=== This is the CP2K Performance-Test === Updating a647768d8..ad1d17f03 Fast-forward tools/dashboard/dashboard.conf | 10 ++++++++++ tools/docker/cp2k-ci.conf | 33 +++++++++++++++++++++++++++++++++ 2 files changed, 43 insertions(+) Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: ad1d17f038befc10386af0239c8c9a26c496fd9e ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.12.0, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.8.2, 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: 28.08.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.12.0 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/01 job id: 48654683 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/02 job id: 48654684 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/03 job id: 48654685 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/04 job id: 48654686 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/05 job id: 48654687 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/06 job id: 48654688 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/07 job id: 48654689 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/08 job id: 48654690 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/09 job id: 48654691 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/10 job id: 48654692 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/11 job id: 48654693 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/12 job id: 48654694 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/13 job id: 48654695 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/14 job id: 48654696 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/15 job id: 48654698 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/16 job id: 48654700 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/17 job id: 48654704 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/18 job id: 48654708 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/19 job id: 48654710 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/20 job id: 48654713 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/21 job id: 48654714 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/22 job id: 48654715 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/23 job id: 48654716 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/24 job id: 48654717 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/25 job id: 48654718 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/26 job id: 48654719 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/27 job id: 48654721 --- 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 0.029 0.048 133.187 133.187 farming_run 1 2.0 132.555 132.557 133.140 133.143 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.488728E+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 0.023 114.308 114.309 qs_energies 1 2.0 0.000 0.000 114.122 114.124 mp2_main 1 3.0 0.000 0.000 112.036 112.038 mp2_gpw_main 1 4.0 0.028 0.034 111.018 111.020 mp2_ri_gpw_compute_in 1 5.0 0.186 0.205 92.019 92.402 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.645 55.030 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.165 41.031 46.134 get_2c_integrals 1 6.0 0.008 0.009 36.456 37.187 integrate_v_rspace 273 8.0 0.436 0.453 24.590 29.465 pw_transfer 6555 10.6 0.374 0.389 26.710 27.272 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 25.338 25.985 grid_integrate_task_list 273 9.0 20.505 25.857 20.505 25.857 fft_wrap_pw1pw2_100 2178 12.4 1.098 1.293 22.891 23.524 rpa_ri_compute_en 1 5.0 0.020 0.021 18.882 18.971 compute_2c_integrals 1 7.0 0.003 0.003 18.916 18.916 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.517 18.616 mp2_eri_2c_integrate_gpw 1 9.0 2.393 2.436 18.514 18.614 cp_fm_cholesky_decompose 12 8.2 17.509 18.246 17.509 18.246 cholesky_decomp 1 7.0 0.000 0.000 16.383 17.109 fft3d_s 5443 13.4 16.023 16.491 16.044 16.514 ao_to_mo_and_store_B_mult_1 272 7.0 10.781 15.323 10.781 15.323 calculate_wavefunction 272 8.0 5.428 5.577 12.343 13.037 rpa_num_int 1 6.0 0.000 0.003 10.675 10.676 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.642 10.657 calc_mat_Q 8 8.0 0.000 0.000 9.476 9.584 contract_S_to_Q 8 9.0 0.000 0.000 8.896 9.004 calc_potential_gpw 544 9.5 0.005 0.005 8.237 8.678 parallel_gemm_fm 14 9.1 0.000 0.000 8.491 8.607 parallel_gemm_fm_cosma 14 10.1 8.491 8.607 8.491 8.607 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.028 8.257 potential_pw2rs 545 10.0 0.107 0.109 7.434 8.068 create_integ_mat 1 6.0 0.022 0.027 7.730 7.730 collocate_single_gaussian 272 10.0 0.040 0.042 7.292 7.506 array2fm 1 7.0 0.000 0.000 6.731 7.157 pw_scatter_s 2720 13.7 4.364 4.502 4.364 4.502 pw_gather_s 2722 13.2 3.459 3.852 3.459 3.852 array2fm_buffer_send 1 8.0 2.973 3.177 2.973 3.177 pw_poisson_solve 545 10.5 1.121 1.191 2.200 2.352 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.017550, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2807.