_invert 11 10.9 2.636 2.644 2.636 2.644 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.455 2.490 calculate_first_density_matrix 1 7.0 0.001 0.003 2.429 2.434 grid_collocate_task_list 110 9.6 2.181 2.423 2.181 2.423 jit_kernel_multiply 13 15.9 1.098 2.381 1.098 2.381 qs_energies_init_hamiltonians 11 5.9 0.002 0.006 2.303 2.306 potential_pw2rs 110 12.3 0.012 0.013 2.207 2.215 mp_alltoall_d11v 2046 13.8 1.809 2.132 1.809 2.132 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.682 1.718 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.598 1.699 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.683 1.695 cp_fm_cholesky_decompose 22 10.9 1.627 1.680 1.627 1.680 mp_waitany 13376 13.8 1.072 1.600 1.072 1.600 mp_allgather_i34 2055 14.4 0.643 1.563 0.643 1.563 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.522 1.543 acc_transpose_blocks 24660 15.4 0.107 0.109 1.509 1.528 mp_sum_l 6514 12.8 0.898 1.517 0.898 1.517 multiply_cannon_metrocomm4 20550 15.4 0.057 0.059 0.839 1.445 mp_irecv_dv 62702 16.1 0.741 1.370 0.741 1.370 rs_pw_transfer_RS2PW_140 121 11.5 0.170 0.178 0.839 1.363 qs_env_update_s_mstruct 11 6.9 0.001 0.005 1.246 1.348 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.169000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=691.909091, yerr=10.492028 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/13/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 807.299199E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1438408 0.0% 0.0% 100.0% average stack size 0.0 0.0 684.2 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 847.081472E+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 0.017 0.030 55.456 55.457 qs_mol_dyn_low 1 2.0 0.003 0.004 55.215 55.225 qs_forces 11 3.9 0.002 0.002 55.146 55.149 qs_energies 11 4.9 0.001 0.002 51.445 51.451 scf_env_do_scf 11 5.9 0.000 0.001 43.201 43.201 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.451 35.452 velocity_verlet 10 3.0 0.001 0.001 31.055 31.058 dbcsr_multiply_generic 2055 12.4 0.105 0.106 22.626 22.775 qs_scf_new_mos 99 7.5 0.001 0.001 20.115 20.164 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.114 20.164 ot_scf_mini 99 9.5 0.002 0.002 18.884 18.907 multiply_cannon 2055 13.4 0.248 0.266 17.231 18.573 multiply_cannon_loop 2055 14.4 0.321 0.335 15.901 16.204 rebuild_ks_matrix 110 8.3 0.000 0.000 12.061 12.080 qs_ks_build_kohn_sham_matrix 110 9.3 0.016 0.017 12.060 12.080 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.719 10.739 ot_mini 99 10.5 0.001 0.001 10.198 10.215 multiply_cannon_multrec 8220 15.4 3.220 4.434 7.452 8.416 init_scf_loop 11 6.9 0.000 0.000 7.701 7.703 mp_waitall_1 103326 16.6 5.932 7.455 5.932 7.455 sum_up_and_integrate 110 10.3 0.079 0.080 6.578 6.587 integrate_v_rspace 110 11.3 0.003 0.003 6.499 6.509 qs_ot_get_derivative 99 11.5 0.001 0.001 6.469 6.494 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.477 6.490 calculate_rho_elec 110 8.6 0.115 0.116 6.476 6.489 prepare_preconditioner 11 7.9 0.000 0.000 6.023 6.027 make_preconditioner 11 8.9 0.000 0.000 6.023 6.027 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.608 5.684 init_scf_run 11 5.9 0.000 0.001 5.111 5.111 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.110 5.111 dbcsr_mm_accdrv_process 17442 15.9 2.876 3.960 4.103 5.001 qs_ot_get_p 110 10.4 0.001 0.001 4.787 4.802 make_m2s 4110 13.4 0.038 0.040 4.153 4.410 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.968 4.400 make_images 4110 14.4 0.639 0.689 4.023 4.279 pw_transfer 1331 11.6 0.066 0.073 4.212 4.230 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.105 4.126 ot_diis_step 99 11.5 0.012 0.012 3.707 3.707 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.633 3.656 apply_single 110 13.6 0.000 0.000 3.633 3.656 fft_wrap_pw1pw2_140 451 13.1 0.217 0.220 3.549 3.569 density_rs2pw 110 9.6 0.004 0.004 3.182 3.562 grid_integrate_task_list 110 12.3 3.352 3.547 3.352 3.547 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.434 3.437 fft3d_ps 1111 14.6 1.154 1.199 3.345 3.367 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.139 3.139 multiply_cannon_sync_h2d 8220 15.4 2.917 3.031 2.917 3.031 cp_fm_cholesky_invert 11 10.9 2.877 2.881 2.877 2.881 make_images_data 4110 15.4 0.038 0.043 2.374 2.796 hybrid_alltoall_any 4261 16.3 0.199 0.862 2.319 2.738 wfi_extrapolate 11 7.9 0.001 0.001 2.713 2.713 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.691 2.694 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.639 2.641 cp_fm_redistribute_end 48 14.0 0.670 2.615 0.674 2.615 cp_fm_diag_elpa_base 48 14.0 1.762 2.428 1.934 2.590 rs_pw_transfer 902 11.9 0.010 0.010 2.161 2.567 grid_collocate_task_list 110 9.6 2.277 2.542 2.277 2.542 calculate_dm_sparse 110 9.5 0.001 0.001 2.472 2.522 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.515 2.517 calculate_first_density_matrix 1 7.0 0.000 0.000 2.281 2.282 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.174 2.180 potential_pw2rs 110 12.3 0.015 0.016 2.116 2.122 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 1.995 mp_alltoall_d11v 2046 13.8 1.659 1.920 1.659 1.920 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.870 1.883 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.797 1.808 cp_fm_cholesky_decompose 22 10.9 1.682 1.696 1.682 1.696 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.676 1.678 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.528 1.648 mp_allgather_i34 2055 14.4 0.515 1.603 0.515 1.603 mp_waitany 9240 13.8 1.066 1.500 1.066 1.500 dbcsr_complete_redistribute 325 12.2 0.564 0.583 1.415 1.499 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.441 1.454 jit_kernel_multiply 8 15.8 0.920 1.381 0.920 1.381 qs_create_task_list 11 7.9 0.000 0.001 1.230 1.333 generate_qs_task_list 11 8.9 0.380 0.449 1.230 1.332 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.764 1.237 rs_pw_transfer_RS2PW_140 121 11.5 0.161 0.164 0.797 1.206 rs_gather_matrices 110 12.3 0.325 0.368 0.942 1.189 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.138 1.150 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.484 1.136 yz_to_x 561 15.1 0.524 0.537 1.104 1.117 mp_alltoall_z22v 1111 16.6 1.082 1.112 1.082 1.112 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.457000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=791.090909, yerr=15.535616 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_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 1.312059E+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 0.050 0.080 89.083 89.091 qs_mol_dyn_low 1 2.0 0.003 0.004 88.603 88.624 qs_forces 11 3.9 0.002 0.002 87.812 87.812 qs_energies 11 4.9 0.002 0.004 83.612 83.615 scf_env_do_scf 11 5.9 0.001 0.001 73.084 73.084 velocity_verlet 10 3.0 0.001 0.001 56.730 