=== 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: b8a11310c43cf35ba551cf3896354d85cdb3a872 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 21.06.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/01 job id: 47749670 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/02 job id: 47749671 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/03 job id: 47749673 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/04 job id: 47749674 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/05 job id: 47749677 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/06 job id: 47749679 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/07 job id: 47749681 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/08 job id: 47749683 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/09 job id: 47749684 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/10 job id: 47749685 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/11 job id: 47749686 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/12 job id: 47749687 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/13 job id: 47749688 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/14 job id: 47749689 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/15 job id: 47749690 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/16 job id: 47749691 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/17 job id: 47749692 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/18 job id: 47749693 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/19 job id: 47749694 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/20 job id: 47749695 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/21 job id: 47749696 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/22 job id: 47749697 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/23 job id: 47749698 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/24 job id: 47749699 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/25 job id: 47749700 --- 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/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/26 job id: 47749701 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/27 job id: 47749702 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.035 133.957 133.958 farming_run 1 2.0 133.393 133.395 133.928 133.931 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.506869E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.026 114.558 114.558 qs_energies 1 2.0 0.000 0.000 114.367 114.370 mp2_main 1 3.0 0.000 0.000 112.295 112.298 mp2_gpw_main 1 4.0 0.027 0.032 111.307 111.310 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 92.277 92.681 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.936 55.340 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.165 41.318 46.461 get_2c_integrals 1 6.0 0.008 0.009 36.441 37.169 integrate_v_rspace 273 8.0 0.441 0.454 24.748 29.720 pw_transfer 6555 10.6 0.382 0.392 27.252 27.644 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 25.898 26.321 grid_integrate_task_list 273 9.0 20.570 26.023 20.570 26.023 fft_wrap_pw1pw2_100 2178 12.4 1.139 1.199 23.430 23.839 compute_2c_integrals 1 7.0 0.002 0.002 19.177 19.179 rpa_ri_compute_en 1 5.0 0.019 0.021 18.912 19.098 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.774 18.904 mp2_eri_2c_integrate_gpw 1 9.0 2.397 2.439 18.771 18.901 cp_fm_cholesky_decompose 12 8.2 17.186 17.927 17.186 17.927 cholesky_decomp 1 7.0 0.000 0.000 16.104 16.856 fft3d_s 5443 13.4 16.092 16.544 16.114 16.566 ao_to_mo_and_store_B_mult_1 272 7.0 10.776 15.333 10.776 15.333 calculate_wavefunction 272 8.0 5.398 5.554 12.484 13.127 rpa_num_int 1 6.0 0.002 0.026 10.649 10.650 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.601 10.627 calc_mat_Q 8 8.0 0.000 0.000 9.471 9.593 contract_S_to_Q 8 9.0 0.000 0.000 8.891 9.014 calc_potential_gpw 544 9.5 0.005 0.006 8.229 8.592 parallel_gemm_fm 14 9.1 0.000 0.000 8.480 8.590 parallel_gemm_fm_cosma 14 10.1 8.480 8.590 8.480 8.590 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.149 8.382 potential_pw2rs 545 10.0 0.107 0.109 7.640 8.291 create_integ_mat 1 6.0 0.022 0.027 7.765 7.766 collocate_single_gaussian 272 10.0 0.039 0.042 7.414 7.662 array2fm 1 7.0 0.000 0.000 6.763 7.225 pw_scatter_s 2720 13.7 4.415 4.560 4.415 4.560 pw_gather_s 2722 13.2 3.851 4.107 3.851 4.107 array2fm_buffer_send 1 8.0 2.916 3.153 2.916 3.153 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.309655, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2823.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.040 397.007 397.008 farming_run 1 2.0 395.843 395.847 396.970 396.973 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.241575E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 28169588. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.030 209.757 209.758 qs_energies 1 2.0 0.000 0.000 209.165 209.173 scf_env_do_scf 1 3.0 0.000 0.000 106.784 106.784 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.867 105.874 rebuild_ks_matrix 4 6.0 0.000 0.000 105.844 105.851 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 105.844 105.851 hfx_ks_matrix 4 8.0 0.001 0.001 105.466 105.470 integrate_four_center 4 9.0 0.143 0.473 105.465 105.470 mp2_main 1 3.0 0.000 0.000 102.091 102.098 mp2_gpw_main 1 4.0 0.048 0.072 101.255 101.265 integrate_four_center_main 4 10.0 0.072 0.472 97.225 99.702 integrate_four_center_bin 267 11.0 97.153 99.694 97.153 99.694 init_scf_loop 1 4.0 0.000 0.000 92.542 92.543 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.381 75.441 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.008 55.070 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.164 41.848 46.768 integrate_v_rspace 95 8.0 0.400 0.563 28.223 32.963 pw_transfer 2240 10.6 0.145 0.166 29.901 30.344 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.912 29.379 ao_to_mo_and_store_B_mult_1 91 7.0 10.485 29.095 10.485 29.095 mp2_ri_gpw_compute_en 1 5.0 0.054 0.062 26.713 28.873 grid_integrate_task_list 95 9.0 23.528 28.469 23.528 28.469 fft_wrap_pw1pw2_100 730 12.4 1.280 1.462 26.624 27.088 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.832 1.887 25.001 25.011 get_2c_integrals 1 6.0 0.000 0.000 20.281 20.309 compute_2c_integrals 1 7.0 0.002 0.003 19.268 19.271 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.851 19.147 mp2_eri_2c_integrate_gpw 1 9.0 1.739 1.870 18.850 19.145 fft3d_s 1823 13.4 18.421 18.816 18.435 18.829 scf_env_do_scf_inner_loop 4 4.0 0.001 0.023 14.240 14.240 calculate_wavefunction 91 8.0 2.015 2.045 9.735 9.952 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.582 8.760 9.484 potential_pw2rs 186 10.0 0.033 0.036 8.590 9.102 local_gemm 172 8.0 8.202 8.920 8.202 8.920 mp2_ri_gpw_compute_en_comm 22 7.0 0.505 0.526 8.007 8.581 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.238 8.564 calc_potential_gpw 182 9.5 0.002 0.002 7.940 8.219 collocate_single_gaussian 91 10.0 0.017 0.019 7.875 8.102 mp_sendrecv_dm3 2068 8.0 6.027 6.570 6.027 6.570 mp2_ri_gpw_compute_en_ener 172 7.0 6.356 6.421 6.356 6.421 mp_sync 37 10.5 2.934 5.879 2.934 5.879 pw_gather_s 912 13.2 4.871 5.369 4.871 5.369 pw_scatter_s 910 13.7 3.951 4.360 3.951 4.360 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.252850, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1499.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.674112E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 588782. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.028 56.898 56.899 qs_mol_dyn_low 1 2.0 0.003 0.003 56.661 56.669 qs_forces 11 3.9 0.002 0.002 56.600 56.600 qs_energies 11 4.9 0.001 0.001 55.091 55.108 scf_env_do_scf 11 5.9 0.000 0.001 48.457 48.457 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 46.457 46.457 qs_scf_new_mos 108 7.5 0.000 0.001 36.207 36.511 qs_scf_loop_do_ot 108 8.5 0.000 0.001 36.207 36.510 dbcsr_multiply_generic 2286 12.5 0.094 0.097 35.891 36.281 ot_scf_mini 108 9.5 0.002 0.002 34.492 34.679 multiply_cannon 2286 13.5 0.188 0.205 27.640 29.244 multiply_cannon_loop 2286 14.5 1.853 1.969 26.955 28.567 velocity_verlet 10 3.0 0.003 0.007 27.847 27.848 ot_mini 108 10.5 0.001 0.001 21.031 21.258 qs_ot_get_derivative 108 11.5 0.001 0.001 17.713 17.885 mp_waitall_1 245248 16.5 9.440 15.861 9.440 15.861 multiply_cannon_metrocomm3 54864 15.5 0.073 0.078 6.478 13.675 multiply_cannon_multrec 54864 15.5 3.633 5.678 7.887 11.343 qs_ot_get_p 119 10.4 0.001 0.001 8.173 8.503 rebuild_ks_matrix 119 8.3 0.000 0.000 7.886 8.040 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.886 8.040 mp_sum_l 7287 12.8 6.019 7.586 6.019 7.586 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.943 7.080 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.371 6.854 multiply_cannon_sync_h2d 54864 15.5 5.141 6.565 5.141 6.565 dbcsr_mm_accdrv_process 76910 16.1 1.845 2.906 4.167 5.977 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.818 5.929 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.370 5.436 init_scf_run 11 5.9 0.000 0.001 5.358 5.359 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.358 5.358 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.526 4.526 sum_up_and_integrate 119 10.3 0.012 0.014 4.439 4.446 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.303 4.444 calculate_rho_elec 119 8.7 0.012 0.017 4.302 4.443 integrate_v_rspace 119 11.3 0.002 0.003 4.427 4.435 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.341 4.341 cp_fm_redistribute_end 50 14.0 2.215 4.319 2.222 4.322 cp_fm_diag_elpa_base 50 14.0 2.095 4.213 2.099 4.222 multiply_cannon_metrocomm1 54864 15.5 0.057 0.062 2.236 3.802 calculate_dm_sparse 119 9.5 0.000 0.001 3.216 3.375 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.023 3.303 apply_single 119 13.6 0.000 0.000 3.022 3.302 acc_transpose_blocks 54864 15.5 0.230 0.254 2.425 3.236 calculate_first_density_matrix 1 7.0 0.000 0.000 3.143 3.146 jit_kernel_multiply 13 15.8 2.259 3.014 2.259 3.014 ot_diis_step 108 11.5 0.006 0.006 3.012 3.013 rs_pw_transfer 974 11.9 0.012 0.013 2.702 2.793 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.583 2.660 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.648 2.649 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.369 2.420 density_rs2pw 119 9.7 0.004 0.005 2.189 2.281 mp_sum_d 4135 12.0 1.570 2.199 1.570 2.199 wfi_extrapolate 11 7.9 0.001 0.001 2.147 2.147 grid_integrate_task_list 119 12.3 2.026 2.112 2.026 2.112 init_scf_loop 11 6.9 0.000 0.000 1.984 1.984 make_m2s 4572 13.5 0.053 0.055 1.914 1.974 make_images 4572 14.5 0.133 0.138 1.832 1.892 potential_pw2rs 119 12.3 0.004 0.004 1.794 1.803 pw_transfer 1439 11.6 0.052 0.057 1.702 1.767 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.669 1.695 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.627 1.694 mp_alltoall_d11v 2130 13.8 1.440 1.606 1.440 1.606 acc_transpose_blocks_kernels 54864 16.5 0.249 0.386 0.937 1.495 mp_waitany 12084 13.8 1.296 1.463 1.296 1.463 acc_transpose_blocks_sync 164592 16.5 1.212 1.451 1.212 1.451 grid_collocate_task_list 119 9.7 1.355 1.430 1.355 1.430 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.965 1.399 fft_wrap_pw1pw2_140 487 13.2 0.188 0.209 1.265 1.334 fft3d_ps 1201 14.6 0.369 0.475 1.266 1.331 dbcsr_new_transposed 954 13.3 0.007 0.008 1.030 1.234 dbcsr_redistribute 145 14.7 0.004 0.004 1.005 1.213 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=56.899000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.272727, yerr=1.052349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_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 487.403520E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1971066. 