56.826 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 44.745 44.747 dbcsr_multiply_generic 2055 12.4 0.119 0.120 29.220 29.266 init_scf_loop 11 6.9 0.000 0.000 28.267 28.270 qs_scf_new_mos 99 7.5 0.001 0.001 26.199 26.235 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.198 26.234 prepare_preconditioner 11 7.9 0.000 0.000 26.213 26.218 make_preconditioner 11 8.9 0.000 0.000 26.213 26.218 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.461 25.695 ot_scf_mini 99 9.5 0.002 0.002 24.422 24.446 multiply_cannon 2055 13.4 0.372 0.405 22.083 22.767 multiply_cannon_loop 2055 14.4 0.342 0.346 20.261 20.704 cp_fm_upper_to_full 70 14.2 12.615 18.086 12.615 18.086 rebuild_ks_matrix 110 8.3 0.001 0.001 14.336 14.381 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.335 14.380 ot_mini 99 10.5 0.001 0.001 13.694 13.718 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.991 13.031 dbcsr_complete_redistribute 325 12.2 1.021 1.032 7.330 10.489 multiply_cannon_multrec 8220 15.4 4.475 4.679 9.991 10.082 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.298 9.454 qs_ot_get_derivative 99 11.5 0.001 0.001 9.078 9.106 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.736 8.872 mp_waitall_1 84994 16.7 7.748 8.607 7.748 8.607 mp_alltoall_i22 605 13.7 5.372 8.522 5.372 8.522 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.824 7.859 calculate_rho_elec 110 8.6 0.227 0.227 7.823 7.858 sum_up_and_integrate 110 10.3 0.150 0.151 7.418 7.435 integrate_v_rspace 110 11.3 0.004 0.004 7.268 7.285 init_scf_run 11 5.9 0.000 0.001 6.208 6.208 scf_env_initial_rho_setup 11 6.9 0.003 0.004 6.208 6.208 make_m2s 4110 13.4 0.043 0.043 5.385 5.890 make_images 4110 14.4 0.881 0.917 5.197 5.701 dbcsr_mm_accdrv_process 11614 15.7 3.268 3.672 5.372 5.614 qs_ot_get_p 110 10.4 0.001 0.001 5.496 5.523 pw_transfer 1331 11.6 0.075 0.076 5.337 5.341 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.823 5.327 cp_fm_cholesky_invert 11 10.9 5.315 5.319 5.315 5.319 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.220 5.224 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.723 5.184 apply_single 110 13.6 0.000 0.000 4.723 5.183 ot_diis_step 99 11.5 0.015 0.015 4.593 4.593 fft_wrap_pw1pw2_140 451 13.1 0.221 0.223 4.550 4.556 fft3d_ps 1111 14.6 1.294 1.299 4.414 4.422 qs_ot_p2m_diag 48 11.0 0.151 0.155 3.982 3.989 multiply_cannon_sync_h2d 8220 15.4 3.950 3.959 3.950 3.959 density_rs2pw 110 9.6 0.004 0.004 3.848 3.881 hybrid_alltoall_any 4261 16.3 0.258 0.559 3.037 3.806 grid_integrate_task_list 110 12.3 3.669 3.734 3.669 3.734 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.723 3.724 make_images_data 4110 15.4 0.041 0.044 3.028 3.674 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.537 3.537 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.879 3.331 wfi_extrapolate 11 7.9 0.001 0.001 3.325 3.325 calculate_dm_sparse 110 9.5 0.001 0.001 3.296 3.311 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.207 3.213 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.933 2.934 cp_fm_diag_elpa_base 48 14.0 2.407 2.597 2.931 2.932 calculate_first_density_matrix 1 7.0 0.000 0.000 2.759 2.760 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.700 2.717 grid_collocate_task_list 110 9.6 2.639 2.663 2.639 2.663 potential_pw2rs 110 12.3 0.021 0.021 2.515 2.520 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.391 2.406 rs_pw_transfer 902 11.9 0.011 0.011 2.296 2.326 qs_env_update_s_mstruct 11 6.9 0.006 0.021 2.242 2.301 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.085 2.179 jit_kernel_multiply 10 15.3 1.904 2.090 1.904 2.090 mp_alltoall_d11v 2046 13.8 1.959 2.057 1.959 2.057 cp_fm_cholesky_decompose 22 10.9 1.976 1.995 1.976 1.995 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.950 1.955 qs_create_task_list 11 7.9 0.001 0.002 1.902 1.951 generate_qs_task_list 11 8.9 0.736 0.788 1.901 1.950 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.783 1.841 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.817 1.835 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=89.091000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1184.000000, yerr=51.249834 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_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 625.844224E+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 2520945. 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 0.055 0.143 205.149 205.163 qs_mol_dyn_low 1 2.0 0.004 0.006 204.562 204.576 qs_forces 11 3.9 0.003 0.003 204.459 204.462 qs_energies 11 4.9 0.001 0.002 198.846 198.859 scf_env_do_scf 11 5.9 0.001 0.001 182.195 182.201 scf_env_do_scf_inner_loop 117 6.6 0.003 0.007 161.497 161.501 dbcsr_multiply_generic 2507 12.6 0.179 0.183 123.899 124.934 velocity_verlet 10 3.0 0.001 0.001 123.552 123.555 qs_scf_new_mos 117 7.6 0.001 0.001 121.102 121.412 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.101 121.411 ot_scf_mini 117 9.6 0.003 0.003 114.532 114.780 multiply_cannon 2507 13.6 0.251 0.274 100.635 102.822 multiply_cannon_loop 2507 14.6 2.115 2.188 98.460 100.665 ot_mini 117 10.6 0.001 0.001 65.548 65.870 multiply_cannon_multrec 60168 15.6 33.077 36.009 41.481 43.642 qs_ot_get_derivative 117 11.6 0.001 0.002 40.710 40.977 rebuild_ks_matrix 128 8.3 0.001 0.001 33.938 34.243 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.018 33.937 34.242 mp_waitall_1 267128 16.5 28.114 30.916 28.114 30.916 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.492 30.776 multiply_cannon_sync_h2d 60168 15.6 27.315 30.096 27.315 30.096 qs_ot_get_p 128 10.4 0.001 0.001 27.129 27.523 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.222 25.122 apply_single 128 13.6 0.001 0.001 24.221 25.121 ot_diis_step 117 11.6 0.007 0.008 24.529 24.531 init_scf_loop 11 6.9 0.000 0.001 20.617 20.618 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.464 20.601 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.937 19.166 multiply_cannon_metrocomm3 60168 15.6 0.114 0.121 15.493 18.116 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.900 17.902 prepare_preconditioner 11 7.9 0.000 0.000 15.994 16.049 make_preconditioner 11 8.9 0.000 0.000 15.994 16.049 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.226 15.385 cp_fm_diag_elpa 83 13.4 0.000 0.000 15.092 15.097 cp_fm_redistribute_end 83 14.4 11.929 15.018 11.942 15.023 cp_fm_diag_elpa_base 83 14.4 3.035 14.650 3.068 14.766 sum_up_and_integrate 128 10.3 0.089 0.106 14.407 14.425 integrate_v_rspace 128 11.3 0.003 0.004 14.318 14.338 make_m2s 5014 13.6 0.104 0.114 13.789 14.237 make_images 5014 14.6 0.399 0.419 13.607 14.067 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.903 14.001 calculate_rho_elec 128 8.7 0.045 0.064 13.902 14.001 init_scf_run 11 5.9 0.000 0.001 12.448 12.449 