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 0.011 0.029 39.004 39.005 qs_mol_dyn_low 1 2.0 0.003 0.004 38.792 38.821 qs_forces 11 3.9 0.002 0.003 38.679 38.680 qs_energies 11 4.9 0.001 0.001 36.975 36.980 scf_env_do_scf 11 5.9 0.000 0.001 31.970 31.970 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.583 29.583 dbcsr_multiply_generic 2286 12.5 0.100 0.103 21.946 22.296 qs_scf_new_mos 108 7.5 0.001 0.001 20.511 20.741 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.510 20.741 ot_scf_mini 108 9.5 0.002 0.003 19.592 19.756 velocity_verlet 10 3.0 0.029 0.060 18.517 18.518 multiply_cannon 2286 13.5 0.208 0.217 16.871 18.321 multiply_cannon_loop 2286 14.5 1.204 1.255 15.722 17.215 ot_mini 108 10.5 0.001 0.001 12.086 12.316 mp_waitall_1 200699 16.5 5.445 10.697 5.445 10.697 qs_ot_get_derivative 108 11.5 0.001 0.002 9.724 9.891 multiply_cannon_metrocomm3 27432 15.5 0.071 0.075 3.995 9.511 multiply_cannon_multrec 27432 15.5 1.837 4.216 6.150 8.937 rebuild_ks_matrix 119 8.3 0.000 0.000 7.184 7.324 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.183 7.324 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.327 6.456 dbcsr_mm_accdrv_process 47894 16.0 3.438 5.636 4.232 6.301 qs_ot_get_p 119 10.4 0.001 0.001 4.599 4.823 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.635 4.563 mp_sum_l 7287 12.8 2.058 4.101 2.058 4.101 sum_up_and_integrate 119 10.3 0.025 0.028 4.095 4.100 integrate_v_rspace 119 11.3 0.002 0.003 4.070 4.079 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.080 4.059 apply_single 119 13.6 0.000 0.000 3.080 4.058 init_scf_run 11 5.9 0.000 0.001 3.819 3.819 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.819 3.819 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.735 3.760 calculate_rho_elec 119 8.7 0.021 0.024 3.734 3.759 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.987 3.005 rs_pw_transfer 974 11.9 0.010 0.012 2.427 2.929 make_m2s 4572 13.5 0.052 0.053 2.566 2.915 make_images 4572 14.5 0.206 0.246 2.479 2.829 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.589 2.590 density_rs2pw 119 9.7 0.004 0.004 2.001 2.524 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.298 2.384 init_scf_loop 11 6.9 0.000 0.000 2.368 2.368 calculate_first_density_matrix 1 7.0 0.000 0.000 2.314 2.316 ot_diis_step 108 11.5 0.010 0.011 2.315 2.316 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.279 2.279 cp_fm_redistribute_end 50 14.0 1.155 2.254 1.158 2.256 multiply_cannon_sync_h2d 27432 15.5 1.704 2.224 1.704 2.224 cp_fm_diag_elpa_base 50 14.0 1.066 2.162 1.095 2.206 calculate_dm_sparse 119 9.5 0.000 0.001 2.059 2.141 acc_transpose_blocks 27432 15.5 0.115 0.121 1.632 1.975 grid_integrate_task_list 119 12.3 1.838 1.943 1.838 1.943 jit_kernel_multiply 10 16.2 0.736 1.919 0.736 1.919 pw_transfer 1439 11.6 0.065 0.069 1.873 1.906 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.896 1.898 potential_pw2rs 119 12.3 0.006 0.006 1.816 1.825 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.781 1.815 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.702 1.743 make_images_data 4572 15.5 0.048 0.055 1.228 1.560 prepare_preconditioner 11 7.9 0.000 0.000 1.459 1.485 make_preconditioner 11 8.9 0.000 0.000 1.459 1.485 fft_wrap_pw1pw2_140 487 13.2 0.204 0.216 1.428 1.461 wfi_extrapolate 11 7.9 0.001 0.001 1.446 1.446 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.086 1.432 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.363 1.417 grid_collocate_task_list 119 9.7 1.279 1.417 1.279 1.417 fft3d_ps 1201 14.6 0.517 0.573 1.331 1.355 mp_alltoall_d11v 2130 13.8 1.154 1.328 1.154 1.328 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.247 1.256 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.192 1.239 rs_pw_transfer_RS2PW_140 130 11.5 0.140 0.148 0.534 1.040 mp_waitany 5720 13.7 0.504 1.027 0.504 1.027 mp_sum_d 4135 12.0 0.526 1.017 0.526 1.017 mp_allgather_i34 2286 14.5 0.571 0.962 0.571 0.962 acc_transpose_blocks_sync 82296 16.5 0.827 0.949 0.827 0.949 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.929 0.930 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.912 0.926 acc_transpose_blocks_kernels 27432 16.5 0.188 0.275 0.665 0.880 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.005000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.090909, yerr=1.831767 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_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 520.204288E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607820. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.035 32.730 32.737 qs_mol_dyn_low 1 2.0 0.003 0.004 32.388 32.396 qs_forces 11 3.9 0.002 0.002 32.313 32.314 qs_energies 11 4.9 0.016 0.085 30.611 30.676 scf_env_do_scf 11 5.9 0.001 0.002 25.809 25.816 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 23.345 23.347 dbcsr_multiply_generic 2286 12.5 0.093 0.096 16.805 16.868 velocity_verlet 10 3.0 0.001 0.002 15.291 15.293 qs_scf_new_mos 108 7.5 0.001 0.001 15.252 15.273 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.252 15.273 ot_scf_mini 108 9.5 0.002 0.003 14.501 14.517 multiply_cannon 2286 13.5 0.194 0.199 13.623 14.384 multiply_cannon_loop 2286 14.5 0.856 0.892 12.865 13.636 ot_mini 108 10.5 0.001 0.001 8.969 8.987 qs_ot_get_derivative 108 11.5 0.001 0.001 7.484 7.500 multiply_cannon_multrec 18288 15.5 1.891 2.965 6.960 7.318 rebuild_ks_matrix 119 8.3 0.000 0.000 6.359 6.379 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.359 6.379 dbcsr_mm_accdrv_process 38222 16.0 4.880 5.681 4.975 5.746 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.573 5.592 sum_up_and_integrate 119 10.3 0.032 0.032 3.934 3.943 integrate_v_rspace 119 11.3 0.003 0.003 3.902 3.916 mp_waitall_1 158411 16.6 2.527 3.738 2.527 3.738 init_scf_run 11 5.9 0.000 0.001 3.569 3.569 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.568 3.568 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.510 3.525 calculate_rho_elec 119 8.7 0.031 0.032 3.509 3.525 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.813 3.415 qs_ot_get_p 119 10.4 0.001 0.001 3.324 3.350 rs_pw_transfer 974 11.9 0.009 0.010 2.246 2.520 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.414 2.479 init_scf_loop 11 6.9 0.002 0.009 2.447 2.455 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.007 2.349 apply_single 119 13.6 0.000 0.000 2.007 2.349 density_rs2pw 119 9.7 0.004 0.004 1.987 2.282 calculate_first_density_matrix 1 7.0 0.001 0.002 2.262 2.264 qs_ot_p2m_diag 50 11.0 0.013 0.017 2.237 2.243 pw_transfer 1439 11.6 0.065 0.070 1.932 1.947 cp_dbcsr_syevd 50 12.0 0.009 0.028 1.941 1.942 make_m2s 4572 13.5 0.044 0.045 1.762 1.918 grid_integrate_task_list 119 12.3 1.802 1.909 1.802 1.909 calculate_dm_sparse 119 9.5 0.000 0.001 1.870 1.877 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.839 1.856 make_images 4572 14.5 0.193 0.205 1.678 1.831 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.694 1.708 potential_pw2rs 119 12.3 0.007 0.008 1.695 1.706 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.690 1.698 acc_transpose_blocks 18288 15.5 0.078 0.080 1.636 1.685 cp_fm_diag_elpa_base 50 14.0 1.652 1.666 1.674 1.683 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.675 1.677 prepare_preconditioner 11 7.9 0.000 0.000 1.621 1.625 make_preconditioner 11 8.9 0.001 0.006 1.621 1.625 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.478 1.562 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.531 1.538 fft_wrap_pw1pw2_140 487 13.2 0.255 0.264 1.490 1.507 ot_diis_step 108 11.5 0.011 0.011 1.472 1.473 mp_sum_l 7287 12.8 1.056 1.448 1.056 1.448 grid_collocate_task_list 119 9.7 1.259 1.400 1.259 1.400 fft3d_ps 1201 14.6 0.529 0.548 1.326 1.341 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.215 1.218 multiply_cannon_sync_h2d 18288 15.5 1.061 1.172 1.061 1.172 wfi_extrapolate 11 7.9 0.001 0.001 1.160 1.160 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.983 0.984 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.931 0.947 make_images_data 4572 15.5 0.048 0.051 0.753 0.936 hybrid_alltoall_any 4725 16.4 0.058 0.114 0.649 0.862 mp_waitany 9880 13.7 0.574 0.862 0.574 0.862 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.857 0.859 rs_pw_transfer_RS2PW_140 130 11.5 0.121 0.127 0.571 0.847 acc_transpose_blocks_kernels 18288 16.5 0.216 0.225 0.797 0.809 acc_transpose_blocks_sync 54864 16.5 0.741 0.796 0.741 0.796 jit_kernel_multiply 4 15.7 0.036 0.791 0.036 0.791 build_core_hamiltonian_matrix_ 11 4.9 0.011 0.083 0.690 0.756 mp_alltoall_d11v 2130 13.8 0.627 0.746 0.627 0.746 mp_alltoall_z22v 1201 16.6 0.639 0.714 0.639 0.714 cp_fm_cholesky_invert 11 10.9 0.680 0.684 0.680 0.684 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.737000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.727273, yerr=1.542778 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 556.851200E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.041 37.462 37.463 qs_mol_dyn_low 1 2.0 0.003 0.004 37.253 37.263 qs_forces 11 3.9 0.002 0.003 36.970 36.973 qs_energies 11 4.9 0.002 0.006 35.234 35.242 scf_env_do_scf 11 5.9 0.001 0.002 29.939 29.940 scf_env_do_scf_inner_loop 108 6.5 0.003 0.010 26.500 26.501 dbcsr_multiply_generic 2286 12.5 0.097 0.102 20.079 20.203 velocity_verlet 10 3.0 0.001 0.002 18.894 18.905 qs_scf_new_mos 108 7.5 0.001 0.001 17.991 18.048 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.990 18.048 ot_scf_mini 108 9.5 0.002 0.003 16.969 17.024 multiply_cannon 2286 13.5 0.217 0.224 16.171 16.845 multiply_cannon_loop 2286 14.5 1.514 1.608 15.216 15.745 ot_mini 108 10.5 0.001 0.001 10.308 10.370 multiply_cannon_multrec 27432 15.5 2.463 3.118 8.999 9.252 qs_ot_get_derivative 108 11.5 0.001 0.001 8.453 8.510 dbcsr_mm_accdrv_process 47916 15.9 5.888 7.608 6.431 7.754 rebuild_ks_matrix 119 8.3 0.000 0.000 6.602 6.652 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.601 6.651 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.849 5.892 init_scf_run 11 5.9 0.000 0.001 3.871 3.872 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.871 3.871 sum_up_and_integrate 119 10.3 0.036 0.038 3.759 3.766 qs_ot_get_p 119 10.4 0.001 0.002 3.660 3.733 integrate_v_rspace 119 11.3 0.003 0.003 3.724 3.731 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.597 3.639 calculate_rho_elec 119 8.7 0.040 0.046 3.597 3.638 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.179 3.499 init_scf_loop 11 6.9 0.001 0.004 3.421 3.425 acc_transpose_blocks 27432 15.5 0.118 0.121 2.392 2.605 prepare_preconditioner 11 7.9 0.000 0.000 2.558 2.567 make_preconditioner 11 8.9 0.000 0.001 2.558 2.567 make_m2s 4572 13.5 0.053 0.055 2.406 2.526 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.173 2.494 make_images 4572 14.5 0.272 0.334 2.298 2.417 calculate_first_density_matrix 1 7.0 0.000 0.002 2.414 2.416 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 2.377 2.406 mp_waitall_1 137007 16.6 1.854 2.277 1.854 2.277 rs_pw_transfer 974 11.9 0.009 0.010 2.030 2.230 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.203 2.216 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.024 2.208 apply_single 119 13.6 0.000 0.000 2.024 2.208 calculate_dm_sparse 119 9.5 0.000 0.000 2.133 2.189 density_rs2pw 119 9.7 0.004 0.005 1.957 2.148 pw_transfer 1439 11.6 0.065 0.069 1.996 2.028 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.983 1.984 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.903 1.939 grid_integrate_task_list 119 12.3 1.836 1.920 1.836 1.920 ot_diis_step 108 11.5 0.012 0.012 1.814 1.815 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.780 1.802 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.801 1.801 jit_kernel_multiply 9 15.9 0.478 1.640 0.478 1.640 acc_transpose_blocks_sync 82296 16.5 1.423 1.627 1.423 1.627 fft_wrap_pw1pw2_140 487 13.2 0.290 0.306 1.580 1.620 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.525 1.536 cp_fm_diag_elpa_base 50 14.0 1.491 1.510 1.523 1.533 potential_pw2rs 119 12.3 0.009 0.009 1.502 1.505 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.445 1.464 mp_sum_l 7287 12.8 1.084 1.432 1.084 1.432 wfi_extrapolate 11 7.9 0.001 0.001 1.397 1.397 fft3d_ps 1201 14.6 0.557 0.606 1.340 1.367 grid_collocate_task_list 119 9.7 1.249 1.367 1.249 1.367 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.860 1.325 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.239 1.249 dbcsr_complete_redistribute 329 12.2 0.122 0.160 0.849 1.182 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.162 1.164 cp_fm_upper_to_full 72 14.2 0.801 1.141 0.801 1.141 make_images_data 4572 15.5 0.048 0.051 0.896 1.053 hybrid_alltoall_any 4725 16.4 0.065 0.154 0.789 1.025 mp_alltoall_d11v 2130 13.8 0.787 0.921 0.787 0.921 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.555 0.879 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.798 0.875 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.842 0.847 acc_transpose_blocks_kernels 27432 16.5 0.270 0.278 0.822 0.832 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.463000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.909091, yerr=3.088234 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_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 614.649856E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.026 28.906 28.906 qs_mol_dyn_low 1 2.0 0.003 0.004 28.730 28.738 qs_forces 11 3.9 0.003 0.003 28.673 28.674 qs_energies 11 4.9 0.001 0.001 26.946 26.949 scf_env_do_scf 11 5.9 0.000 0.001 22.223 22.223 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.805 19.805 velocity_verlet 10 3.0 0.001 0.002 14.781 14.784 dbcsr_multiply_generic 2286 12.5 0.091 0.100 13.156 13.242 qs_scf_new_mos 108 7.5 0.001 0.001 11.828 11.851 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.828 11.851 ot_scf_mini 108 9.5 0.002 0.002 11.127 11.152 multiply_cannon 2286 13.5 0.223 0.229 10.586 11.051 multiply_cannon_loop 2286 14.5 0.645 0.671 9.670 9.873 ot_mini 108 10.5 0.001 0.001 6.498 6.527 multiply_cannon_multrec 9144 15.5 1.684 1.921 5.930 6.084 rebuild_ks_matrix 119 8.3 0.000 0.000 6.010 6.032 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.010 6.031 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.352 5.370 qs_ot_get_derivative 108 11.5 0.001 0.001 5.222 5.245 dbcsr_mm_accdrv_process 12550 15.8 3.565 4.148 4.135 4.197 sum_up_and_integrate 119 10.3 0.037 0.041 3.682 3.686 integrate_v_rspace 119 11.3 0.003 0.003 3.644 3.649 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.490 3.493 calculate_rho_elec 119 8.7 0.060 0.061 3.489 3.493 init_scf_run 11 5.9 0.000 0.001 3.275 3.275 scf_env_initial_rho_setup 11 6.9 0.001 0.004 3.275 3.275 qs_ot_get_p 119 10.4 0.001 0.001 2.757 2.794 init_scf_loop 11 6.9 0.000 0.000 2.399 2.400 calculate_first_density_matrix 1 7.0 0.000 0.000 2.144 2.145 mp_waitall_1 115863 16.7 1.541 2.058 1.541 2.058 pw_transfer 1439 11.6 0.066 0.068 2.024 2.036 grid_integrate_task_list 119 12.3 1.848 1.975 1.848 1.975 make_m2s 4572 13.5 0.034 0.035 1.789 1.955 density_rs2pw 119 9.7 0.004 0.004 1.797 1.945 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.930 1.942 make_images 4572 14.5 0.270 0.301 1.702 1.866 rs_pw_transfer 974 11.9 0.008 0.008 1.686 1.829 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.782 1.784 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.728 1.749 calculate_dm_sparse 119 9.5 0.000 0.000 1.725 1.741 prepare_preconditioner 11 7.9 0.000 0.000 1.661 1.666 make_preconditioner 11 8.9 0.000 0.000 1.661 1.666 fft_wrap_pw1pw2_140 487 13.2 0.366 0.375 1.585 1.595 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.553 1.579 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.576 1.577 