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.448 12.449 density_rs2pw 128 9.7 0.006 0.007 7.262 10.426 mp_sum_l 7870 13.0 8.261 9.418 8.261 9.418 cp_fm_cholesky_invert 11 10.9 9.171 9.179 9.171 9.179 wfi_extrapolate 11 7.9 0.001 0.001 9.159 9.159 rs_pw_transfer 1046 11.9 0.016 0.019 5.802 8.952 dbcsr_mm_accdrv_process 124484 16.2 3.323 3.457 7.966 8.512 multiply_cannon_metrocomm1 60168 15.6 0.088 0.094 6.051 8.512 calculate_dm_sparse 128 9.5 0.001 0.001 8.395 8.494 pw_transfer 1547 11.6 0.076 0.096 8.103 8.339 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.897 8.139 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.943 8.086 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.785 7.933 make_images_data 5014 15.6 0.065 0.070 6.797 7.664 grid_integrate_task_list 128 12.3 7.065 7.542 7.065 7.542 hybrid_alltoall_any 5200 16.5 0.290 2.268 5.964 7.359 fft_wrap_pw1pw2_140 523 13.2 0.445 0.512 6.888 7.176 fft3d_ps 1291 14.7 2.112 2.895 6.688 6.868 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.715 6.724 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.874 5.971 mp_waitany 16020 13.9 2.708 5.951 2.708 5.951 mp_alltoall_d11v 2415 14.1 4.249 5.850 4.249 5.850 grid_collocate_task_list 128 9.7 4.557 5.793 4.557 5.793 rs_pw_transfer_RS2PW_140 139 11.5 0.280 0.302 2.121 5.306 potential_pw2rs 128 12.3 0.009 0.011 4.767 4.793 mp_sum_d 4464 12.1 3.642 4.658 3.642 4.658 cp_fm_cholesky_decompose 22 10.9 4.607 4.624 4.607 4.624 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.163000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=593.727273, yerr=5.327304 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 828.436480E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4107 56904. MP_Allreduce 11307 944. MP_Sync 170 MP_Alltoall 1983 5625022. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_comm_split 83 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.033 193.055 193.058 qs_mol_dyn_low 1 2.0 0.004 0.006 192.723 192.737 qs_forces 11 3.9 0.003 0.004 192.545 192.548 qs_energies 11 4.9 0.001 0.002 185.702 185.712 scf_env_do_scf 11 5.9 0.001 0.001 169.303 169.313 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 136.202 136.204 velocity_verlet 10 3.0 0.001 0.001 121.968 121.972 dbcsr_multiply_generic 2527 12.6 0.187 0.193 97.734 99.032 qs_scf_new_mos 118 7.6 0.001 0.001 95.278 95.776 qs_scf_loop_do_ot 118 8.6 0.001 0.001 95.277 95.775 ot_scf_mini 118 9.6 0.004 0.004 90.426 90.982 multiply_cannon 2527 13.6 0.483 0.539 77.533 81.952 multiply_cannon_loop 2527 14.6 1.278 1.330 74.548 76.872 ot_mini 118 10.6 0.001 0.001 50.185 50.658 mp_waitall_1 216438 16.6 24.648 37.513 24.648 37.513 multiply_cannon_multrec 30324 15.6 22.192 26.088 31.870 36.085 rebuild_ks_matrix 129 8.3 0.001 0.001 33.766 34.290 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.023 33.765 34.289 init_scf_loop 11 6.9 0.000 0.000 33.014 33.015 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.402 30.881 prepare_preconditioner 11 7.9 0.000 0.000 28.584 28.628 make_preconditioner 11 8.9 0.000 0.000 28.584 28.628 qs_ot_get_derivative 118 11.6 0.001 0.002 28.086 28.626 multiply_cannon_metrocomm3 30324 15.6 0.096 0.103 15.742 28.357 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.259 27.777 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.107 23.337 apply_single 129 13.6 0.001 0.001 22.107 23.337 qs_ot_get_p 129 10.4 0.001 0.001 21.958 22.518 ot_diis_step 118 11.6 0.014 0.015 21.926 21.928 multiply_cannon_sync_h2d 30324 15.6 19.388 21.420 19.388 21.420 qs_ot_p2m_diag 83 11.4 0.188 0.216 17.056 17.091 cp_fm_cholesky_invert 11 10.9 16.685 16.698 16.685 16.698 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.901 15.902 make_m2s 5054 13.6 0.090 0.096 14.453 15.717 make_images 5054 14.6 1.164 1.353 14.241 15.506 sum_up_and_integrate 129 10.3 0.115 0.133 15.017 15.047 integrate_v_rspace 129 11.3 0.003 0.004 14.902 14.938 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.531 14.565 calculate_rho_elec 129 8.7 0.089 0.106 14.530 14.564 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.667 12.678 cp_fm_redistribute_end 83 14.4 7.398 12.605 7.407 12.607 cp_fm_diag_elpa_base 83 14.4 4.961 12.090 5.179 12.459 init_scf_run 11 5.9 0.000 0.001 11.616 11.618 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.616 11.618 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.989 11.425 multiply_cannon_metrocomm4 27797 15.6 0.094 0.107 3.784 10.388 make_images_data 5054 15.6 0.064 0.073 8.803 10.301 density_rs2pw 129 9.7 0.006 0.007 7.741 10.131 mp_irecv_dv 70031 16.3 3.593 10.008 3.593 10.008 hybrid_alltoall_any 5240 16.5 0.343 1.529 7.674 9.720 dbcsr_mm_accdrv_process 62734 16.2 4.583 5.332 9.135 9.703 pw_transfer 1559 11.6 0.088 0.103 9.235 9.315 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 9.007 9.089 wfi_extrapolate 11 7.9 0.001 0.001 8.448 8.448 rs_pw_transfer 1054 12.0 0.014 0.016 5.838 8.237 fft_wrap_pw1pw2_140 527 13.2 0.475 0.528 7.936 8.038 grid_integrate_task_list 129 12.3 7.235 7.625 7.235 7.625 fft3d_ps 1301 14.7 2.774 2.948 7.359 7.394 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.388 7.183 cp_fm_cholesky_decompose 22 10.9 6.964 7.045 6.964 7.045 calculate_dm_sparse 129 9.5 0.001 0.001 6.517 6.672 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.189 6.199 grid_collocate_task_list 129 9.7 4.742 5.880 4.742 5.880 mp_alltoall_d11v 2423 14.1 4.643 5.634 4.643 5.634 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.383 5.452 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.281 5.434 mp_sum_l 7930 13.1 3.955 5.353 3.955 5.353 potential_pw2rs 129 12.3 0.015 0.018 5.188 5.212 mp_waitany 11836 13.9 2.507 4.993 2.507 4.993 mp_allgather_i34 2527 14.6 1.545 4.657 1.545 4.657 rs_pw_transfer_RS2PW_140 140 11.5 0.355 0.383 2.101 4.491 dbcsr_complete_redistribute 395 12.7 0.770 0.852 3.197 4.043 mp_sum_d 4493 12.2 2.702 4.027 2.702 4.027 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=193.058000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=789.818182, yerr=3.128118 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 969.183232E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+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 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172905552 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57194. MP_Allreduce 11251 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_comm_split 82 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.051 187.085 187.087 qs_mol_dyn_low 1 2.0 0.009 0.012 186.606 186.619 qs_forces 11 3.9 0.003 0.003 186.481 186.489 qs_energies 11 4.9 0.001 0.002 179.702 179.715 scf_env_do_scf 11 5.9 0.001 0.001 163.511 163.512 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 126.037 126.041 velocity_verlet 10 3.0 0.001 0.001 119.307 119.310 qs_scf_new_mos 116 7.6 0.001 0.001 85.150 85.470 qs_scf_loop_do_ot 116 8.6 0.001 0.001 85.149 85.469 dbcsr_multiply_generic 2485 12.5 0.185 0.188 84.306 85.354 ot_scf_mini 116 9.6 0.003 0.004 80.768 81.119 multiply_cannon 2485 13.5 0.498 0.528 62.818 67.297 multiply_cannon_loop 2485 14.5 0.852 0.881 59.334 61.754 ot_mini 116 10.6 0.001 0.001 44.226 44.576 init_scf_loop 11 6.9 0.000 0.000 37.327 37.330 mp_waitall_1 169034 16.6 27.068 36.422 27.068 36.422 prepare_preconditioner 11 7.9 0.000 0.000 32.998 33.042 make_preconditioner 11 8.9 0.000 0.000 32.998 33.042 rebuild_ks_matrix 127 8.3 0.001 0.001 32.468 32.875 qs_ks_build_kohn_sham_matrix 127 9.3 0.015 0.017 32.467 32.875 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.463 31.957 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.266 29.637 multiply_cannon_metrocomm3 19880 15.5 0.059 0.063 15.505 25.236 multiply_cannon_multrec 19880 15.5 13.281 16.371 21.938 24.921 qs_ot_get_derivative 116 11.6 0.001 0.002 24.429 24.791 qs_ot_get_p 127 10.4 0.001 0.001 21.576 21.991 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.835 20.803 apply_single 127 13.6 0.001 0.001 19.835 20.803 ot_diis_step 116 11.6 0.017 0.019 19.695 19.697 make_m2s 4970 13.5 0.080 0.085 16.006 17.269 make_images 4970 14.5 1.172 1.269 15.775 17.037 qs_ot_p2m_diag 82 11.4 0.262 0.269 16.898 16.906 sum_up_and_integrate 127 10.3 0.130 0.141 15.909 15.936 multiply_cannon_sync_h2d 19880 15.5 14.163 15.889 14.163 15.889 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.850 15.851 integrate_v_rspace 127 11.3 0.004 0.004 15.778 15.801 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.677 15.734 calculate_rho_elec 127 8.7 0.131 0.146 15.676 15.733 cp_fm_cholesky_invert 11 10.9 15.548 15.559 15.548 15.559 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.546 12.549 cp_fm_redistribute_end 82 14.4 4.762 12.471 4.784 12.473 cp_fm_diag_elpa_base 82 14.4 7.240 11.813 7.665 12.288 make_images_data 4970 15.5 0.060 0.067 9.823 11.684 init_scf_run 11 5.9 0.000 0.001 10.965 10.965 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.965 10.965 hybrid_alltoall_any 5155 16.4 0.427 1.964 8.567 10.652 pw_transfer 1535 11.6 0.086 0.105 10.418 10.536 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 10.195 10.319 density_rs2pw 127 9.7 0.006 0.006 8.091 10.034 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.750 9.995 multiply_cannon_metrocomm4 17395 15.5 0.060 0.069 3.404 9.045 fft_wrap_pw1pw2_140 519 13.2 0.475 0.519 8.788 8.914 mp_irecv_dv 49801 16.2 3.287 8.809 3.287 8.809 fft3d_ps 1281 14.7 2.672 2.900 8.466 8.561 dbcsr_mm_accdrv_process 41158 16.2 4.453 5.267 8.113 8.308 cp_fm_cholesky_decompose 22 10.9 7.934 8.001 7.934 8.001 wfi_extrapolate 11 7.9 0.001 0.001 7.851 7.851 rs_pw_transfer 1038 11.9 0.013 0.014 5.818 7.778 grid_integrate_task_list 127 12.3 7.216 7.752 7.216 7.752 cp_fm_upper_to_full 104 14.5 5.856 7.652 5.856 7.652 dbcsr_complete_redistribute 393 12.7 1.164 1.195 5.103 6.923 mp_alltoall_d11v 2401 14.1 5.602 6.542 5.602 6.542 calculate_dm_sparse 127 9.5 0.001 0.001 6.050 6.149 potential_pw2rs 127 12.3 0.020 0.022 5.733 5.748 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.682 5.689 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.967 5.673 grid_collocate_task_list 127 9.7 4.881 5.658 4.881 5.658 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.689 5.509 mp_sum_l 7804 13.0 3.676 5.010 3.676 5.010 mp_allgather_i34 2485 14.5 2.016 4.982 2.016 4.982 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.590 4.742 mp_waitany 11660 13.9 2.486 4.495 2.486 4.495 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.515 4.286 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.119 4.162 mp_alltoall_i22 712 14.1 2.127 4.094 2.127 4.094 mp_alltoall_z22v 1281 16.7 3.812 4.037 3.812 4.037 rs_pw_transfer_RS2PW_140 138 11.5 0.335 0.359 1.996 3.939 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.915 3.920 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.727 3.765 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=187.087000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=905.909091, yerr=24.062247 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_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 1.123066E+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 4085 57194. MP_Allreduce 11251 1067. 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 0.058 0.149 195.920 195.970 qs_mol_dyn_low 1 2.0 0.006 0.055 195.167 195.189 qs_forces 11 3.9 0.004 0.018 195.031 195.041 qs_energies 11 4.9 0.007 0.025 187.647 187.659 scf_env_do_scf 11 5.9 0.003 0.035 170.133 170.141 velocity_verlet 10 3.0 0.001 0.001 129.704 129.708 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 121.617 121.619 qs_scf_new_mos 117 7.6 0.001 0.001 82.104 82.416 qs_scf_loop_do_ot 117 8.6 0.001 0.001 82.104 82.415 dbcsr_multiply_generic 2507 12.6 0.190 0.193 81.242 82.005 ot_scf_mini 117 9.6 0.003 0.004 77.536 77.850 multiply_cannon 2507 13.6 0.554 0.599 55.057 58.811 multiply_cannon_loop 2507 14.6 1.181 1.208 51.289 53.137 init_scf_loop 11 6.9 0.001 0.005 48.350 48.352 prepare_preconditioner 11 7.9 0.000 0.001 44.078 44.113 make_preconditioner 11 8.9 0.000 0.001 44.078 44.113 ot_mini 117 10.6 0.001 0.001 43.438 43.747 make_full_inverse_cholesky 11 9.9 0.000 0.000 37.541 42.665 rebuild_ks_matrix 128 8.3 0.001 0.001 31.367 31.696 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 31.366 31.696 multiply_cannon_multrec 30084 15.6 14.111 19.662 26.004 30.985 mp_waitall_1 147882 16.7 18.755 28.670 18.755 28.670 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.225 28.526 qs_ot_get_derivative 117 11.6 0.001 0.002 23.634 23.949 make_m2s 5014 13.6 0.095 0.099 21.587 22.443 make_images 5014 14.6 1.966 2.237 21.280 22.138 qs_ot_get_p 128 10.4 0.001 0.001 20.079 20.428 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.242 19.786 apply_single 128 13.6 0.001 0.001 19.242 19.786 ot_diis_step 117 11.6 0.017 0.018 19.673 19.675 cp_fm_cholesky_invert 11 10.9 17.126 17.139 17.126 17.139 cp_fm_upper_to_full 105 14.7 11.298 16.655 11.298 16.655 qs_ot_p2m_diag 83 11.4 0.343 0.390 15.756 15.811 sum_up_and_integrate 128 10.3 0.140 0.153 15.668 15.696 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.531 15.562 calculate_rho_elec 128 8.7 0.175 0.191 15.531 15.561 integrate_v_rspace 128 11.3 0.004 0.005 15.528 15.560 multiply_cannon_metrocomm3 30084 15.6 0.047 0.049 6.496 15.163 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.402 14.403 make_images_data 5014 15.6 0.063 0.068 11.825 13.306 dbcsr_complete_redistribute 395 12.7 1.521 1.643 9.286 13.135 multiply_cannon_sync_h2d 30084 15.6 11.619 12.660 11.619 12.660 hybrid_alltoall_any 