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.542 1.543 acc_transpose_blocks 9144 15.5 0.042 0.043 1.456 1.503 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.413 1.423 potential_pw2rs 119 12.3 0.010 0.010 1.416 1.419 grid_collocate_task_list 119 9.7 1.294 1.411 1.294 1.411 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.332 1.345 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.276 1.284 cp_fm_diag_elpa_base 50 14.0 1.249 1.267 1.274 1.283 fft3d_ps 1201 14.6 0.558 0.570 1.262 1.274 ot_diis_step 108 11.5 0.012 0.013 1.264 1.264 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.209 1.241 apply_single 119 13.6 0.000 0.000 1.209 1.241 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.236 1.238 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.122 1.126 wfi_extrapolate 11 7.9 0.001 0.001 1.079 1.079 jit_kernel_multiply 6 15.8 0.531 1.020 0.531 1.020 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.752 0.985 make_images_data 4572 15.5 0.042 0.046 0.771 0.954 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.872 0.930 mp_alltoall_d11v 2130 13.8 0.746 0.851 0.746 0.851 cp_fm_cholesky_invert 11 10.9 0.820 0.822 0.820 0.822 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.809 0.816 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.494 0.794 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.781 0.783 acc_transpose_blocks_sync 27432 16.5 0.725 0.767 0.725 0.767 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.683 0.734 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 0.672 0.682 mp_allgather_i34 2286 14.5 0.234 0.654 0.234 0.654 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.288 0.624 mp_alltoall_z22v 1201 16.6 0.579 0.610 0.579 0.610 mp_waitany 5200 13.7 0.467 0.595 0.467 0.595 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.906000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=584.000000, yerr=5.116817 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_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 766.087168E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.048 42.624 42.625 qs_mol_dyn_low 1 2.0 0.003 0.004 42.382 42.389 qs_forces 11 3.9 0.002 0.002 42.309 42.310 qs_energies 11 4.9 0.001 0.001 40.302 40.305 scf_env_do_scf 11 5.9 0.001 0.001 34.694 34.695 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.930 26.931 velocity_verlet 10 3.0 0.002 0.002 24.045 24.052 dbcsr_multiply_generic 2286 12.5 0.098 0.102 18.883 19.132 qs_scf_new_mos 108 7.5 0.001 0.001 17.168 17.263 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.167 17.263 ot_scf_mini 108 9.5 0.002 0.002 16.008 16.107 multiply_cannon 2286 13.5 0.297 0.301 15.019 16.013 multiply_cannon_loop 2286 14.5 0.881 0.899 13.784 14.806 ot_mini 108 10.5 0.001 0.001 9.818 9.932 multiply_cannon_multrec 9144 15.5 3.473 4.821 8.820 8.906 qs_ot_get_derivative 108 11.5 0.001 0.001 7.785 7.883 init_scf_loop 11 6.9 0.000 0.000 7.739 7.742 rebuild_ks_matrix 119 8.3 0.000 0.000 7.313 7.451 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.313 7.450 prepare_preconditioner 11 7.9 0.000 0.000 6.758 6.771 make_preconditioner 11 8.9 0.000 0.000 6.758 6.771 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.585 6.710 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.328 6.651 dbcsr_mm_accdrv_process 12550 15.8 4.535 6.135 5.210 6.532 cp_fm_upper_to_full 72 14.2 3.208 4.585 3.208 4.585 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.259 4.265 calculate_rho_elec 119 8.7 0.118 0.121 4.258 4.265 sum_up_and_integrate 119 10.3 0.064 0.066 4.016 4.021 integrate_v_rspace 119 11.3 0.004 0.006 3.951 3.959 init_scf_run 11 5.9 0.000 0.001 3.543 3.543 scf_env_initial_rho_setup 11 6.9 0.005 0.005 3.543 3.543 qs_ot_get_p 119 10.4 0.001 0.001 3.292 3.424 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.785 3.222 mp_waitall_1 94719 16.7 2.152 3.119 2.152 3.119 dbcsr_complete_redistribute 329 12.2 0.287 0.290 1.951 2.803 pw_transfer 1439 11.6 0.069 0.070 2.610 2.614 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.511 2.514 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.215 2.500 apply_single 119 13.6 0.000 0.000 2.215 2.500 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.659 2.488 make_m2s 4572 13.5 0.037 0.037 2.286 2.427 make_images 4572 14.5 0.356 0.391 2.167 2.308 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.425 2.246 calculate_dm_sparse 119 9.5 0.000 0.000 2.223 2.240 mp_alltoall_i22 627 13.8 1.397 2.196 1.397 2.196 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.275 2.173 density_rs2pw 119 9.7 0.004 0.004 2.134 2.166 fft_wrap_pw1pw2_140 487 13.2 0.617 0.621 2.129 2.133 grid_integrate_task_list 119 12.3 2.092 2.131 2.092 2.131 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.046 2.097 calculate_first_density_matrix 1 7.0 0.000 0.000 2.091 2.093 ot_diis_step 108 11.5 0.014 0.015 2.006 2.007 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.941 1.943 acc_transpose_blocks 9144 15.5 0.044 0.045 1.844 1.873 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.810 1.810 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.780 1.781 mp_sum_l 7287 12.8 1.034 1.747 1.034 1.747 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.569 1.608 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.585 1.586 grid_collocate_task_list 119 9.7 1.531 1.562 1.531 1.562 fft3d_ps 1201 14.6 0.593 0.605 1.527 1.531 rs_pw_transfer 974 11.9 0.009 0.009 1.456 1.499 wfi_extrapolate 11 7.9 0.001 0.001 1.383 1.383 cp_fm_cholesky_invert 11 10.9 1.379 1.382 1.379 1.382 potential_pw2rs 119 12.3 0.014 0.014 1.356 1.359 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.300 1.319 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.306 1.306 cp_fm_diag_elpa_base 50 14.0 1.160 1.213 1.304 1.305 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.229 1.236 hybrid_alltoall_any 4725 16.4 0.089 0.149 1.016 1.215 acc_transpose_blocks_sync 27432 16.5 1.133 1.165 1.133 1.165 mp_alltoall_d11v 2130 13.8 1.092 1.143 1.092 1.143 qs_env_update_s_mstruct 11 6.9 0.004 0.005 1.108 1.120 make_images_data 4572 15.5 0.046 0.049 0.933 1.103 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.992 1.032 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.940 0.953 qs_create_task_list 11 7.9 0.000 0.000 0.934 0.944 generate_qs_task_list 11 8.9 0.372 0.391 0.934 0.944 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.625000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=717.181818, yerr=13.319966 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_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 501.616640E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1116689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.082 81.711 81.728 qs_mol_dyn_low 1 2.0 0.003 0.004 81.165 81.177 qs_forces 11 3.9 0.003 0.003 81.096 81.098 qs_energies 11 4.9 0.001 0.001 78.223 78.234 scf_env_do_scf 11 5.9 0.000 0.001 69.517 69.519 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.153 64.153 dbcsr_multiply_generic 2055 12.4 0.104 0.106 50.835 51.099 qs_scf_new_mos 99 7.5 0.000 0.001 46.900 47.031 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.899 47.030 ot_scf_mini 99 9.5 0.002 0.002 44.569 44.680 multiply_cannon 2055 13.4 0.182 0.186 42.566 43.561 multiply_cannon_loop 2055 14.4 1.820 1.869 41.662 42.690 velocity_verlet 10 3.0 0.001 0.002 42.629 42.629 ot_mini 99 10.5 0.001 0.001 26.460 26.558 qs_ot_get_derivative 99 11.5 0.001 0.001 19.664 19.757 multiply_cannon_multrec 49320 15.4 11.390 12.117 17.399 18.143 rebuild_ks_matrix 110 8.3 0.000 0.000 14.547 14.755 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.016 14.547 14.755 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.739 12.931 mp_waitall_1 220248 16.4 10.612 11.475 10.612 11.475 multiply_cannon_sync_h2d 49320 15.4 9.644 10.201 9.644 10.201 qs_ot_get_p 110 10.4 0.001 0.001 9.668 9.803 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.389 7.933 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.199 7.720 apply_single 110 13.6 0.000 0.000 7.199 7.720 multiply_cannon_metrocomm3 49320 15.4 0.084 0.088 6.520 7.599 sum_up_and_integrate 110 10.3 0.036 0.043 7.038 7.053 integrate_v_rspace 110 11.3 0.003 0.004 7.001 7.025 init_scf_run 11 5.9 0.000 0.001 6.691 6.691 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.691 6.691 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.549 6.673 calculate_rho_elec 110 8.6 0.021 0.026 6.548 6.672 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.575 6.604 ot_diis_step 99 11.5 0.005 0.005 6.578 6.579 dbcsr_mm_accdrv_process 87628 16.1 3.053 3.146 5.878 6.121 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.595 5.595 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.459 5.511 init_scf_loop 11 6.9 0.000 0.000 5.338 5.339 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.046 5.071 cp_fm_diag_elpa_base 48 14.0 5.035 5.061 5.045 5.070 mp_sum_l 6594 12.7 3.994 4.856 3.994 4.856 rs_pw_transfer 902 11.9 0.012 0.013 3.590 4.171 wfi_extrapolate 11 7.9 0.001 0.001 4.016 4.016 make_m2s 4110 13.4 0.060 0.064 3.823 3.921 density_rs2pw 110 9.6 0.004 0.005 3.308 3.831 make_images 4110 14.4 0.176 0.188 3.728 3.830 calculate_dm_sparse 110 9.5 0.001 0.001 3.694 3.791 multiply_cannon_metrocomm1 49320 15.4 0.066 0.068 2.359 3.601 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.524 3.528 grid_integrate_task_list 110 12.3 3.270 3.408 3.270 3.408 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.195 3.246 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.160 3.225 prepare_preconditioner 11 7.9 0.000 0.000 3.146 3.178 make_preconditioner 11 8.9 0.000 0.000 3.146 3.178 pw_transfer 1331 11.6 0.055 0.072 3.041 3.115 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 2.952 3.029 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.940 2.991 acc_transpose_blocks 49320 15.4 0.214 0.221 2.593 2.673 fft_wrap_pw1pw2_140 451 13.1 0.458 0.505 2.538 2.621 calculate_first_density_matrix 1 7.0 0.000 0.000 2.574 2.578 jit_kernel_multiply 13 15.9 2.545 2.564 2.545 2.564 potential_pw2rs 110 12.3 0.006 0.007 2.546 2.561 mp_alltoall_d11v 2046 13.8 1.983 2.461 1.983 2.461 mp_waitany 14300 13.8 1.816 2.436 1.816 2.436 grid_collocate_task_list 110 9.6 2.157 2.370 2.157 2.370 fft3d_ps 1111 14.6 0.783 0.872 2.138 2.195 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.937 1.962 mp_sum_d 3889 11.9 1.365 1.892 1.365 1.892 make_images_data 4110 15.4 0.045 0.049 1.740 1.880 hybrid_alltoall_any 4261 16.3 0.085 0.484 1.518 1.778 cp_fm_cholesky_invert 11 10.9 1.713 1.716 1.713 1.716 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.642 1.662 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.728000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.363636, yerr=2.804954 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_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 590.581760E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 1439748. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.051 68.926 68.927 qs_mol_dyn_low 1 2.0 0.003 0.004 68.605 68.615 qs_forces 11 3.9 0.003 0.003 68.540 68.541 qs_energies 11 4.9 0.001 0.001 65.174 65.177 scf_env_do_scf 11 5.9 0.000 0.001 56.744 56.747 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.251 49.251 dbcsr_multiply_generic 2055 12.4 0.113 0.117 37.707 37.952 velocity_verlet 10 3.0 0.001 0.002 36.049 36.050 qs_scf_new_mos 99 7.5 0.001 0.001 32.959 33.081 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.958 33.080 multiply_cannon 2055 13.4 0.223 0.242 31.189 32.335 ot_scf_mini 99 9.5 0.003 0.003 31.299 31.411 multiply_cannon_loop 2055 14.4 1.166 1.185 29.994 31.018 ot_mini 99 10.5 0.001 0.001 18.269 18.389 multiply_cannon_multrec 24660 15.4 7.022 8.589 14.023 15.846 rebuild_ks_matrix 110 8.3 0.000 0.000 13.768 13.891 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.768 13.890 qs_ot_get_derivative 99 11.5 0.001 0.001 12.483 12.596 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.093 12.208 mp_waitall_1 176588 16.5 7.431 9.756 7.431 9.756 multiply_cannon_metrocomm3 24660 15.4 0.072 0.074 5.088 7.932 multiply_cannon_sync_h2d 24660 15.4 6.353 7.712 6.353 7.712 init_scf_loop 11 6.9 0.000 0.000 7.458 7.459 dbcsr_mm_accdrv_process 52282 16.1 5.467 6.136 6.832 7.143 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.464 7.093 apply_single 110 13.6 0.000 0.001 6.463 7.093 sum_up_and_integrate 110 10.3 0.052 0.059 6.627 6.637 integrate_v_rspace 110 11.3 0.003 0.003 6.575 6.589 qs_ot_get_p 110 10.4 0.001 0.001 6.411 6.558 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.175 6.185 calculate_rho_elec 110 8.6 0.040 0.048 6.175 6.185 init_scf_run 11 5.9 0.000 0.001 6.070 6.071 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.070 6.071 ot_diis_step 99 11.5 0.010 0.010 5.738 5.738 prepare_preconditioner 11 7.9 0.000 0.000 5.442 5.464 make_preconditioner 11 8.9 0.000 0.000 5.441 5.464 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.696 5.393 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.027 5.180 make_m2s 4110 13.4 0.056 0.059 4.147 4.558 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.528 4.548 make_images 4110 14.4 0.410 0.470 4.041 4.448 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.067 4.067 density_rs2pw 110 9.6 0.004 0.005 3.261 3.842 pw_transfer 1331 11.6 0.067 0.078 3.597 3.731 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.490 3.626 rs_pw_transfer 902 11.9 0.012 0.015 2.948 3.552 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.509 3.522 cp_fm_diag_elpa_base 48 14.0 3.462 3.479 3.507 3.519 wfi_extrapolate 11 7.9 0.001 0.001 3.498 3.498 grid_integrate_task_list 110 12.3 3.153 3.376 3.153 3.376 