5200 16.5 0.524 2.198 10.355 12.143 dbcsr_mm_accdrv_process 62264 16.2 7.301 8.174 11.467 12.024 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.834 11.682 init_scf_run 11 5.9 0.000 0.001 11.306 11.308 scf_env_initial_rho_setup 11 6.9 0.002 0.004 11.305 11.307 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.178 11.181 cp_fm_redistribute_end 83 14.4 1.921 11.052 1.941 11.060 cp_fm_diag_elpa_base 83 14.4 8.488 10.417 9.072 10.907 pw_transfer 1547 11.6 0.088 0.105 10.377 10.463 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.515 10.296 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 10.149 10.240 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.894 10.134 mp_alltoall_i22 716 14.1 5.721 9.614 5.721 9.614 density_rs2pw 128 9.7 0.006 0.006 7.739 9.161 fft_wrap_pw1pw2_140 523 13.2 0.483 0.490 8.890 8.976 fft3d_ps 1291 14.7 2.783 2.877 8.371 8.435 cp_fm_cholesky_decompose 22 10.9 8.028 8.138 8.028 8.138 wfi_extrapolate 11 7.9 0.001 0.001 8.072 8.073 grid_integrate_task_list 128 12.3 7.450 7.969 7.450 7.969 multiply_cannon_metrocomm4 25070 15.6 0.075 0.085 2.798 7.862 mp_irecv_dv 76098 16.2 2.657 7.587 2.657 7.587 mp_alltoall_d11v 2415 14.1 5.662 6.730 5.662 6.730 rs_pw_transfer 1046 11.9 0.013 0.014 5.155 6.634 calculate_dm_sparse 128 9.5 0.001 0.001 6.270 6.351 grid_collocate_task_list 128 9.7 5.075 5.797 5.075 5.797 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.620 5.655 potential_pw2rs 128 12.3 0.023 0.023 5.121 5.135 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.661 4.665 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.559 4.644 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.463 4.542 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.262 4.324 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=195.970000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1063.727273, yerr=17.560756 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.524773E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3981 58509. MP_Allreduce 10950 1177. MP_Sync 86 MP_Alltoall 1700 18828162. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.065 179.544 179.555 qs_mol_dyn_low 1 2.0 0.003 0.004 179.104 179.118 qs_forces 11 3.9 0.003 0.003 178.950 178.956 qs_energies 11 4.9 0.002 0.004 171.267 171.276 scf_env_do_scf 11 5.9 0.001 0.001 153.591 153.597 velocity_verlet 10 3.0 0.001 0.001 117.130 117.137 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.970 115.972 qs_scf_new_mos 116 7.6 0.001 0.001 75.901 76.006 qs_scf_loop_do_ot 116 8.6 0.001 0.001 75.900 76.005 dbcsr_multiply_generic 2485 12.5 0.182 0.187 74.935 75.306 ot_scf_mini 116 9.6 0.003 0.004 71.394 71.455 multiply_cannon 2485 13.5 0.582 0.618 54.479 59.120 multiply_cannon_loop 2485 14.5 0.442 0.451 49.852 50.546 ot_mini 116 10.6 0.001 0.001 40.070 40.135 init_scf_loop 11 6.9 0.000 0.000 37.458 37.462 mp_waitall_1 124680 16.7 26.700 33.850 26.700 33.850 prepare_preconditioner 11 7.9 0.000 0.000 33.404 33.427 make_preconditioner 11 8.9 0.000 0.000 33.404 33.427 make_full_inverse_cholesky 11 9.9 0.000 0.000 31.086 31.414 rebuild_ks_matrix 127 8.3 0.001 0.001 30.810 30.882 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 30.810 30.881 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.981 28.051 multiply_cannon_multrec 9940 15.5 10.298 14.215 17.556 20.370 qs_ot_get_derivative 116 11.6 0.001 0.002 20.296 20.357 ot_diis_step 116 11.6 0.019 0.020 19.702 19.704 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.273 19.548 apply_single 127 13.6 0.001 0.001 19.273 19.547 cp_fm_cholesky_invert 11 10.9 18.995 19.002 18.995 19.002 make_m2s 4970 13.5 0.065 0.070 16.734 18.775 multiply_cannon_metrocomm3 9940 15.5 0.023 0.025 12.073 18.765 make_images 4970 14.5 2.266 2.616 16.431 18.466 qs_ot_get_p 127 10.4 0.001 0.001 18.368 18.442 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.156 16.197 calculate_rho_elec 127 8.7 0.257 0.266 16.155 16.196 sum_up_and_integrate 127 10.3 0.178 0.188 15.711 15.759 integrate_v_rspace 127 11.3 0.004 0.004 15.532 15.584 qs_ot_p2m_diag 82 11.4 0.489 0.495 14.574 14.591 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.310 13.311 make_images_data 4970 15.5 0.051 0.058 10.226 12.801 hybrid_alltoall_any 5155 16.4 0.808 3.798 9.893 12.223 multiply_cannon_sync_h2d 9940 15.5 11.538 11.795 11.538 11.795 init_scf_run 11 5.9 0.000 0.001 10.756 10.757 scf_env_initial_rho_setup 11 6.9 0.002 0.003 10.756 10.756 pw_transfer 1535 11.6 0.085 0.095 10.565 10.612 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 10.344 10.394 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.248 10.261 cp_fm_diag_elpa_base 82 14.4 9.998 10.083 10.241 10.253 fft_wrap_pw1pw2_140 519 13.2 0.491 0.509 9.009 9.054 cp_fm_cholesky_decompose 22 10.9 8.483 8.586 8.483 8.586 fft3d_ps 1281 14.7 2.686 2.765 8.524 8.556 density_rs2pw 127 9.7 0.005 0.006 7.582 8.460 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.191 8.240 grid_integrate_task_list 127 12.3 7.680 8.129 7.680 8.129 multiply_cannon_metrocomm1 9940 15.5 0.029 0.029 5.026 7.698 wfi_extrapolate 11 7.9 0.001 0.001 7.671 7.671 mp_allgather_i34 2485 14.5 2.839 7.542 2.839 7.542 dbcsr_mm_accdrv_process 20590 16.1 2.622 3.444 6.890 7.535 mp_alltoall_d11v 2401 14.1 5.848 6.787 5.848 6.787 calculate_dm_sparse 127 9.5 0.001 0.001 6.231 6.325 grid_collocate_task_list 127 9.7 5.348 6.021 5.348 6.021 dbcsr_complete_redistribute 393 12.7 2.111 2.186 5.592 6.003 qs_energies_init_hamiltonians 11 5.9 0.003 0.005 5.369 5.369 rs_pw_transfer 1038 11.9 0.012 0.013 4.451 5.368 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.266 5.276 potential_pw2rs 127 12.3 0.026 0.026 4.948 4.962 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.095 4.122 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.668 4.024 multiply_cannon_metrocomm4 7455 15.5 0.023 0.025 1.853 3.949 mp_irecv_dv 28618 15.9 1.819 3.887 1.819 3.887 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.567 3.885 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.695 3.774 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.726 3.753 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.739 3.746 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.605 3.624 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=179.555000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1427.545455, yerr=50.209314 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 11.785519E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 1979552 0.0% 0.0% 100.0% average stack size 0.0 0.0 3438.3 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 3.136025E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101080 MPI messages size (bytes): total size 1.144901E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.326685E+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 45568 35366371328 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726591807936 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 58746. MP_Allreduce 11072 1512. MP_Sync 87 MP_Alltoall 1724 36993717. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.066 305.507 305.515 qs_mol_dyn_low 1 2.0 0.005 0.009 304.873 304.887 qs_forces 11 3.9 0.006 0.009 304.398 304.401 qs_energies 11 4.9 0.005 0.006 295.012 295.026 scf_env_do_scf 11 5.9 0.001 0.001 272.180 272.197 velocity_verlet 10 3.0 0.001 0.001 220.062 220.115 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 145.719 145.721 init_scf_loop 11 6.9 0.000 0.000 126.201 126.203 prepare_preconditioner 11 7.9 0.000 0.000 120.946 120.971 make_preconditioner 11 8.9 0.000 0.000 120.946 120.971 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.928 118.090 qs_scf_new_mos 118 7.6 0.001 0.001 92.660 92.775 qs_scf_loop_do_ot 118 8.6 0.001 0.001 92.659 92.774 ot_scf_mini 118 9.6 0.004 0.004 87.829 87.873 dbcsr_multiply_generic 2527 12.6 0.215 0.222 84.562 85.119 cp_fm_upper_to_full 105 14.8 52.750 75.831 52.750 75.831 multiply_cannon 2527 13.6 0.718 0.762 59.076 59.648 multiply_cannon_loop 2527 14.6 0.476 0.486 55.427 56.761 ot_mini 118 10.6 0.001 0.001 45.140 45.180 dbcsr_complete_redistribute 395 12.7 3.971 4.022 29.938 42.861 rebuild_ks_matrix 129 8.3 0.001 0.001 39.770 39.821 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 39.770 39.821 copy_fm_to_dbcsr 209 11.7 0.002 0.002 26.399 39.301 qs_ks_update_qs_env 129 7.6 0.001 0.001 36.709 36.760 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.973 36.721 mp_alltoall_i22 718 14.1 21.831 34.617 21.831 34.617 cp_fm_cholesky_invert 11 10.9 34.508 34.516 34.508 34.516 mp_waitall_1 104500 16.8 28.932 32.862 28.932 32.862 qs_ot_get_p 129 10.4 0.001 0.001 26.916 26.931 qs_ot_get_derivative 118 11.6 0.002 0.002 25.020 25.062 qs_ot_p2m_diag 83 11.4 0.878 0.882 22.594 22.620 qs_rho_update_rho_low 129 7.7 0.001 0.001 21.741 21.751 calculate_rho_elec 129 8.7 0.486 0.488 21.741 21.750 make_m2s 5054 13.6 0.077 0.080 20.737 21.666 make_images 5054 14.6 3.768 3.978 20.254 21.185 cp_dbcsr_syevd 83 12.4 0.006 0.006 20.832 20.834 sum_up_and_integrate 129 10.3 0.323 0.324 20.600 20.680 integrate_v_rspace 129 11.3 0.004 0.004 20.276 20.356 ot_diis_step 118 11.6 0.022 0.023 20.090 20.092 multiply_cannon_metrocomm3 10108 15.6 0.024 0.024 18.862 20.003 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.292 19.503 apply_single 129 13.6 0.001 0.001 19.292 19.503 multiply_cannon_multrec 10108 15.6 10.569 12.341 18.065 18.134 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.539 17.540 cp_fm_diag_elpa_base 83 14.4 13.086 14.758 17.534 17.535 multiply_cannon_sync_h2d 10108 15.6 15.776 15.798 15.776 15.798 pw_transfer 1559 11.6 0.094 0.095 13.372 13.385 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 13.135 13.148 hybrid_alltoall_any 5240 16.5 1.309 3.061 11.147 13.052 make_images_data 5054 15.6 0.060 0.065 11.139 12.915 init_scf_run 11 5.9 0.000 0.001 12.629 12.630 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.629 12.630 fft_wrap_pw1pw2_140 527 13.2 0.546 0.547 11.563 11.574 fft3d_ps 1301 14.7 2.774 2.784 11.157 11.166 mp_alltoall_d11v 2423 14.1 9.034 9.880 9.034 9.880 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.754 9.795 cp_fm_cholesky_decompose 22 10.9 9.521 9.598 9.521 9.598 wfi_extrapolate 11 7.9 0.001 0.001 9.379 9.379 dbcsr_mm_accdrv_process 20926 16.1 3.899 5.661 7.256 9.145 density_rs2pw 129 9.7 0.005 0.005 8.992 9.124 grid_integrate_task_list 129 12.3 8.633 8.801 8.633 8.801 qs_energies_init_hamiltonians 11 5.9 0.003 0.006 8.130 8.135 calculate_dm_sparse 129 9.5 0.001 0.001 6.724 6.818 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.492 6.583 rs_scatter_matrices 140 9.7 3.579 4.508 6.293 6.526 grid_collocate_task_list 129 9.7 6.408 6.445 6.408 6.445 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.198 6.290 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=305.515000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2769.727273, yerr=183.937287 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_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 1.262117E+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 0.033 0.070 87.903 87.905 qs_energies 1 2.0 0.000 0.000 87.162 87.183 ls_scf 1 3.0 0.003 0.025 86.167 86.174 dbcsr_multiply_generic 111 6.7 0.014 0.015 74.266 74.458 multiply_cannon 111 7.7 0.018 0.021 56.215 57.550 multiply_cannon_loop 111 8.7 0.208 0.221 52.645 54.088 ls_scf_main 1 4.0 0.002 0.019 53.977 53.980 density_matrix_trs4 2 5.0 0.002 0.003 48.144 48.234 ls_scf_init_scf 1 4.0 0.000 0.001 29.069 29.080 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.828 27.876 mp_waitall_1 11031 10.9 24.270 26.651 24.270 26.651 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.635 25.652 multiply_cannon_multrec 2664 9.7 8.126 8.804 15.427 17.049 multiply_cannon_sync_h2d 2664 9.7 13.227 14.903 13.227 14.903 make_m2s 222 7.7 0.008 0.010 14.308 14.896 make_images 222 8.7 0.099 0.110 14.286 14.876 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 10.269 13.618 make_images_data 222 9.7 0.004 0.005 8.746 9.346 hybrid_alltoall_any 227 10.6 0.215 1.839 7.537 8.996 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.557 8.759 dbcsr_mm_accdrv_process 4760 10.4 0.511 0.618 6.922 7.875 dbcsr_mm_accdrv_process_sort 4760 11.4 6.208 7.073 6.208 7.073 calculate_norms 4752 9.8 5.472 6.050 5.472 6.050 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.197 5.313 mp_sum_l 807 5.4 3.241 4.601 3.241 4.601 make_images_sizes 222 9.7 0.000 0.000 0.947 3.936 mp_alltoall_i44 222 10.7 0.947 3.936 0.947 3.936 arnoldi_extremal 4 6.8 0.000 0.000 3.615 3.636 arnoldi_normal_ev 4 7.8 0.001 0.003 3.615 3.636 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.059 3.585 mp_irecv_dv 6231 10.9 2.043 3.570 2.043 3.570 build_subspace 16 8.4 0.009 0.012 3.525 3.528 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.315 3.385 ls_scf_post 1 4.0 0.002 0.035 3.118 3.127 ls_scf_store_result 1 5.0 0.000 0.000 2.936 2.986 dbcsr_matrix_vector_mult 304 9.0 0.005 0.012 2.471 2.714 dbcsr_special_finalize 555 9.7 0.005 0.006 2.293 2.653 dbcsr_merge_single_wm 555 10.7 0.460 0.593 2.285 2.646 make_images_pack 222 9.7 2.176 2.603 2.178 2.604 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.457 2.514 dbcsr_matrix_vector_mult_local 304 10.0 2.055 2.445 2.057 2.447 dbcsr_sort_data 658 11.4 2.082 2.374 2.082 2.374 buffer_matrices_ensure_size 222 8.7 1.760 2.133 1.760 2.133 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.922 1.938 rebuild_ks_matrix 3 7.3 0.000 0.000 1.912 1.929 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.021 1.912 1.929 