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.253 3.316 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.229 3.230 fft_wrap_pw1pw2_140 451 13.1 0.522 0.536 3.003 3.138 calculate_dm_sparse 110 9.5 0.001 0.001 2.963 2.990 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.897 2.947 make_images_data 4110 15.4 0.050 0.054 2.266 2.669 hybrid_alltoall_any 4261 16.3 0.107 0.459 1.957 2.668 fft3d_ps 1111 14.6 1.107 1.330 2.460 2.607 grid_collocate_task_list 110 9.6 2.095 2.529 2.095 2.529 cp_fm_cholesky_invert 11 10.9 2.501 2.508 2.501 2.508 calculate_first_density_matrix 1 7.0 0.000 0.000 2.475 2.477 potential_pw2rs 110 12.3 0.008 0.009 2.406 2.421 mp_sum_l 6594 12.7 1.778 2.344 1.778 2.344 jit_kernel_multiply 11 16.2 1.003 2.068 1.003 2.068 acc_transpose_blocks 24660 15.4 0.116 0.119 1.998 2.026 mp_alltoall_d11v 2046 13.8 1.705 2.000 1.705 2.000 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.953 1.974 qs_energies_init_hamiltonians 11 5.9 0.000 0.004 1.856 1.857 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.832 1.846 mp_waitany 10164 13.8 1.244 1.807 1.244 1.807 rs_pw_transfer_RS2PW_140 121 11.5 0.208 0.219 0.981 1.582 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.538 1.550 multiply_cannon_metrocomm4 22605 15.4 0.079 0.084 0.780 1.513 mp_allgather_i34 2055 14.4 0.506 1.502 0.506 1.502 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.385 1.491 mp_irecv_dv 57340 16.2 0.649 1.389 0.649 1.389 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.927000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=558.363636, yerr=6.650670 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 668.160000E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.037 60.054 60.055 qs_mol_dyn_low 1 2.0 0.003 0.004 59.497 59.507 qs_forces 11 3.9 0.003 0.003 59.420 59.421 qs_energies 11 4.9 0.001 0.001 56.203 56.205 scf_env_do_scf 11 5.9 0.000 0.001 48.363 48.363 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.949 39.950 velocity_verlet 10 3.0 0.001 0.002 32.515 32.517 dbcsr_multiply_generic 2055 12.4 0.104 0.109 28.587 28.882 qs_scf_new_mos 99 7.5 0.001 0.001 24.914 25.023 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.914 25.022 ot_scf_mini 99 9.5 0.002 0.002 23.674 23.798 multiply_cannon 2055 13.4 0.213 0.224 22.489 23.679 multiply_cannon_loop 2055 14.4 0.816 0.850 21.334 22.426 ot_mini 99 10.5 0.001 0.001 13.575 13.707 rebuild_ks_matrix 110 8.3 0.000 0.000 12.395 12.554 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.395 12.554 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.895 11.034 multiply_cannon_multrec 16440 15.4 3.667 4.664 9.833 10.745 mp_waitall_1 139946 16.5 6.867 9.880 6.867 9.880 qs_ot_get_derivative 99 11.5 0.001 0.001 9.158 9.283 init_scf_loop 11 6.9 0.000 0.000 8.380 8.381 multiply_cannon_metrocomm3 16440 15.4 0.045 0.049 4.364 7.162 prepare_preconditioner 11 7.9 0.000 0.000 6.607 6.623 make_preconditioner 11 8.9 0.000 0.000 6.607 6.623 sum_up_and_integrate 110 10.3 0.060 0.061 6.532 6.547 integrate_v_rspace 110 11.3 0.003 0.003 6.472 6.487 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.926 6.291 dbcsr_mm_accdrv_process 34862 16.1 5.322 5.682 6.009 6.184 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.979 5.990 calculate_rho_elec 110 8.6 0.059 0.059 5.978 5.989 init_scf_run 11 5.9 0.000 0.001 5.452 5.452 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.452 5.452 qs_ot_get_p 110 10.4 0.001 0.001 5.267 5.423 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.857 5.357 apply_single 110 13.6 0.000 0.000 4.857 5.356 make_m2s 4110 13.4 0.049 0.050 4.116 4.495 ot_diis_step 99 11.5 0.010 0.011 4.388 4.388 make_images 4110 14.4 0.399 0.521 4.003 4.381 density_rs2pw 110 9.6 0.005 0.005 3.066 4.314 rs_pw_transfer 902 11.9 0.010 0.011 2.641 3.880 multiply_cannon_sync_h2d 16440 15.4 3.223 3.866 3.223 3.866 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.193 3.811 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.675 3.678 pw_transfer 1331 11.6 0.066 0.073 3.623 3.636 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.516 3.532 grid_integrate_task_list 110 12.3 3.172 3.419 3.172 3.419 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.311 3.312 fft_wrap_pw1pw2_140 451 13.1 0.640 0.647 3.055 3.071 wfi_extrapolate 11 7.9 0.001 0.001 2.954 2.954 make_images_data 4110 15.4 0.046 0.051 2.333 2.808 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.792 2.800 cp_fm_diag_elpa_base 48 14.0 2.724 2.754 2.790 2.798 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.794 2.795 hybrid_alltoall_any 4261 16.3 0.109 0.385 2.086 2.771 calculate_dm_sparse 110 9.5 0.001 0.001 2.525 2.560 cp_fm_cholesky_invert 11 10.9 2.553 2.558 2.553 2.558 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.480 2.526 grid_collocate_task_list 110 9.6 2.120 2.523 2.120 2.523 multiply_cannon_metrocomm4 14385 15.4 0.047 0.051 0.894 2.493 mp_waitany 17072 13.8 1.189 2.483 1.189 2.483 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.397 2.463 calculate_first_density_matrix 1 7.0 0.000 0.000 2.405 2.407 mp_irecv_dv 48980 15.7 0.819 2.362 0.819 2.362 fft3d_ps 1111 14.6 1.091 1.106 2.335 2.350 potential_pw2rs 110 12.3 0.011 0.011 2.261 2.271 rs_pw_transfer_RS2PW_140 121 11.5 0.179 0.184 0.908 2.148 mp_alltoall_d11v 2046 13.8 1.671 2.078 1.671 2.078 mp_sum_l 6594 12.7 1.359 2.012 1.359 2.012 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.961 1.962 dbcsr_complete_redistribute 325 12.2 0.331 0.364 1.432 1.898 cp_fm_upper_to_full 70 14.2 1.401 1.792 1.401 1.792 acc_transpose_blocks 16440 15.4 0.079 0.081 1.612 1.723 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.694 1.715 cp_fm_cholesky_decompose 22 10.9 1.537 1.557 1.537 1.557 mp_allgather_i34 2055 14.4 0.468 1.557 0.468 1.557 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.469 1.480 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.365 1.471 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.964 1.411 rs_gather_matrices 110 12.3 0.233 0.264 0.961 1.345 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.264 1.272 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.055000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=632.636364, yerr=7.425665 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_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 744.239104E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.049 0.083 66.220 66.221 qs_mol_dyn_low 1 2.0 0.003 0.004 65.905 65.915 qs_forces 11 3.9 0.006 0.033 65.796 65.798 qs_energies 11 4.9 0.005 0.033 62.315 62.318 scf_env_do_scf 11 5.9 0.001 0.001 53.884 53.887 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.386 42.386 velocity_verlet 10 3.0 0.005 0.009 37.319 37.323 dbcsr_multiply_generic 2055 12.4 0.111 0.116 30.677 30.943 qs_scf_new_mos 99 7.5 0.001 0.001 27.313 27.430 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.312 27.430 ot_scf_mini 99 9.5 0.003 0.003 25.644 25.760 multiply_cannon 2055 13.4 0.236 0.250 23.601 24.774 multiply_cannon_loop 2055 14.4 1.393 1.473 22.195 22.883 ot_mini 99 10.5 0.001 0.001 14.874 14.996 multiply_cannon_multrec 24660 15.4 4.122 6.820 13.035 14.290 rebuild_ks_matrix 110 8.3 0.000 0.000 12.261 12.372 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.261 12.372 init_scf_loop 11 6.9 0.001 0.004 11.459 11.460 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.806 10.902 qs_ot_get_derivative 99 11.5 0.001 0.001 10.707 10.818 dbcsr_mm_accdrv_process 52304 16.0 7.734 8.926 8.756 9.753 prepare_preconditioner 11 7.9 0.000 0.000 9.700 9.715 make_preconditioner 11 8.9 0.000 0.001 9.700 9.715 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.904 9.380 sum_up_and_integrate 110 10.3 0.067 0.070 6.488 6.501 integrate_v_rspace 110 11.3 0.003 0.004 6.421 6.434 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.145 6.155 calculate_rho_elec 110 8.6 0.078 0.082 6.144 6.155 mp_waitall_1 121746 16.5 4.322 6.090 4.322 6.090 make_m2s 4110 13.4 0.058 0.061 5.447 5.830 qs_ot_get_p 110 10.4 0.001 0.001 5.624 5.792 make_images 4110 14.4 0.581 0.701 5.308 5.686 init_scf_run 11 5.9 0.000 0.001 5.590 5.591 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.590 5.591 cp_fm_upper_to_full 70 14.2 3.342 4.860 3.342 4.860 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.058 4.155 apply_single 110 13.6 0.000 0.000 4.058 4.155 ot_diis_step 99 11.5 0.011 0.011 4.130 4.131 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.808 3.827 dbcsr_complete_redistribute 325 12.2 0.423 0.463 2.623 3.769 pw_transfer 1331 11.6 0.066 0.075 3.709 3.740 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.603 3.637 density_rs2pw 110 9.6 0.004 0.005 3.018 3.546 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.463 3.523 grid_integrate_task_list 110 12.3 3.267 3.413 3.267 3.413 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.494 3.390 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.311 3.376 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.124 3.263 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.244 3.244 fft_wrap_pw1pw2_140 451 13.1 0.671 0.694 3.112 3.149 hybrid_alltoall_any 4261 16.3 0.123 0.458 2.311 3.146 make_images_data 4110 15.4 0.049 0.052 2.745 3.113 calculate_dm_sparse 110 9.5 0.001 0.001 2.967 2.998 wfi_extrapolate 11 7.9 0.001 0.001 2.983 2.983 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.787 2.910 rs_pw_transfer 902 11.9 0.010 0.011 2.365 2.903 mp_alltoall_i22 605 13.7 1.663 2.839 1.663 2.839 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.791 2.793 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.731 2.743 cp_fm_diag_elpa_base 48 14.0 2.584 2.642 2.728 2.741 cp_fm_cholesky_invert 11 10.9 2.553 2.561 2.553 2.561 multiply_cannon_sync_h2d 24660 15.4 2.366 2.517 2.366 2.517 acc_transpose_blocks 24660 15.4 0.110 0.112 2.384 2.508 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.445 2.487 calculate_first_density_matrix 1 7.0 0.000 0.002 2.476 2.479 grid_collocate_task_list 110 9.6 2.227 2.446 2.227 2.446 fft3d_ps 1111 14.6 1.083 1.113 2.360 2.373 qs_energies_init_hamiltonians 11 5.9 0.002 0.007 2.272 2.273 potential_pw2rs 110 12.3 0.012 0.013 2.117 2.122 mp_alltoall_d11v 2046 13.8 1.741 1.930 1.741 1.930 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.775 1.819 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.699 1.712 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.006 1.614 1.706 cp_fm_cholesky_decompose 22 10.9 1.605 1.647 1.605 1.647 mp_waitany 13376 13.8 1.091 1.614 1.091 1.614 mp_sum_l 6594 12.7 0.961 1.590 0.961 1.590 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.540 1.557 multiply_cannon_metrocomm4 20550 15.4 0.061 0.064 0.854 1.521 acc_transpose_blocks_sync 73980 16.4 1.372 1.485 1.372 1.485 jit_kernel_multiply 8 15.8 0.674 1.437 0.674 1.437 mp_irecv_dv 62702 16.1 0.749 1.433 0.749 1.433 mp_allgather_i34 2055 14.4 0.489 1.432 0.489 1.432 rs_pw_transfer_RS2PW_140 121 11.5 0.171 0.177 0.861 1.403 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.221000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=701.363636, yerr=13.792440 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_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 850.903040E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.027 55.335 55.335 qs_mol_dyn_low 1 2.0 0.003 0.004 55.108 55.117 qs_forces 11 3.9 0.003 0.003 54.980 54.981 qs_energies 11 4.9 0.001 0.001 51.288 51.293 scf_env_do_scf 11 5.9 0.000 0.001 42.931 42.931 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.334 35.335 velocity_verlet 10 3.0 0.001 0.002 31.188 31.195 dbcsr_multiply_generic 2055 12.4 0.103 0.108 23.039 23.261 qs_scf_new_mos 99 7.5 0.001 0.001 20.208 20.254 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.208 20.253 multiply_cannon 2055 13.4 0.237 0.246 17.570 19.056 ot_scf_mini 99 9.5 0.002 0.002 18.962 18.993 multiply_cannon_loop 2055 14.4 0.602 0.628 16.317 16.720 rebuild_ks_matrix 110 8.3 0.000 0.000 11.883 11.922 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.883 11.922 ot_mini 99 10.5 0.001 0.001 10.570 10.593 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.542 10.573 multiply_cannon_multrec 8220 15.4 3.194 4.394 7.587 8.564 init_scf_loop 11 6.9 0.000 0.000 7.552 7.553 mp_waitall_1 103326 16.6 5.684 7.165 5.684 7.165 qs_ot_get_derivative 99 11.5 0.001 0.001 6.891 6.921 sum_up_and_integrate 110 10.3 0.079 0.080 6.494 6.509 integrate_v_rspace 110 11.3 0.003 0.003 6.415 6.429 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.262 6.277 calculate_rho_elec 110 8.6 0.115 0.115 6.261 6.277 prepare_preconditioner 11 7.9 0.000 0.000 5.896 5.899 make_preconditioner 11 8.9 0.000 0.000 5.896 5.899 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.486 5.561 init_scf_run 11 5.9 0.000 0.001 5.235 5.235 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.234 5.235 dbcsr_mm_accdrv_process 17442 15.9 3.129 4.350 4.253 5.187 qs_ot_get_p 110 10.4 0.001 0.001 4.470 4.489 make_m2s 4110 13.4 0.038 0.038 4.179 4.464 make_images 4110 14.4 0.652 0.719 4.051 4.337 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 2.972 4.329 pw_transfer 1331 11.6 0.066 0.071 3.881 3.893 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.726 3.801 apply_single 110 13.6 0.000 0.000 3.726 3.801 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.773 3.788 ot_diis_step 99 11.5 0.012 0.012 3.657 3.657 grid_integrate_task_list 110 12.3 3.365 3.507 3.365 3.507 density_rs2pw 110 9.6 0.004 0.004 2.979 3.350 fft_wrap_pw1pw2_140 451 13.1 0.834 0.848 3.271 3.301 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.157 3.161 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.824 2.824 cp_fm_cholesky_invert 11 10.9 2.786 2.789 2.786 2.789 hybrid_alltoall_any 4261 16.3 0.201 0.856 2.217 2.745 wfi_extrapolate 11 7.9 0.001 0.001 2.701 2.701 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.671 2.672 make_images_data 4110 15.4 0.041 0.046 2.238 2.632 grid_collocate_task_list 110 9.6 2.317 2.577 2.317 2.577 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.570 2.572 rs_pw_transfer 902 11.9 0.010 0.010 2.142 2.540 calculate_dm_sparse 110 9.5 0.001 0.001 2.483 2.520 multiply_cannon_sync_h2d 8220 15.4 2.374 2.495 2.374 2.495 calculate_first_density_matrix 1 7.0 0.000 0.000 2.433 2.434 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.332 2.340 cp_fm_diag_elpa_base 48 14.0 2.276 2.297 2.331 2.338 fft3d_ps 1111 14.6 1.136 1.192 2.318 2.335 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.204 2.216 potential_pw2rs 110 12.3 0.015 0.015 2.036 2.037 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.989 2.018 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.785 1.993 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.865 1.880 mp_alltoall_d11v 2046 13.8 1.591 1.816 1.591 1.816 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.670 1.677 cp_fm_cholesky_decompose 22 10.9 1.659 1.675 1.659 1.675 qs_env_update_s_mstruct 11 6.9 0.001 0.004 1.495 1.614 mp_allgather_i34 2055 14.4 0.428 1.511 0.428 1.511 mp_waitany 9240 13.8 1.076 1.506 1.076 1.506 dbcsr_complete_redistribute 325 12.2 0.574 0.607 1.416 1.505 acc_transpose_blocks 8220 15.4 0.039 0.040 1.423 1.477 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.440 1.451 qs_create_task_list 11 7.9 0.000 0.000 1.210 1.318 generate_qs_task_list 11 8.9 0.373 0.441 1.210 1.318 rs_pw_transfer_RS2PW_140 121 11.5 0.163 0.167 0.807 1.208 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.144 1.165 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.335000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=803.181818, yerr=9.113608 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_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 1.379484E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.035 90.944 90.945 qs_mol_dyn_low 1 2.0 0.003 0.004 90.641 90.651 qs_forces 11 3.9 0.003 0.003 89.883 89.896 qs_energies 11 4.9 0.001 0.001 85.652 85.666 scf_env_do_scf 11 5.9 0.000 0.001 75.384 75.384 velocity_verlet 10 3.0 0.020 0.021 58.328 58.411 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 46.213 46.215 dbcsr_multiply_generic 2055 12.4 0.119 0.123 30.091 30.223 init_scf_loop 11 6.9 0.000 0.000 29.095 29.098 qs_scf_new_mos 99 7.5 0.001 0.001 27.652 27.734 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.651 27.734 prepare_preconditioner 11 7.9 0.000 0.000 27.018 27.032 make_preconditioner 11 8.9 0.000 0.000 27.018 27.032 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.037 26.477 ot_scf_mini 99 9.5 0.002 0.002 25.840 25.919 multiply_cannon 2055 13.4 0.336 0.369 22.676 23.364 multiply_cannon_loop 2055 14.4 0.824 0.839 20.842 21.221 cp_fm_upper_to_full 70 14.2 13.069 18.765 13.069 18.765 ot_mini 99 10.5 0.001 0.001 14.475 14.553 rebuild_ks_matrix 110 8.3 0.000 0.001 14.378 14.469 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.378 14.469 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.003 13.085 dbcsr_complete_redistribute 325 12.2 1.015 1.051 7.614 10.963 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.553 9.907 multiply_cannon_multrec 8220 15.4 4.048 4.219 9.735 9.869 qs_ot_get_derivative 99 11.5 0.001 0.001 9.648 9.727 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.965 9.286 mp_alltoall_i22 605 13.7 5.601 8.939 5.601 8.939 mp_waitall_1 84994 16.7 7.879 8.901 7.879 8.901 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.837 7.867 calculate_rho_elec 110 8.6 0.225 0.226 7.836 7.866 sum_up_and_integrate 110 10.3 0.150 0.150 7.455 7.471 integrate_v_rspace 110 11.3 0.004 0.004 7.305 7.321 make_m2s 4110 13.4 0.043 0.043 5.529 6.050 qs_ot_get_p 110 10.4 0.001 0.001 5.971 6.050 init_scf_run 11 5.9 0.000 0.001 5.896 5.896 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.896 5.896 make_images 4110 14.4 0.888 0.944 5.341 5.861 dbcsr_mm_accdrv_process 11614 15.7 3.869 4.114 5.538 5.753 cp_fm_cholesky_invert 11 10.9 5.419 5.424 5.419 5.424 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.890 5.391 apply_single 110 13.6 0.000 0.000 4.889 5.391 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 5.074 5.333 pw_transfer 1331 11.6 0.075 0.076 5.133 5.140 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.016 5.022 ot_diis_step 99 11.5 0.016 0.016 4.804 4.804 fft_wrap_pw1pw2_140 451 13.1 1.336 1.342 4.357 4.365 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.259 4.268 density_rs2pw 110 9.6 0.004 0.004 3.800 3.826 grid_integrate_task_list 110 12.3 3.701 3.784 3.701 3.784 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.730 3.744 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.739 3.740 hybrid_alltoall_any 4261 16.3 0.264 0.568 2.890 3.632 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.090 3.558 make_images_data 4110 15.4 0.045 0.048 2.895 3.545 wfi_extrapolate 11 7.9 0.001 0.001 3.451 3.451 calculate_dm_sparse 110 9.5 0.001 0.001 3.219 3.238 multiply_cannon_sync_h2d 8220 15.4 3.127 3.150 3.127 3.150 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.127 3.127 cp_fm_diag_elpa_base 48 14.0 2.572 2.780 3.125 3.125 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.966 2.971 fft3d_ps 1111 14.6 1.325 1.335 2.945 2.957 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.677 2.719 grid_collocate_task_list 110 9.6 2.688 2.708 2.688 2.708 potential_pw2rs 110 12.3 0.021 0.021 2.490 2.497 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.439 2.469 rs_pw_transfer 902 11.9 0.010 0.011 2.357 2.396 calculate_first_density_matrix 1 7.0 0.000 0.000 2.321 2.324 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.222 2.291 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.105 2.207 mp_alltoall_d11v 2046 13.8 2.054 2.165 2.054 2.165 cp_fm_cholesky_decompose 22 10.9 2.029 2.050 2.029 2.050 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.975 1.984 qs_create_task_list 11 7.9 0.000 0.000 1.898 1.955 generate_qs_task_list 11 8.9 0.739 0.793 1.898 1.955 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.829 1.884 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=90.945000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1238.454545, yerr=47.790002 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 632.369152E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2718423. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.076 210.968 210.974 qs_mol_dyn_low 1 2.0 0.006 0.026 210.354 210.369 qs_forces 11 3.9 0.009 0.043 210.195 210.220 qs_energies 11 4.9 0.001 0.003 204.467 204.485 scf_env_do_scf 11 5.9 0.001 0.001 187.866 187.870 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 167.203 167.204 dbcsr_multiply_generic 2507 12.6 0.178 0.184 126.393 127.160 qs_scf_new_mos 117 7.6 0.001 0.001 126.080 126.312 qs_scf_loop_do_ot 117 8.6 0.001 0.001 126.079 126.311 velocity_verlet 10 3.0 0.004 0.020 125.983 125.984 ot_scf_mini 117 9.6 0.003 0.004 119.338 119.599 multiply_cannon 2507 13.6 0.238 0.249 101.858 104.000 multiply_cannon_loop 2507 14.6 2.426 2.475 99.632 101.745 ot_mini 117 10.6 0.001 0.002 66.718 66.954 multiply_cannon_multrec 60168 15.6 31.939 34.297 41.679 43.241 qs_ot_get_derivative 117 11.6 0.001 0.002 41.829 42.044 rebuild_ks_matrix 128 8.3 0.001 0.001 34.141 34.539 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.016 34.140 34.538 mp_waitall_1 267128 16.5 29.244 32.698 29.244 32.698 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.607 30.994 qs_ot_get_p 128 10.4 0.001 0.001 30.280 30.596 multiply_cannon_sync_h2d 60168 15.6 26.200 27.694 26.200 27.694 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.361 25.228 apply_single 128 13.6 0.001 0.001 24.360 25.228 ot_diis_step 117 11.6 0.010 0.032 24.642 24.660 qs_ot_p2m_diag 83 11.4 0.079 0.092 23.553 23.637 cp_dbcsr_syevd 83 12.4 0.005 0.006 20.651 20.653 init_scf_loop 11 6.9 0.000 0.001 20.591 20.592 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.154 20.347 multiply_cannon_metrocomm3 60168 15.6 0.119 0.124 16.360 18.797 cp_fm_diag_elpa 83 13.4 0.000 0.001 17.592 17.625 cp_fm_diag_elpa_base 83 14.4 17.519 17.567 17.587 17.621 prepare_preconditioner 11 7.9 0.000 0.000 15.954 16.017 make_preconditioner 11 8.9 0.000 0.000 15.954 16.017 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.164 15.337 sum_up_and_integrate 128 10.3 0.089 0.107 14.501 14.519 integrate_v_rspace 128 11.3 0.004 0.008 14.412 14.434 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.224 14.352 calculate_rho_elec 128 8.7 0.046 0.065 14.224 14.351 make_m2s 5014 13.6 0.104 0.114 14.021 14.345 make_images 5014 14.6 0.399 0.420 13.844 14.176 init_scf_run 11 5.9 0.000 0.001 12.388 12.388 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.387 12.388 density_rs2pw 128 9.7 0.006 0.007 7.408 10.678 mp_sum_l 7950 12.9 9.308 10.588 9.308 10.588 dbcsr_mm_accdrv_process 124484 16.2 4.699 4.834 9.314 9.840 rs_pw_transfer 1046 11.9 0.016 0.019 6.106 9.352 wfi_extrapolate 11 7.9 0.001 0.001 9.168 9.168 cp_fm_cholesky_invert 11 10.9 9.049 9.057 9.049 9.057 calculate_dm_sparse 128 9.5 0.001 0.001 8.569 8.704 multiply_cannon_metrocomm1 60168 15.6 0.095 0.098 6.199 8.587 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.253 8.396 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.145 8.211 pw_transfer 1547 11.6 0.075 0.092 7.984 8.175 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 7.780 7.972 make_images_data 5014 15.6 0.071 0.077 6.863 7.823 grid_integrate_task_list 128 12.3 7.071 7.632 7.071 7.632 hybrid_alltoall_any 5200 16.5 0.298 2.280 6.000 7.242 fft_wrap_pw1pw2_140 523 13.2 1.277 1.320 6.733 6.932 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.654 6.666 mp_waitany 16020 13.9 2.775 6.056 2.775 6.056 grid_collocate_task_list 128 9.7 4.670 5.949 4.670 5.949 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.777 5.899 fft3d_ps 1291 14.7 2.135 2.767 5.541 5.851 mp_alltoall_d11v 2415 14.1 4.442 5.844 4.442 5.844 rs_pw_transfer_RS2PW_140 139 11.5 0.281 0.305 2.190 5.448 potential_pw2rs 128 12.3 0.009 0.011 4.801 4.841 cp_fm_cholesky_decompose 22 10.9 4.707 4.721 4.707 4.721 mp_sum_d 4465 12.1 3.533 4.250 3.533 4.250 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=210.974000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.000000, yerr=6.252272 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_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 838.938624E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57903. MP_Allreduce 11138 958. MP_Sync 87 MP_Alltoall 1983 6488560. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.037 194.097 194.098 qs_mol_dyn_low 1 2.0 0.003 0.004 193.709 193.751 qs_forces 11 3.9 0.004 0.005 193.407 193.410 qs_energies 11 4.9 0.001 0.001 186.562 186.576 scf_env_do_scf 11 5.9 0.001 0.001 169.810 169.819 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 136.282 136.286 velocity_verlet 10 3.0 0.001 0.002 121.783 121.789 dbcsr_multiply_generic 2527 12.6 0.190 0.196 99.228 100.407 qs_scf_new_mos 118 7.6 0.001 0.001 96.192 96.845 qs_scf_loop_do_ot 118 8.6 0.001 0.001 96.191 96.844 ot_scf_mini 118 9.6 0.004 0.004 91.319 91.918 multiply_cannon 2527 13.6 0.506 0.564 78.357 82.744 multiply_cannon_loop 2527 14.6 1.582 1.650 74.796 77.532 ot_mini 118 10.6 0.001 0.001 51.172 51.806 mp_waitall_1 216438 16.6 24.839 39.935 24.839 39.935 multiply_cannon_multrec 30324 15.6 21.084 25.642 31.696 36.503 rebuild_ks_matrix 129 8.3 0.001 0.001 33.218 33.976 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 33.217 33.975 init_scf_loop 11 6.9 0.000 0.000 33.439 33.442 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.867 30.546 qs_ot_get_derivative 118 11.6 0.001 0.002 28.950 29.546 multiply_cannon_metrocomm3 30324 15.6 0.097 0.103 15.996 29.227 prepare_preconditioner 11 7.9 0.000 0.000 28.821 28.889 make_preconditioner 11 8.9 0.000 0.000 28.821 28.889 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.524 28.061 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.244 23.409 apply_single 129 13.6 0.001 0.001 22.243 23.408 qs_ot_get_p 129 10.4 0.001 0.001 22.024 22.771 ot_diis_step 118 11.6 0.014 0.015 22.048 22.050 multiply_cannon_sync_h2d 30324 15.6 18.094 20.944 18.094 20.944 cp_fm_cholesky_invert 11 10.9 17.105 17.117 17.105 17.117 make_m2s 5054 13.6 0.093 0.101 14.723 16.523 make_images 5054 14.6 1.187 1.401 14.513 16.311 qs_ot_p2m_diag 83 11.4 0.189 0.217 16.163 16.196 cp_dbcsr_syevd 83 12.4 0.006 0.006 14.970 14.970 sum_up_and_integrate 129 10.3 0.117 0.133 14.740 14.770 integrate_v_rspace 129 11.3 0.004 0.004 14.623 14.656 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.174 14.202 calculate_rho_elec 129 8.7 0.090 0.107 14.173 14.201 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.785 12.217 init_scf_run 11 5.9 0.000 0.001 11.957 11.958 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.957 11.958 cp_fm_diag_elpa 83 13.4 0.001 0.001 11.855 11.884 cp_fm_diag_elpa_base 83 14.4 11.599 11.687 11.849 11.873 multiply_cannon_metrocomm4 27797 15.6 0.108 0.121 3.901 10.823 make_images_data 5054 15.6 0.069 0.080 8.672 10.789 dbcsr_mm_accdrv_process 62734 16.2 5.406 6.087 10.062 10.617 mp_irecv_dv 70031 16.3 3.690 10.411 3.690 10.411 hybrid_alltoall_any 5240 16.5 0.352 1.525 7.337 10.296 density_rs2pw 129 9.7 0.007 0.008 7.425 9.946 pw_transfer 1559 11.6 0.087 0.100 8.767 8.835 wfi_extrapolate 11 7.9 0.001 0.001 8.678 8.679 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 8.540 8.612 rs_pw_transfer 1054 12.0 0.015 0.017 5.649 8.164 fft_wrap_pw1pw2_140 527 13.2 1.350 1.372 7.581 7.675 grid_integrate_task_list 129 12.3 7.231 7.659 7.231 7.659 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.391 7.208 cp_fm_cholesky_decompose 22 10.9 6.845 6.946 6.845 6.946 calculate_dm_sparse 129 9.5 0.001 0.001 6.542 6.685 mp_sum_l 8010 12.9 4.331 6.597 4.331 6.597 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.292 6.301 grid_collocate_task_list 129 9.7 4.815 6.003 4.815 6.003 fft3d_ps 1301 14.7 2.833 3.014 5.819 5.858 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.390 5.474 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.256 5.466 mp_allgather_i34 2527 14.6 2.075 5.293 2.075 5.293 mp_waitany 11836 13.9 2.509 5.026 2.509 5.026 mp_alltoall_d11v 2423 14.1 4.267 4.918 4.267 4.918 potential_pw2rs 129 12.3 0.016 0.019 4.898 4.912 rs_pw_transfer_RS2PW_140 140 11.5 0.357 0.381 2.122 4.604 dbcsr_complete_redistribute 395 12.7 0.775 0.906 3.341 4.170 mp_sum_d 4499 12.1 2.733 4.156 2.733 4.156 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=194.098000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=799.272727, yerr=2.415515 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/17/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 2.906045E+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 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 950.824960E+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 925172558800 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3995 58307. MP_Allreduce 11065 1000. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 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 0.016 0.046 177.044 177.044 qs_mol_dyn_low 1 2.0 0.003 0.005 176.452 176.465 qs_forces 11 3.9 0.005 0.005 176.363 176.365 qs_energies 11 4.9 0.001 0.001 169.712 169.723 scf_env_do_scf 11 5.9 0.001 0.001 153.662 153.663 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 119.294 119.294 velocity_verlet 10 3.0 0.001 0.002 112.710 112.712 dbcsr_multiply_generic 2485 12.5 0.177 0.183 82.527 83.746 qs_scf_new_mos 116 7.6 0.001 0.001 81.340 81.668 qs_scf_loop_do_ot 116 8.6 0.001 0.001 81.339 81.667 ot_scf_mini 116 9.6 0.003 0.004 77.183 77.562 multiply_cannon 2485 13.5 0.506 0.523 62.418 67.298 multiply_cannon_loop 2485 14.5 1.117 1.180 59.210 61.720 ot_mini 116 10.6 0.001 0.001 42.404 42.789 mp_waitall_1 169034 16.6 24.920 34.670 24.920 34.670 init_scf_loop 11 6.9 0.000 0.000 34.272 34.273 rebuild_ks_matrix 127 8.3 0.001 0.001 30.520 31.031 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.017 30.520 31.031 prepare_preconditioner 11 7.9 0.000 0.000 30.261 30.304 make_preconditioner 11 8.9 0.000 0.000 30.261 30.304 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.910 29.316 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.396 27.856 multiply_cannon_multrec 19880 15.5 13.015 16.471 22.299 25.849 multiply_cannon_metrocomm3 19880 15.5 0.061 0.063 15.137 24.887 qs_ot_get_derivative 116 11.6 0.001 0.002 22.978 23.367 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.513 20.559 apply_single 127 13.6 0.001 0.001 19.512 20.558 qs_ot_get_p 127 10.4 0.001 0.001 19.764 20.261 ot_diis_step 116 11.6 0.017 0.018 19.309 19.310 make_m2s 4970 13.5 0.081 0.085 14.892 16.313 make_images 4970 14.5 1.140 1.233 14.661 16.079 multiply_cannon_sync_h2d 19880 15.5 13.489 15.506 13.489 15.506 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.768 14.796 calculate_rho_elec 127 8.7 0.131 0.146 14.768 14.795 sum_up_and_integrate 127 10.3 0.132 0.144 14.508 14.532 qs_ot_p2m_diag 82 11.4 0.262 0.269 14.420 14.444 integrate_v_rspace 127 11.3 0.004 0.004 14.376 14.404 cp_fm_cholesky_invert 11 10.9 14.013 14.021 14.013 14.021 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.107 13.124 init_scf_run 11 5.9 0.000 0.001 11.033 11.033 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.032 11.033 make_images_data 4970 15.5 0.063 0.073 8.763 10.651 hybrid_alltoall_any 5155 16.4 0.445 2.017 7.639 9.790 cp_fm_diag_elpa 82 13.4 0.000 0.001 9.704 9.724 cp_fm_diag_elpa_base 82 14.4 9.257 9.439 9.695 9.716 density_rs2pw 127 9.7 0.006 0.007 7.461 9.369 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.059 9.339 multiply_cannon_metrocomm4 17395 15.5 0.064 0.074 3.435 9.196 mp_irecv_dv 49801 16.2 3.310 8.938 3.310 8.938 pw_transfer 1535 11.6 0.085 0.103 8.776 8.883 dbcsr_mm_accdrv_process 41158 16.2 5.610 5.920 8.741 8.854 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.554 8.665 wfi_extrapolate 11 7.9 0.001 0.001 7.847 7.847 fft_wrap_pw1pw2_140 519 13.2 1.405 1.434 7.615 7.740 grid_integrate_task_list 127 12.3 7.244 7.664 7.244 7.664 rs_pw_transfer 1038 11.9 0.013 0.015 5.392 7.310 cp_fm_upper_to_full 104 14.8 5.726 7.282 5.726 7.282 cp_fm_cholesky_decompose 22 10.9 7.154 7.189 7.154 7.189 dbcsr_complete_redistribute 393 12.7 1.170 1.198 4.726 6.497 calculate_dm_sparse 127 9.5 0.001 0.001 5.756 5.859 fft3d_ps 1281 14.7 2.716 2.935 5.691 5.774 grid_collocate_task_list 127 9.7 4.951 5.754 4.951 5.754 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.393 5.399 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.521 5.286 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.623 5.259 mp_alltoall_d11v 2401 14.1 4.543 5.098 4.543 5.098 mp_sum_l 7884 12.9 3.403 4.881 3.403 4.881 mp_allgather_i34 2485 14.5 1.717 4.718 1.717 4.718 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.524 4.676 potential_pw2rs 127 12.3 0.020 0.022 4.605 4.618 mp_waitany 11660 13.9 2.315 4.281 2.315 4.281 transfer_fm_to_dbcsr 11 9.9 0.017 0.022 2.332 4.077 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.054 4.076 mp_alltoall_i22 712 14.1 2.000 3.927 2.000 3.927 rs_pw_transfer_RS2PW_140 138 11.5 0.333 0.352 1.900 3.797 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.755 3.756 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.618 3.670 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.044000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=901.909091, yerr=10.561117 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239167967232 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239167967232 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233007E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383584 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.152123E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+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 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922014800 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57626. MP_Allreduce 11184 1079. MP_Sync 88 MP_Alltoall 1724 12509628. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.114 189.852 189.853 qs_mol_dyn_low 1 2.0 0.003 0.005 189.248 189.261 qs_forces 11 3.9 0.004 0.004 189.142 189.145 qs_energies 11 4.9 0.013 0.094 181.912 181.921 scf_env_do_scf 11 5.9 0.001 0.001 165.010 165.024 velocity_verlet 10 3.0 0.002 0.002 125.736 125.739 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 118.931 118.932 dbcsr_multiply_generic 2529 12.6 0.187 0.193 81.403 82.161 qs_scf_new_mos 118 7.6 0.001 0.001 80.977 81.231 qs_scf_loop_do_ot 118 8.6 0.001 0.002 80.976 81.231 ot_scf_mini 118 9.6 0.003 0.004 76.365 76.674 multiply_cannon 2529 13.6 0.557 0.587 56.199 59.739 multiply_cannon_loop 2529 14.6 1.841 1.940 52.572 54.328 init_scf_loop 11 6.9 0.008 0.063 45.960 45.967 ot_mini 118 10.6 0.001 0.001 43.561 43.841 prepare_preconditioner 11 7.9 0.000 0.000 41.550 41.578 make_preconditioner 11 8.9 0.000 0.001 41.550 41.578 make_full_inverse_cholesky 11 9.9 0.012 0.022 35.246 40.232 multiply_cannon_multrec 30348 15.6 13.803 18.986 26.774 31.543 rebuild_ks_matrix 129 8.3 0.001 0.001 30.400 30.719 qs_ks_build_kohn_sham_matrix 129 9.3 0.022 0.058 30.400 30.718 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.415 27.697 mp_waitall_1 149172 16.7 17.367 26.448 17.367 26.448 qs_ot_get_derivative 118 11.6 0.002 0.002 23.689 23.995 make_m2s 5058 13.6 0.096 0.101 20.615 21.402 make_images 5058 14.6 1.974 2.332 20.306 21.091 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.247 19.831 apply_single 129 13.6 0.001 0.001 19.247 19.830 ot_diis_step 118 11.6 0.018 0.019 19.742 19.743 qs_ot_get_p 129 10.4 0.001 0.002 18.853 19.174 cp_fm_upper_to_full 106 14.8 11.349 16.734 11.349 16.734 cp_fm_cholesky_invert 11 10.9 15.789 15.798 15.789 15.798 multiply_cannon_metrocomm3 30348 15.6 0.050 0.054 6.344 15.226 sum_up_and_integrate 129 10.3 0.141 0.152 14.802 14.826 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.765 14.813 calculate_rho_elec 129 8.7 0.177 0.193 14.764 14.812 integrate_v_rspace 129 11.3 0.004 0.005 14.661 14.689 qs_ot_p2m_diag 84 11.4 0.347 0.394 14.603 14.656 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.193 13.195 dbcsr_mm_accdrv_process 62780 16.2 8.510 9.475 12.537 13.083 dbcsr_complete_redistribute 397 12.7 1.502 1.616 9.015 12.732 make_images_data 5058 15.6 0.067 0.072 11.158 12.561 multiply_cannon_sync_h2d 30348 15.6 10.606 11.441 10.606 11.441 copy_fm_to_dbcsr 210 11.7 0.002 0.002 7.691 11.401 hybrid_alltoall_any 5245 16.5 0.535 2.236 9.798 11.139 init_scf_run 11 5.9 0.000 0.001 10.930 10.931 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.930 10.931 qs_ot_get_derivative_diag 78 12.4 0.009 0.055 10.099 10.319 cp_fm_diag_elpa 84 13.4 0.000 0.001 9.993 10.009 cp_fm_diag_elpa_base 84 14.4 9.032 9.364 9.985 9.999 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.281 9.921 mp_alltoall_i22 720 14.1 5.689 9.568 5.689 9.568 pw_transfer 1559 11.6 0.087 0.102 9.212 9.289 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.984 9.068 density_rs2pw 129 9.7 0.010 0.034 7.041 8.595 fft_wrap_pw1pw2_140 527 13.2 1.577 1.603 7.979 8.066 grid_integrate_task_list 129 12.3 7.555 8.050 7.555 8.050 wfi_extrapolate 11 7.9 0.001 0.001 7.709 7.709 cp_fm_cholesky_decompose 22 10.9 7.372 7.467 7.372 7.467 multiply_cannon_metrocomm4 25290 15.6 0.084 0.093 2.825 7.075 mp_irecv_dv 76751 16.2 2.668 6.792 2.668 6.792 calculate_dm_sparse 129 9.5 0.001 0.001 6.329 6.426 rs_pw_transfer 1054 12.0 0.013 0.014 4.631 6.253 fft3d_ps 1301 14.7 2.858 2.914 5.877 5.925 grid_collocate_task_list 129 9.7 5.188 5.880 5.188 5.880 mp_alltoall_d11v 2429 14.1 4.988 5.757 4.988 5.757 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.496 5.593 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.484 4.563 potential_pw2rs 129 12.3 0.023 0.023 4.481 4.496 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.407 4.487 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.468 4.468 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.234 4.313 mp_waitany 15064 13.9 2.296 3.880 2.296 3.880 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.853000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1091.181818, yerr=17.240640 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/19/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 5.909858E+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 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.527685E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+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 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591885696 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57892. MP_Allreduce 11137 1164. MP_Sync 87 MP_Alltoall 1724 18848418. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.035 171.020 171.020 qs_mol_dyn_low 1 2.0 0.003 0.004 170.541 170.573 qs_forces 11 3.9 0.004 0.004 170.443 170.449 qs_energies 11 4.9 0.001 0.001 162.782 162.788 scf_env_do_scf 11 5.9 0.001 0.001 145.466 145.472 velocity_verlet 10 3.0 0.002 0.002 112.944 112.948 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 109.918 109.920 dbcsr_multiply_generic 2527 12.6 0.180 0.192 72.142 72.539 qs_scf_new_mos 118 7.6 0.001 0.001 71.014 71.124 qs_scf_loop_do_ot 118 8.6 0.001 0.001 71.014 71.124 ot_scf_mini 118 9.6 0.003 0.004 66.563 66.656 multiply_cannon 2527 13.6 0.561 0.594 53.453 56.864 multiply_cannon_loop 2527 14.6 0.817 0.840 50.555 51.330 ot_mini 118 10.6 0.001 0.001 37.473 37.560 init_scf_loop 11 6.9 0.000 0.000 35.401 35.404 prepare_preconditioner 11 7.9 0.000 0.000 31.465 31.489 make_preconditioner 11 8.9 0.000 0.000 31.465 31.489 mp_waitall_1 126780 16.7 24.649 30.678 24.649 30.678 rebuild_ks_matrix 129 8.3 0.001 0.001 30.132 30.253 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 30.132 30.252 make_full_inverse_cholesky 11 9.9 0.012 0.027 29.251 29.531 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.332 27.444 multiply_cannon_multrec 10108 15.6 10.487 14.588 18.101 21.100 qs_ot_get_derivative 118 11.6 0.002 0.002 20.398 20.490 multiply_cannon_metrocomm3 10108 15.6 0.026 0.027 12.495 19.470 cp_fm_cholesky_invert 11 10.9 17.881 17.886 17.881 17.886 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.048 17.245 apply_single 129 13.6 0.001 0.001 17.048 17.244 ot_diis_step 118 11.6 0.020 0.020 17.008 17.008 qs_ot_get_p 129 10.4 0.001 0.001 16.301 16.395 make_m2s 5054 13.6 0.067 0.071 14.841 15.667 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.455 15.488 calculate_rho_elec 129 8.7 0.261 0.271 15.454 15.487 make_images 5054 14.6 2.197 2.622 14.533 15.356 sum_up_and_integrate 129 10.3 0.181 0.190 15.085 15.130 integrate_v_rspace 129 11.3 0.004 0.004 14.904 14.959 qs_ot_p2m_diag 83 11.4 0.496 0.501 12.571 12.588 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.415 11.415 multiply_cannon_sync_h2d 10108 15.6 10.964 11.289 10.964 11.289 init_scf_run 11 5.9 0.000 0.001 10.480 10.481 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.480 10.480 hybrid_alltoall_any 5240 16.5 0.853 3.832 8.395 9.835 make_images_data 5054 15.6 