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=87.905000, 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/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_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 2.119475E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.040 99.756 99.757 qs_energies 1 2.0 0.000 0.000 99.124 99.129 ls_scf 1 3.0 0.000 0.000 97.775 97.780 dbcsr_multiply_generic 111 6.7 0.015 0.015 82.168 82.771 multiply_cannon 111 7.7 0.027 0.034 56.076 62.076 ls_scf_main 1 4.0 0.000 0.000 61.067 61.074 multiply_cannon_loop 111 8.7 0.115 0.123 52.631 57.632 density_matrix_trs4 2 5.0 0.002 0.003 54.759 55.046 mp_waitall_1 9105 10.9 26.357 36.428 26.357 36.428 ls_scf_init_scf 1 4.0 0.000 0.001 32.992 32.994 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.610 31.743 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.056 29.077 multiply_cannon_multrec 1332 9.7 12.798 15.891 21.974 25.841 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 14.236 24.906 make_m2s 222 7.7 0.006 0.008 17.598 18.442 make_images 222 8.7 1.578 1.981 17.568 18.414 make_images_data 222 9.7 0.004 0.004 10.817 11.964 mp_sum_l 807 5.4 7.223 11.904 7.223 11.904 hybrid_alltoall_any 227 10.6 0.517 2.479 9.981 11.021 dbcsr_mm_accdrv_process 4041 10.4 0.241 0.455 8.785 10.387 dbcsr_mm_accdrv_process_sort 4041 11.4 8.371 9.932 8.371 9.932 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 5.687 9.640 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.220 8.014 mp_irecv_dv 3311 11.0 3.201 7.968 3.201 7.968 calculate_norms 2376 9.8 5.949 6.575 5.949 6.575 multiply_cannon_sync_h2d 1332 9.7 4.706 6.165 4.706 6.165 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.493 5.825 arnoldi_extremal 4 6.8 0.000 0.000 5.527 5.551 arnoldi_normal_ev 4 7.8 0.001 0.005 5.527 5.551 build_subspace 16 8.4 0.014 0.020 5.254 5.260 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.914 3.790 ls_scf_post 1 4.0 0.000 0.000 3.715 3.719 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.375 3.620 ls_scf_store_result 1 5.0 0.000 0.000 3.424 3.540 dbcsr_matrix_vector_mult_local 304 10.0 2.725 3.200 2.727 3.202 mp_allgather_i34 111 8.7 1.148 3.139 1.148 3.139 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.867 2.936 make_images_sizes 222 9.7 0.000 0.000 0.857 2.556 mp_alltoall_i44 222 10.7 0.856 2.556 0.856 2.556 make_images_pack 222 9.7 2.021 2.392 2.023 2.394 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.087 2.089 rebuild_ks_matrix 3 7.3 0.000 0.000 2.074 2.076 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.074 2.076 dbcsr_sort_data 436 11.2 1.804 2.054 1.804 2.054 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=99.757000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1746.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_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 2.702930E+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 0.068 0.104 101.992 102.001 qs_energies 1 2.0 0.000 0.000 101.045 101.059 ls_scf 1 3.0 0.000 0.001 99.546 99.558 dbcsr_multiply_generic 111 6.7 0.015 0.016 82.651 83.024 ls_scf_main 1 4.0 0.000 0.000 62.607 62.613 multiply_cannon 111 7.7 0.043 0.127 55.932 62.211 density_matrix_trs4 2 5.0 0.002 0.003 56.060 56.242 multiply_cannon_loop 111 8.7 0.099 0.107 52.222 55.538 mp_waitall_1 7281 11.0 28.875 37.121 28.875 37.121 ls_scf_init_scf 1 4.0 0.000 0.002 33.122 33.127 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.754 31.849 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.190 29.209 multiply_cannon_multrec 888 9.7 12.481 15.044 20.977 23.953 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 13.033 23.745 make_m2s 222 7.7 0.006 0.006 19.495 20.810 make_images 222 8.7 1.982 2.273 19.457 20.773 make_images_data 222 9.7 0.003 0.004 11.567 12.913 hybrid_alltoall_any 227 10.6 0.615 2.834 11.067 12.837 mp_sum_l 807 5.4 5.758 10.078 5.758 10.078 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 4.849 9.299 dbcsr_mm_accdrv_process 3754 10.4 0.253 0.426 8.037 9.288 dbcsr_mm_accdrv_process_sort 3754 11.4 7.663 8.862 7.663 8.862 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.347 7.703 multiply_cannon_sync_h2d 888 9.7 5.976 7.316 5.976 7.316 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.445 6.834 mp_irecv_dv 2335 11.1 2.430 6.773 2.430 6.773 arnoldi_extremal 4 6.8 0.000 0.000 5.989 6.013 arnoldi_normal_ev 4 7.8 0.001 0.005 5.989 6.013 build_subspace 16 8.4 0.014 0.020 5.678 5.684 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.325 5.570 mp_allgather_i34 111 8.7 1.539 4.696 1.539 4.696 calculate_norms 1584 9.8 4.285 4.647 4.285 4.647 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.722 4.049 ls_scf_post 1 4.0 0.004 0.018 3.816 3.833 ls_scf_store_result 1 5.0 0.000 0.000 3.554 3.639 dbcsr_matrix_vector_mult_local 304 10.0 3.007 3.591 3.009 3.593 make_images_sizes 222 9.7 0.000 0.000 1.623 3.341 mp_alltoall_i44 222 10.7 1.623 3.340 1.623 3.340 ls_scf_dm_to_ks 2 5.0 0.000 0.001 3.133 3.205 make_images_pack 222 9.7 1.818 2.133 1.820 2.136 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.116 2.118 dbcsr_sort_data 325 11.1 1.863 2.109 1.863 2.109 rebuild_ks_matrix 3 7.3 0.000 0.000 2.090 2.093 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.017 2.090 2.093 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=102.001000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2166.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_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 3.333677E+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 0.074 0.094 97.278 97.278 qs_energies 1 2.0 0.000 0.000 96.519 96.527 ls_scf 1 3.0 0.000 0.000 94.859 94.867 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.583 78.844 ls_scf_main 1 4.0 0.000 0.000 58.859 58.868 multiply_cannon 111 7.7 0.061 0.120 51.746 55.336 density_matrix_trs4 2 5.0 0.002 0.003 52.721 52.818 multiply_cannon_loop 111 8.7 0.115 0.125 46.715 49.952 ls_scf_init_scf 1 4.0 0.000 0.000 32.723 32.739 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.491 31.565 mp_waitall_1 6369 11.0 22.925 30.878 22.925 30.878 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.009 29.024 multiply_cannon_multrec 1332 9.7 14.119 17.079 21.963 24.461 make_m2s 222 7.7 0.006 0.008 21.098 22.462 make_images 222 8.7 3.142 3.617 21.048 22.414 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.438 17.387 make_images_data 222 9.7 0.004 0.004 11.768 13.380 hybrid_alltoall_any 227 10.6 0.797 3.812 11.214 12.970 dbcsr_mm_accdrv_process 3641 10.4 0.221 0.418 7.488 9.011 dbcsr_mm_accdrv_process_sort 3641 11.4 7.109 8.591 7.109 8.591 mp_sum_l 807 5.4 4.170 7.301 4.170 7.301 multiply_cannon_sync_h2d 1332 9.7 5.485 6.311 5.485 6.311 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.088 6.097 mp_irecv_dv 3229 10.9 2.065 6.051 2.065 6.051 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.178 5.604 arnoldi_extremal 4 6.8 0.000 0.000 5.226 5.242 arnoldi_normal_ev 4 7.8 0.001 0.005 5.226 5.242 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.552 4.911 build_subspace 16 8.4 0.014 0.021 4.842 4.852 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.533 4.663 calculate_norms 2376 9.8 4.228 4.647 4.228 4.647 mp_allgather_i34 111 8.7 2.154 4.435 2.154 4.435 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.560 3.865 dbcsr_matrix_vector_mult_local 304 10.0 3.176 3.690 3.178 3.692 dbcsr_sort_data 658 11.4 3.119 3.564 3.119 3.564 ls_scf_post 1 4.0 0.000 0.000 3.277 3.285 dbcsr_special_finalize 555 9.7 0.006 0.007 2.863 3.243 dbcsr_merge_single_wm 555 10.7 0.539 0.666 2.855 3.235 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.051 3.101 ls_scf_store_result 1 5.0 0.000 0.000 3.023 3.082 dbcsr_data_release 10477 10.7 1.575 2.443 1.575 2.443 dbcsr_finalize 304 7.8 0.049 0.061 1.804 2.064 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.988 2.002 rebuild_ks_matrix 3 7.3 0.000 0.000 1.963 1.978 qs_ks_build_kohn_sham_matrix 3 8.3 0.009 0.023 1.963 1.978 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.278000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2747.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_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 4.638335E+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 0.077 0.092 99.160 99.160 qs_energies 1 2.0 0.000 0.000 98.254 98.260 ls_scf 1 3.0 0.000 0.000 96.301 96.313 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.563 77.748 ls_scf_main 1 4.0 0.000 0.000 61.902 61.903 multiply_cannon 111 7.7 0.092 0.162 55.158 60.206 density_matrix_trs4 2 5.0 0.002 0.003 54.635 54.712 multiply_cannon_loop 111 8.7 0.069 0.077 50.604 51.934 mp_waitall_1 5436 11.0 26.223 32.937 26.223 32.937 ls_scf_init_scf 1 4.0 0.000 0.000 30.862 30.868 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.501 29.538 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.342 27.353 multiply_cannon_multrec 444 9.7 14.002 16.330 21.032 23.641 make_m2s 222 7.7 0.004 0.005 17.680 20.204 make_images 222 8.7 3.726 4.431 17.618 20.145 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.950 16.007 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.162 15.655 make_images_data 222 9.7 0.003 0.004 9.917 12.430 hybrid_alltoall_any 227 10.6 0.792 3.777 9.733 12.236 multiply_cannon_sync_h2d 444 9.7 6.578 8.448 6.578 8.448 dbcsr_mm_accdrv_process 3003 10.4 0.164 0.338 6.733 7.879 dbcsr_mm_accdrv_process_sort 3003 11.4 6.418 7.533 6.418 7.533 mp_allgather_i34 111 8.7 2.780 6.963 2.780 6.963 arnoldi_extremal 4 6.8 0.000 0.000 5.805 5.818 arnoldi_normal_ev 4 7.8 0.002 0.006 5.805 5.818 build_subspace 16 8.4 0.015 0.020 5.411 5.418 mp_sum_l 807 5.4 2.770 5.058 2.770 5.058 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.516 4.674 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.146 4.325 dbcsr_matrix_vector_mult_local 304 10.0 3.649 4.119 3.651 4.121 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.662 3.945 mp_irecv_dv 1241 11.2 1.644 3.922 1.644 3.922 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.957 3.890 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.669 3.779 calculate_norms 792 9.8 3.543 3.676 3.543 3.676 make_images_sizes 222 9.7 0.000 0.000 1.119 3.564 mp_alltoall_i44 222 10.7 1.119 3.564 1.119 3.564 ls_scf_post 1 4.0 0.000 0.000 3.536 3.543 ls_scf_store_result 1 5.0 0.000 0.000 3.321 3.362 dbcsr_finalize 304 7.8 0.062 0.078 2.200 2.257 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.255 2.256 rebuild_ks_matrix 3 7.3 0.000 0.000 2.222 2.224 qs_ks_build_kohn_sham_matrix 3 8.3 0.011 0.011 2.222 2.224 dbcsr_merge_all 275 8.9 0.473 0.523 2.048 2.088 dbcsr_data_release 10123 10.8 1.326 1.991 1.326 1.991 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.160000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3680.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9_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 8.766591E+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 0.095 0.106 114.398 114.399 qs_energies 1 2.0 0.000 0.000 112.906 112.920 ls_scf 1 3.0 0.000 0.000 109.959 109.972 dbcsr_multiply_generic 111 6.7 0.023 0.025 82.334 82.454 ls_scf_main 1 4.0 0.000 0.000 70.386 70.388 density_matrix_trs4 2 5.0 0.002 0.003 60.302 60.365 multiply_cannon 111 7.7 0.120 0.196 53.008 55.401 multiply_cannon_loop 111 8.7 0.067 0.069 49.218 50.274 ls_scf_init_scf 1 4.0 0.000 0.000 35.675 35.676 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.894 33.908 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 31.021 31.031 mp_waitall_1 4527 11.1 26.329 30.653 26.329 30.653 make_m2s 222 7.7 0.005 0.005 25.549 26.851 make_images 222 8.7 4.591 4.969 25.443 26.743 multiply_cannon_multrec 444 9.7 17.852 18.584 22.442 23.052 hybrid_alltoall_any 227 10.6 1.663 3.649 14.082 17.634 make_images_data 222 9.7 0.003 0.003 14.414 17.502 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 13.272 13.905 multiply_cannon_sync_h2d 444 9.7 8.843 8.888 8.843 8.888 arnoldi_extremal 4 6.8 0.000 0.000 7.751 7.762 arnoldi_normal_ev 4 7.8 0.003 0.009 7.751 7.762 build_subspace 16 8.4 0.026 0.036 7.157 7.168 dbcsr_matrix_vector_mult 304 9.0 0.016 0.032 5.629 5.774 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.646 5.736 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.439 5.685 dbcsr_matrix_vector_mult_local 304 10.0 4.924 5.215 4.926 5.218 dbcsr_mm_accdrv_process 1814 10.4 0.254 0.319 4.428 4.549 make_images_sizes 222 9.7 0.000 0.000 1.806 4.318 mp_alltoall_i44 222 10.7 1.806 4.318 1.806 4.318 mp_allgather_i34 111 8.7 1.432 4.310 1.432 4.310 dbcsr_mm_accdrv_process_sort 1814 11.4 4.126 4.253 4.126 4.253 ls_scf_post 1 4.0 0.000 0.000 3.897 3.910 ls_scf_store_result 1 5.0 0.000 0.000 3.597 3.614 calculate_norms 792 9.8 3.252 3.368 3.252 3.368 dbcsr_finalize 304 7.8 0.082 0.089 3.085 3.139 dbcsr_complete_redistribute 5 7.6 1.433 1.465 2.905 3.032 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.917 2.917 dbcsr_merge_all 275 8.9 0.894 0.925 2.870 2.916 dbcsr_data_release 12724 10.6 2.323 2.872 2.323 2.872 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.797 2.799 rebuild_ks_matrix 3 7.3 0.000 0.000 2.731 2.733 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.731 2.733 matrix_ls_to_qs 2 6.0 0.000 0.000 2.469 2.625 dbcsr_sort_data 325 11.1 2.441 2.507 2.441 2.507 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=114.399000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6911.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 3916455a2c0a6b120ac01e0cd9af67fc9c7ed8e9 Summary: empty Status: OK