0.057 0.067 8.529 9.756 pw_transfer 1559 11.6 0.087 0.096 9.500 9.524 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.275 9.304 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.599 8.612 cp_fm_diag_elpa_base 83 14.4 8.364 8.434 8.596 8.609 fft_wrap_pw1pw2_140 527 13.2 1.928 1.962 8.235 8.266 grid_integrate_task_list 129 12.3 7.807 8.215 7.807 8.215 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 7.941 8.008 dbcsr_mm_accdrv_process 20926 16.1 3.136 4.514 7.259 7.968 cp_fm_cholesky_decompose 22 10.9 7.799 7.894 7.799 7.894 density_rs2pw 129 9.7 0.006 0.007 6.884 7.815 wfi_extrapolate 11 7.9 0.001 0.001 7.466 7.466 multiply_cannon_metrocomm1 10108 15.6 0.030 0.031 4.444 7.450 grid_collocate_task_list 129 9.7 5.520 6.241 5.520 6.241 calculate_dm_sparse 129 9.5 0.001 0.001 6.099 6.168 mp_alltoall_d11v 2423 14.1 5.055 5.910 5.055 5.910 fft3d_ps 1301 14.7 2.779 2.847 5.721 5.759 dbcsr_complete_redistribute 395 12.7 2.118 2.276 5.286 5.642 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.297 5.298 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.136 5.163 rs_pw_transfer 1054 12.0 0.013 0.013 4.135 5.083 mp_allgather_i34 2527 14.6 1.083 4.549 1.083 4.549 potential_pw2rs 129 12.3 0.026 0.026 4.409 4.420 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.063 4.108 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.588 3.917 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.444 3.760 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.665 3.729 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.708 3.718 multiply_cannon_metrocomm4 7581 15.6 0.026 0.028 1.739 3.665 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.583 3.613 mp_irecv_dv 29086 15.9 1.701 3.597 1.701 3.597 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.434 3.521 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=171.020000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1443.909091, yerr=15.836979 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1979979923456 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.250208E+12 0.0% 0.0% 100.0% flops max/rank 11.788211E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806645728 0.0% 0.0% 100.0% number of processed stacks 1983232 0.0% 0.0% 100.0% average stack size 0.0 0.0 3432.1 marketing flops 145.668111E+12 ------------------------------------------------------------------------------- # multiplications 2537 max memory usage/rank 3.175211E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101480 MPI messages size (bytes): total size 1.145238E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.285357E+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 45968 35701915648 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4042 58572. MP_Allreduce 11182 1500. MP_Sync 87 MP_Alltoall 1724 36993632. 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 0.021 0.038 305.898 305.899 qs_mol_dyn_low 1 2.0 0.003 0.004 305.316 305.329 qs_forces 11 3.9 0.004 0.005 305.195 305.198 qs_energies 11 4.9 0.001 0.001 295.683 295.694 scf_env_do_scf 11 5.9 0.001 0.001 272.514 272.521 velocity_verlet 10 3.0 0.002 0.002 220.300 220.311 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 145.424 145.426 init_scf_loop 11 6.9 0.000 0.000 126.816 126.819 prepare_preconditioner 11 7.9 0.000 0.000 121.664 121.684 make_preconditioner 11 8.9 0.000 0.000 121.664 121.684 make_full_inverse_cholesky 11 9.9 0.037 0.039 97.068 118.798 qs_scf_new_mos 118 7.6 0.001 0.001 93.991 94.065 qs_scf_loop_do_ot 118 8.6 0.001 0.001 93.990 94.064 ot_scf_mini 118 9.6 0.004 0.004 88.984 89.023 dbcsr_multiply_generic 2537 12.6 0.212 0.224 84.484 84.881 cp_fm_upper_to_full 105 14.8 53.519 77.114 53.519 77.114 multiply_cannon 2537 13.6 0.694 0.745 60.608 61.640 multiply_cannon_loop 2537 14.6 1.077 1.114 56.477 57.654 ot_mini 118 10.6 0.001 0.001 45.878 45.917 dbcsr_complete_redistribute 395 12.7 3.993 4.045 30.794 44.437 copy_fm_to_dbcsr 209 11.7 0.001 0.001 27.177 40.758 rebuild_ks_matrix 129 8.3 0.001 0.001 38.512 38.560 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 38.512 38.559 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.553 37.894 mp_alltoall_i22 718 14.1 22.340 36.005 22.340 36.005 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.402 35.440 cp_fm_cholesky_invert 11 10.9 34.381 34.387 34.381 34.387 mp_waitall_1 104900 16.8 29.589 34.046 29.589 34.046 qs_ot_get_p 129 10.4 0.001 0.001 27.647 27.718 qs_ot_get_derivative 118 11.6 0.002 0.002 25.573 25.612 qs_ot_p2m_diag 83 11.4 0.879 0.885 23.297 23.327 cp_dbcsr_syevd 83 12.4 0.006 0.006 21.427 21.430 qs_rho_update_rho_low 129 7.7 0.001 0.001 21.249 21.268 calculate_rho_elec 129 8.7 0.484 0.485 21.248 21.267 multiply_cannon_metrocomm3 10148 15.6 0.025 0.027 20.102 21.264 make_m2s 5074 13.6 0.075 0.079 19.023 20.529 ot_diis_step 118 11.6 0.022 0.023 20.267 20.268 sum_up_and_integrate 129 10.3 0.324 0.326 20.028 20.112 make_images 5074 14.6 3.097 3.292 18.543 20.050 integrate_v_rspace 129 11.3 0.004 0.004 19.704 19.786 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.394 19.628 apply_single 129 13.6 0.001 0.001 19.393 19.628 multiply_cannon_multrec 10148 15.6 10.415 12.250 18.232 18.287 cp_fm_diag_elpa 83 13.4 0.000 0.000 18.133 18.135 cp_fm_diag_elpa_base 83 14.4 13.674 15.342 18.129 18.129 multiply_cannon_sync_h2d 10148 15.6 14.388 14.418 14.388 14.418 hybrid_alltoall_any 5260 16.5 1.327 3.096 10.763 13.011 make_images_data 5074 15.6 0.065 0.071 10.647 12.937 init_scf_run 11 5.9 0.000 0.001 12.696 12.697 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.696 12.697 pw_transfer 1559 11.6 0.094 0.095 12.169 12.175 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 11.932 11.938 fft_wrap_pw1pw2_140 527 13.2 3.141 3.193 10.542 10.548 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.992 10.024 wfi_extrapolate 11 7.9 0.001 0.001 9.551 9.551 dbcsr_mm_accdrv_process 20966 16.1 4.242 6.047 7.565 9.451 mp_alltoall_d11v 2423 14.1 8.461 9.112 8.461 9.112 cp_fm_cholesky_decompose 22 10.9 9.014 9.030 9.014 9.030 grid_integrate_task_list 129 12.3 8.665 8.853 8.665 8.853 density_rs2pw 129 9.7 0.005 0.005 8.558 8.712 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.203 8.204 calculate_dm_sparse 129 9.5 0.001 0.001 6.916 6.995 fft3d_ps 1301 14.7 2.875 2.885 6.924 6.972 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.535 6.599 grid_collocate_task_list 129 9.7 6.491 6.562 6.491 6.562 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.337 6.424 rs_scatter_matrices 140 9.7 3.954 4.827 6.185 6.398 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=305.899000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2865.909091, yerr=156.698004 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_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 1.258721E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 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 0.016 0.061 85.652 85.657 qs_energies 1 2.0 0.000 0.000 85.117 85.125 ls_scf 1 3.0 0.000 0.000 84.217 84.225 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.964 73.180 multiply_cannon 111 7.7 0.017 0.020 56.037 57.537 multiply_cannon_loop 111 8.7 0.227 0.243 52.601 54.271 ls_scf_main 1 4.0 0.000 0.000 52.655 52.655 density_matrix_trs4 2 5.0 0.002 0.003 47.046 47.117 ls_scf_init_scf 1 4.0 0.000 0.000 28.523 28.525 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.327 27.390 mp_waitall_1 11031 10.9 22.644 25.459 22.644 25.459 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.228 25.246 multiply_cannon_multrec 2664 9.7 8.149 8.960 15.698 17.516 multiply_cannon_sync_h2d 2664 9.7 13.436 14.945 13.436 14.945 make_m2s 222 7.7 0.009 0.012 13.226 13.666 make_images 222 8.7 0.100 0.111 13.204 13.646 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.861 12.934 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.367 8.490 make_images_data 222 9.7 0.005 0.006 7.747 8.367 dbcsr_mm_accdrv_process 4760 10.4 0.595 0.709 7.168 8.129 hybrid_alltoall_any 227 10.6 0.219 1.859 6.648 7.997 dbcsr_mm_accdrv_process_sort 4760 11.4 6.374 7.277 6.374 7.277 calculate_norms 4752 9.8 5.513 6.156 5.513 6.156 apply_matrix_preconditioner 6 5.3 0.000 0.001 5.018 5.152 mp_sum_l 887 5.1 3.161 4.851 3.161 4.851 make_images_sizes 222 9.7 0.000 0.000 0.775 3.752 mp_alltoall_i44 222 10.7 0.775 3.752 0.775 3.752 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.371 3.617 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.051 3.569 mp_irecv_dv 6231 10.9 2.035 3.546 2.035 3.546 arnoldi_extremal 4 6.8 0.000 0.000 3.341 3.365 arnoldi_normal_ev 4 7.8 0.001 0.003 3.341 3.365 build_subspace 16 8.4 0.009 0.012 3.192 3.194 ls_scf_post 1 4.0 0.000 0.000 3.039 3.046 ls_scf_store_result 1 5.0 0.000 0.000 2.865 2.901 dbcsr_special_finalize 555 9.7 0.006 0.007 2.344 2.832 dbcsr_merge_single_wm 555 10.7 0.467 0.613 2.334 2.822 make_images_pack 222 9.7 2.211 2.640 2.213 2.642 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.364 2.617 dbcsr_sort_data 658 11.4 2.126 2.535 2.126 2.535 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.471 2.071 2.473 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.375 2.427 buffer_matrices_ensure_size 222 8.7 1.770 2.216 1.770 2.216 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.793 1.794 rebuild_ks_matrix 3 7.3 0.000 0.000 1.784 1.785 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.784 1.785 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.657000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_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 2.163622E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.046 90.225 90.226 qs_energies 1 2.0 0.000 0.000 89.663 89.667 ls_scf 1 3.0 0.000 0.000 88.334 88.337 dbcsr_multiply_generic 111 6.7 0.016 0.017 74.422 74.787 multiply_cannon 111 7.7 0.029 0.046 53.250 57.189 ls_scf_main 1 4.0 0.000 0.000 54.289 54.296 multiply_cannon_loop 111 8.7 0.136 0.147 50.520 53.449 density_matrix_trs4 2 5.0 0.002 0.003 48.565 48.778 ls_scf_init_scf 1 4.0 0.000 0.000 30.283 30.285 mp_waitall_1 9105 10.9 20.889 29.694 20.889 29.694 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.073 29.168 multiply_cannon_multrec 1332 9.7 13.335 17.602 22.999 28.212 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.156 26.165 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.538 21.509 make_m2s 222 7.7 0.007 0.008 15.030 15.604 make_images 222 8.7 1.376 1.707 15.000 15.574 dbcsr_mm_accdrv_process 4041 10.4 0.444 1.057 9.262 10.929 dbcsr_mm_accdrv_process_sort 4041 11.4 8.566 10.132 8.566 10.132 make_images_data 222 9.7 0.005 0.006 8.730 9.507 hybrid_alltoall_any 227 10.6 0.543 2.584 8.115 9.295 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.211 7.933 mp_irecv_dv 3311 11.0 3.190 7.876 3.190 7.876 mp_sum_l 887 5.1 4.649 7.853 4.649 7.853 calculate_norms 2376 9.8 6.080 6.886 6.080 6.886 multiply_cannon_sync_h2d 1332 9.7 4.912 6.211 4.912 6.211 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.443 6.196 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.491 5.739 arnoldi_extremal 4 6.8 0.000 0.000 4.618 4.635 arnoldi_normal_ev 4 7.8 0.001 0.005 4.618 4.635 build_subspace 16 8.4 0.014 0.021 4.367 4.370 ls_scf_post 1 4.0 0.000 0.000 3.761 3.764 ls_scf_store_result 1 5.0 0.000 0.000 3.468 3.583 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.152 3.385 dbcsr_matrix_vector_mult_local 304 10.0 2.767 3.246 2.769 3.248 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.271 3.101 mp_allgather_i34 111 8.7 0.710 2.898 0.710 2.898 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.652 2.761 dbcsr_data_new 4174 10.1 2.118 2.413 2.118 2.413 make_images_pack 222 9.7 1.825 2.143 1.828 2.146 dbcsr_sort_data 436 11.2 1.840 2.084 1.840 2.084 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.881 1.883 rebuild_ks_matrix 3 7.3 0.000 0.000 1.868 1.870 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.007 1.868 1.870 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.226000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1806.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_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 2.921058E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.047 93.141 93.142 qs_energies 1 2.0 0.000 0.000 92.601 92.605 ls_scf 1 3.0 0.000 0.000 91.179 91.184 dbcsr_multiply_generic 111 6.7 0.016 0.017 75.726 76.011 ls_scf_main 1 4.0 0.000 0.000 56.700 56.705 multiply_cannon 111 7.7 0.036 0.099 52.103 56.634 multiply_cannon_loop 111 8.7 0.116 0.131 49.330 53.452 density_matrix_trs4 2 5.0 0.002 0.003 50.691 50.873 mp_waitall_1 7281 11.0 23.553 33.688 23.553 33.688 ls_scf_init_scf 1 4.0 0.000 0.000 30.764 30.767 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.519 29.598 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.852 26.864 multiply_cannon_multrec 888 9.7 12.749 15.575 21.478 24.916 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 10.928 22.887 make_m2s 222 7.7 0.007 0.008 16.490 17.256 make_images 222 8.7 1.588 1.866 16.452 17.217 make_images_data 222 9.7 0.004 0.005 9.689 10.832 hybrid_alltoall_any 227 10.6 0.642 2.955 9.229 10.300 dbcsr_mm_accdrv_process 3754 10.4 0.312 0.493 8.248 9.524 mp_sum_l 887 5.1 5.417 9.167 5.417 9.167 dbcsr_mm_accdrv_process_sort 3754 11.4 7.812 9.030 7.812 9.030 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.469 7.373 mp_irecv_dv 2335 11.1 2.453 7.302 2.453 7.302 multiply_cannon_sync_h2d 888 9.7 6.005 7.278 6.005 7.278 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.162 7.093 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.583 6.753 apply_matrix_preconditioner 6 5.3 0.005 0.022 5.168 5.414 arnoldi_extremal 4 6.8 0.000 0.000 5.082 5.102 arnoldi_normal_ev 4 7.8 0.001 0.005 5.082 5.101 calculate_norms 1584 9.8 4.395 4.793 4.395 4.793 build_subspace 16 8.4 0.014 0.021 4.768 4.775 mp_allgather_i34 111 8.7 0.882 3.965 0.882 3.965 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.462 3.773 ls_scf_post 1 4.0 0.000 0.000 3.715 3.719 dbcsr_matrix_vector_mult_local 304 10.0 3.049 3.606 3.051 3.608 ls_scf_store_result 1 5.0 0.000 0.000 3.458 3.540 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.894 3.004 dbcsr_data_new 4116 9.9 2.112 2.438 2.112 2.438 dbcsr_sort_data 325 11.1 1.864 2.116 1.864 2.116 make_images_sizes 222 9.7 0.000 0.000 0.889 2.044 mp_alltoall_i44 222 10.7 0.889 2.043 0.889 2.043 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.941 1.943 make_images_pack 222 9.7 1.702 1.937 1.706 1.940 rebuild_ks_matrix 3 7.3 0.000 0.000 1.923 1.925 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.923 1.925 dbcsr_finalize 304 7.8 0.027 0.033 1.622 1.882 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.142000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2252.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_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 3.350254E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.049 97.891 97.892 qs_energies 1 2.0 0.000 0.000 97.301 97.307 ls_scf 1 3.0 0.000 0.000 95.647 95.652 dbcsr_multiply_generic 111 6.7 0.017 0.018 79.030 79.288 ls_scf_main 1 4.0 0.000 0.000 59.102 59.103 multiply_cannon 111 7.7 0.045 0.085 52.077 56.685 density_matrix_trs4 2 5.0 0.002 0.003 52.941 53.046 multiply_cannon_loop 111 8.7 0.152 0.167 46.909 49.593 ls_scf_init_scf 1 4.0 0.000 0.000 33.297 33.299 ls_scf_init_matrix_S 1 5.0 0.000 0.001 32.062 32.131 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.434 29.444 mp_waitall_1 6369 11.0 22.736 29.206 22.736 29.206 multiply_cannon_multrec 1332 9.7 14.233 17.360 22.309 24.864 make_m2s 222 7.7 0.008 0.009 21.231 22.709 make_images 222 8.7 3.152 3.624 21.179 22.661 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.197 17.653 make_images_data 222 9.7 0.005 0.005 11.843 13.554 hybrid_alltoall_any 227 10.6 0.800 3.738 11.123 12.903 dbcsr_mm_accdrv_process 3641 10.4 0.287 0.474 7.723 9.283 dbcsr_mm_accdrv_process_sort 3641 11.4 7.283 8.784 7.283 8.784 mp_sum_l 887 5.1 4.133 7.837 4.133 7.837 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.197 6.204 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.096 6.169 mp_irecv_dv 3229 10.9 2.071 6.097 2.071 6.097 multiply_cannon_sync_h2d 1332 9.7 5.452 5.956 5.452 5.956 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.532 5.378 arnoldi_extremal 4 6.8 0.000 0.000 5.250 5.270 arnoldi_normal_ev 4 7.8 0.002 0.010 5.250 5.270 build_subspace 16 8.4 0.014 0.021 4.923 4.931 apply_matrix_preconditioner 6 5.3 0.037 0.046 4.725 4.882 mp_allgather_i34 111 8.7 2.275 4.813 2.275 4.813 calculate_norms 2376 9.8 4.185 4.522 4.185 4.522 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.615 3.917 dbcsr_matrix_vector_mult_local 304 10.0 3.229 3.703 3.231 3.705 dbcsr_sort_data 658 11.4 3.044 3.392 3.044 3.392 ls_scf_post 1 4.0 0.000 0.000 3.248 3.253 dbcsr_special_finalize 555 9.7 0.006 0.007 2.796 3.242 dbcsr_merge_single_wm 555 10.7 0.538 0.658 2.788 3.234 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.042 3.092 ls_scf_store_result 1 5.0 0.000 0.000 2.979 3.049 dbcsr_data_release 10477 10.7 1.571 2.393 1.571 2.393 make_images_pack 222 9.7 1.649 2.008 1.651 2.011 dbcsr_finalize 304 7.8 0.050 0.061 1.807 1.988 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.892000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2806.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_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 4.750430E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.059 91.783 91.785 qs_energies 1 2.0 0.000 0.000 91.019 91.023 ls_scf 1 3.0 0.000 0.000 89.062 89.066 dbcsr_multiply_generic 111 6.7 0.017 0.019 70.589 70.791 ls_scf_main 1 4.0 0.000 0.000 56.369 56.369 multiply_cannon 111 7.7 0.093 0.157 52.525 55.450 multiply_cannon_loop 111 8.7 0.088 0.095 49.922 51.361 density_matrix_trs4 2 5.0 0.002 0.003 49.281 49.364 ls_scf_init_scf 1 4.0 0.000 0.000 29.256 29.258 mp_waitall_1 5436 11.0 24.037 28.267 24.037 28.267 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.963 27.992 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.858 25.875 multiply_cannon_multrec 444 9.7 13.733 16.461 20.928 22.479 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.318 15.140 make_m2s 222 7.7 0.005 0.006 13.344 14.287 make_images 222 8.7 2.049 2.491 13.275 14.218 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.435 14.021 make_images_data 222 9.7 0.003 0.004 8.105 9.623 hybrid_alltoall_any 227 10.6 0.802 3.816 8.050 9.393 multiply_cannon_sync_h2d 444 9.7 6.710 8.053 6.710 8.053 dbcsr_mm_accdrv_process 3003 10.4 0.339 0.401 6.887 8.038 dbcsr_mm_accdrv_process_sort 3003 11.4 6.525 7.637 6.525 7.637 arnoldi_extremal 4 6.8 0.000 0.000 5.769 5.786 arnoldi_normal_ev 4 7.8 0.002 0.005 5.769 5.786 build_subspace 16 8.4 0.015 0.020 5.382 5.391 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.474 4.703 mp_sum_l 887 5.1 2.715 4.362 2.715 4.362 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.166 4.356 dbcsr_matrix_vector_mult_local 304 10.0 3.717 4.168 3.719 4.170 calculate_norms 792 9.8 3.636 3.788 3.636 3.788 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.554 3.728 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.612 3.693 mp_irecv_dv 1241 11.2 1.541 3.689 1.541 3.689 mp_allgather_i34 111 8.7 1.170 3.557 1.170 3.557 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.890 3.490 ls_scf_post 1 4.0 0.000 0.000 3.437 3.441 ls_scf_store_result 1 5.0 0.000 0.000 3.232 3.262 make_images_sizes 222 9.7 0.000 0.000 0.832 3.195 mp_alltoall_i44 222 10.7 0.831 3.195 0.831 3.195 dbcsr_finalize 304 7.8 0.062 0.077 2.206 2.318 dbcsr_data_new 4608 9.7 1.785 2.203 1.785 2.203 dbcsr_merge_all 275 8.9 0.483 0.549 2.063 2.162 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.091 2.092 rebuild_ks_matrix 3 7.3 0.000 0.000 2.059 2.061 qs_ks_build_kohn_sham_matrix 3 8.3 0.010 0.019 2.059 2.061 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 1.941 1.941 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.785000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3764.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_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 8.828600E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.077 0.099 105.810 105.811 qs_energies 1 2.0 0.000 0.000 104.223 104.228 ls_scf 1 3.0 0.000 0.000 101.267 101.272 dbcsr_multiply_generic 111 6.7 0.024 0.027 74.631 74.768 ls_scf_main 1 4.0 0.000 0.000 63.711 63.711 density_matrix_trs4 2 5.0 0.002 0.003 54.665 54.719 multiply_cannon 111 7.7 0.141 0.237 48.231 50.380 multiply_cannon_loop 111 8.7 0.098 0.100 45.253 45.848 ls_scf_init_scf 1 4.0 0.000 0.000 33.773 33.773 ls_scf_init_matrix_S 1 5.0 0.000 0.001 32.199 32.223 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.479 29.485 mp_waitall_1 4527 11.1 21.673 25.314 21.673 25.314 make_m2s 222 7.7 0.006 0.006 22.710 23.765 make_images 222 8.7 3.578 3.873 22.599 23.653 multiply_cannon_multrec 444 9.7 17.812 18.425 22.546 23.197 hybrid_alltoall_any 227 10.6 1.657 3.617 12.849 15.828 make_images_data 222 9.7 0.004 0.004 13.090 15.205 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.984 10.377 multiply_cannon_sync_h2d 444 9.7 8.795 8.875 8.795 8.875 arnoldi_extremal 4 6.8 0.000 0.000 7.416 7.425 arnoldi_normal_ev 4 7.8 0.003 0.009 7.416 7.425 build_subspace 16 8.4 0.026 0.036 6.862 6.874 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.529 5.654 dbcsr_matrix_vector_mult_local 304 10.0 5.095 5.383 5.098 5.385 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.176 5.268 apply_matrix_preconditioner 6 5.3 0.008 0.016 4.877 5.137 dbcsr_mm_accdrv_process 1814 10.4 0.221 0.344 4.546 4.687 dbcsr_mm_accdrv_process_sort 1814 11.4 4.211 4.350 4.211 4.350 ls_scf_post 1 4.0 0.000 0.000 3.784 3.789 make_images_sizes 222 9.7 0.000 0.000 1.459 3.570 mp_alltoall_i44 222 10.7 1.458 3.569 1.458 3.569 ls_scf_store_result 1 5.0 0.000 0.000 3.509 3.543 mp_allgather_i34 111 8.7 1.072 3.434 1.072 3.434 calculate_norms 792 9.8 3.229 3.271 3.229 3.271 dbcsr_finalize 304 7.8 0.082 0.090 3.077 3.177 dbcsr_merge_all 275 8.9 0.885 0.920 2.861 2.954 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.925 2.926 dbcsr_complete_redistribute 5 7.6 1.435 1.480 2.742 2.853 matrix_ls_to_qs 2 6.0 0.000 0.000 2.411 2.513 dbcsr_sort_data 325 11.1 2.441 2.506 2.441 2.506 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.475 2.477 dbcsr_new_transposed 4 7.5 0.313 0.391 2.458 2.469 rebuild_ks_matrix 3 7.3 0.000 0.000 2.409 2.411 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.002 2.409 2.411 dbcsr_data_new 6591 9.6 1.852 2.290 1.852 2.290 dbcsr_frobenius_norm 74 6.6 2.055 2.131 2.207 2.243 dbcsr_add_d 103 6.2 0.000 0.000 2.128 2.211 dbcsr_add_anytype 103 7.2 0.859 0.890 2.127 2.210 dbcsr_redistribute 4 8.5 1.388 1.442 2.118 2.199 dbcsr_data_release 12724 10.6 1.965 2.154 1.965 2.154 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.811000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6998.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b8a11310c43cf35ba551cf3896354d85cdb3a872_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 592.719872E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 1019280897. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.056 0.145 241.611 241.612 qs_mol_dyn_low 1 2.0 0.004 0.023 240.553 240.593 qs_forces 5 3.8 0.005 0.035 240.425 240.446 qs_energies 5 4.8 0.002 0.080 237.054 237.089 scf_env_do_scf 5 5.8 0.001 0.028 219.414 219.417 scf_env_do_scf_inner_loop 105 6.6 0.004 0.080 190.782 190.783 qs_scf_new_mos 105 7.6 0.000 0.001 146.734 146.890 qs_scf_loop_do_ot 105 8.6 0.001 0.007 146.733 146.890 dbcsr_multiply_generic 1445 12.2 0.126 0.133 137.238 137.674 ot_scf_mini 105 9.6 0.003 0.014 136.684 136.839 multiply_cannon 1445 13.2 0.275 0.287 117.036 118.958 multiply_cannon_loop 1445 14.2 2.860 3.023 115.267 116.820 velocity_verlet 4 3.0 0.003 0.021 109.789 109.790 ot_mini 105 10.6 0.002 0.051 61.026 61.158 qs_ot_get_p 112 10.4 0.001 0.001 45.243 45.604 multiply_cannon_multrec 69360 15.2 29.949 34.825 39.877 45.296 mp_waitall_1 488190 16.1 35.462 43.457 35.462 43.457 qs_ot_get_derivative 55 11.6 0.001 0.001 39.211 39.345 qs_ot_p2m_diag 40 11.0 0.021 0.046 34.209 34.343 multiply_cannon_metrocomm3 69360 15.2 0.207 0.220 25.792 33.871 multiply_cannon_sync_h2d 69360 15.2 28.888 33.614 28.888 33.614 rebuild_ks_matrix 110 8.4 0.000 0.000 30.795 30.984 qs_ks_build_kohn_sham_matrix 110 9.4 0.013 0.123 30.795 30.983 cp_dbcsr_syevd 40 12.0 0.002 0.004 30.966 30.967 init_scf_loop 7 6.6 0.001 0.076 28.590 28.591 qs_ks_update_qs_env 112 7.6 0.001 0.001 28.148 28.317 cp_fm_syevd 40 13.0 0.000 0.002 25.608 25.760 prepare_preconditioner 7 7.6 0.000 0.000 23.411 23.439 make_preconditioner 7 8.6 0.000 0.023 23.411 23.439 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.050 23.287 apply_single 62 13.6 0.000 0.000 23.050 23.287 ot_new_cg_direction 55 11.6 0.001 0.025 21.099 21.101 cp_fm_redistribute_end 40 14.0 10.388 20.725 10.395 20.728 cp_fm_syevd_base 40 14.0 10.323 20.659 10.323 20.659 qs_rho_update_rho_low 110 7.6 0.001 0.008 19.372 19.721 calculate_rho_elec 110 8.6 0.031 0.041 19.371 19.721 make_full_inverse_cholesky 7 9.6 0.000 0.001 15.634 15.717 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.042 15.267 init_scf_run 5 5.8 0.000 0.002 14.928 14.929 scf_env_initial_rho_setup 5 6.8 0.003 0.068 14.927 14.928 density_rs2pw 110 9.6 0.006 0.008 13.033 14.393 rs_pw_transfer 690 11.5 0.011 0.012 12.915 14.250 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.035 14.125 pw_transfer 1645 12.4 0.082 0.098 13.690 13.891 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 13.548 13.753 mp_sum_l 4764 12.2 12.373 13.190 12.373 13.190 calculate_dm_sparse 110 9.5 0.000 0.001 11.745 11.921 fft_wrap_pw1pw2_240 915 15.0 1.188 1.268 11.597 11.771 qs_vxc_create 110 10.4 0.002 0.004 11.294 11.332 cp_fm_cholesky_invert 7 10.6 10.878 10.886 10.878 10.886 dbcsr_mm_accdrv_process 154766 15.8 6.176 6.388 9.794 10.703 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.606 10.696 acc_transpose_blocks 69360 15.2 0.355 0.371 9.973 10.489 check_diag 80 13.5 8.596 8.894 9.822 9.975 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.025 9.917 9.936 fft3d_pb 915 16.0 2.365 2.555 9.675 9.854 calculate_first_density_matrix 1 7.0 0.000 0.004 9.768 9.784 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 9.441 9.505 sum_up_and_integrate 60 10.3 0.028 0.032 9.300 9.312 integrate_v_rspace 60 11.3 0.002 0.020 9.272 9.283 xc_rho_set_and_dset_create 110 12.4 0.077 0.097 7.807 8.070 make_m2s 2890 13.2 0.082 0.091 7.369 7.953 multiply_cannon_metrocomm1 69360 15.2 0.097 0.104 4.774 7.904 make_images 2890 14.2 0.240 0.263 7.262 7.842 xc_vxc_pw_create 60 11.3 0.039 0.049 7.571 7.608 acc_transpose_blocks_kernels 69360 16.2 0.847 0.894 7.207 7.589 make_full_single_inverse 7 9.6 0.002 0.071 7.414 7.443 mp_alltoall_z22v 2340 17.7 6.789 7.118 6.789 7.118 xc_pw_derive 510 13.4 0.005 0.006 6.998 7.071 jit_kernel_transpose 5 15.0 6.360 6.717 6.360 6.717 mp_waitany 7680 13.5 4.461 5.776 4.461 5.776 potential_pw2rs 60 12.3 0.003 0.003 5.343 5.375 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=241.612000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.400000, yerr=3.382307 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: b8a11310c43cf35ba551cf3896354d85cdb3a872 Summary: empty Status: OK