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 0.051 0.057 422.628 422.629 farming_run 1 2.0 421.469 421.507 422.565 422.590 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.235403E+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 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 0.013 0.033 217.985 217.986 qs_energies 1 2.0 0.002 0.015 217.730 217.741 scf_env_do_scf 1 3.0 0.000 0.000 115.536 115.536 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.612 114.619 rebuild_ks_matrix 4 6.0 0.000 0.000 114.611 114.618 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 114.611 114.618 hfx_ks_matrix 4 8.0 0.001 0.001 114.259 114.263 integrate_four_center 4 9.0 0.153 0.469 114.258 114.262 integrate_four_center_main 4 10.0 0.129 0.605 102.098 106.114 integrate_four_center_bin 267 11.0 101.969 106.001 101.969 106.001 mp2_main 1 3.0 0.001 0.016 101.893 101.906 mp2_gpw_main 1 4.0 0.056 0.097 100.961 100.973 init_scf_loop 1 4.0 0.000 0.000 97.755 97.755 mp2_ri_gpw_compute_in 1 5.0 0.075 0.143 73.975 75.018 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 53.742 54.788 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.162 41.432 46.488 integrate_v_rspace 95 8.0 0.396 0.560 27.912 32.837 pw_transfer 2240 10.6 0.146 0.170 29.464 29.909 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.478 28.931 mp2_ri_gpw_compute_en 1 5.0 0.060 0.073 26.738 28.742 grid_integrate_task_list 95 9.0 23.330 28.450 23.330 28.450 ao_to_mo_and_store_B_mult_1 91 7.0 10.625 27.271 10.625 27.271 fft_wrap_pw1pw2_100 730 12.4 1.265 1.460 26.231 26.657 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.829 1.894 25.031 25.042 get_2c_integrals 1 6.0 0.000 0.000 20.112 20.150 compute_2c_integrals 1 7.0 0.004 0.011 19.093 19.105 compute_2c_integrals_loop_lm 1 8.0 0.009 0.032 18.587 18.928 fft3d_s 1823 13.4 18.484 18.903 18.497 18.916 mp2_eri_2c_integrate_gpw 1 9.0 1.737 1.841 18.578 18.896 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.780 17.780 calculate_wavefunction 91 8.0 2.014 2.042 9.616 9.835 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.590 8.802 9.612 local_gemm 172 8.0 8.244 9.038 8.244 9.038 potential_pw2rs 186 10.0 0.033 0.034 8.376 8.981 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.125 8.528 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.526 8.017 8.521 calc_potential_gpw 182 9.5 0.002 0.002 7.920 8.124 mp_sync 37 10.5 4.127 8.077 4.127 8.077 collocate_single_gaussian 91 10.0 0.017 0.032 7.732 7.961 integrate_four_center_load 4 10.0 0.000 0.000 6.790 6.794 hfx_load_balance 1 11.0 0.000 0.000 6.790 6.794 mp_sendrecv_dm3 2068 8.0 6.050 6.541 6.050 6.541 mp2_ri_gpw_compute_en_ener 172 7.0 6.339 6.410 6.339 6.410 pw_gather_s 912 13.2 4.452 5.014 4.452 5.014 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.962911, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1496.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 451.805184E+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 592243. 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 0.028 0.034 55.146 55.147 qs_mol_dyn_low 1 2.0 0.003 0.012 54.738 54.746 qs_forces 11 3.9 0.002 0.002 54.641 54.641 qs_energies 11 4.9 0.002 0.002 53.162 53.183 scf_env_do_scf 11 5.9 0.000 0.001 46.300 46.300 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.202 44.202 dbcsr_multiply_generic 2286 12.5 0.093 0.097 34.348 34.788 qs_scf_new_mos 108 7.5 0.000 0.001 34.063 34.343 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.062 34.343 ot_scf_mini 108 9.5 0.002 0.002 32.338 32.502 multiply_cannon 2286 13.5 0.188 0.203 26.762 28.101 multiply_cannon_loop 2286 14.5 1.827 1.931 26.075 27.440 velocity_verlet 10 3.0 0.016 0.042 26.000 26.000 ot_mini 108 10.5 0.001 0.001 19.357 19.602 qs_ot_get_derivative 108 11.5 0.001 0.001 16.358 16.536 mp_waitall_1 245248 16.5 8.670 14.869 8.670 14.869 multiply_cannon_metrocomm3 54864 15.5 0.074 0.080 6.159 13.159 multiply_cannon_multrec 54864 15.5 3.671 5.774 7.917 11.536 qs_ot_get_p 119 10.4 0.001 0.001 8.155 8.474 rebuild_ks_matrix 119 8.3 0.000 0.000 7.913 8.043 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.913 8.043 mp_sum_l 7287 12.8 5.464 7.120 5.464 7.120 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.982 7.106 multiply_cannon_sync_h2d 54864 15.5 5.172 6.551 5.172 6.551 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.605 6.062 dbcsr_mm_accdrv_process 76910 16.1 1.826 2.880 4.160 5.943 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.547 5.653 init_scf_run 11 5.9 0.000 0.001 5.578 5.578 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.578 5.578 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.356 5.429 sum_up_and_integrate 119 10.3 0.001 0.001 4.509 4.516 integrate_v_rspace 119 11.3 0.002 0.003 4.498 4.506 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.503 4.504 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.207 4.328 calculate_rho_elec 119 8.7 0.011 0.017 4.207 4.327 cp_fm_diag_elpa 50 13.0 0.000 0.001 4.296 4.296 cp_fm_redistribute_end 50 14.0 2.190 4.267 2.196 4.270 cp_fm_diag_elpa_base 50 14.0 2.068 4.156 2.073 4.167 calculate_first_density_matrix 1 7.0 0.000 0.000 3.346 3.352 calculate_dm_sparse 119 9.5 0.000 0.000 3.056 3.198 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.947 3.166 apply_single 119 13.6 0.000 0.000 2.947 3.166 jit_kernel_multiply 13 15.8 2.270 3.102 2.270 3.102 acc_transpose_blocks 54864 15.5 0.231 0.258 2.308 2.875 multiply_cannon_metrocomm1 54864 15.5 0.057 0.063 1.778 2.799 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.784 2.786 ot_diis_step 108 11.5 0.006 0.006 2.743 2.744 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.588 2.676 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.511 2.563 density_rs2pw 119 9.7 0.004 0.004 2.206 2.332 wfi_extrapolate 11 7.9 0.001 0.001 2.159 2.159 grid_integrate_task_list 119 12.3 2.005 2.102 2.005 2.102 init_scf_loop 11 6.9 0.000 0.000 2.075 2.076 potential_pw2rs 119 12.3 0.004 0.004 1.874 1.883 mp_sum_d 4135 12.0 1.316 1.860 1.316 1.860 make_m2s 4572 13.5 0.053 0.056 1.810 1.854 pw_transfer 1439 11.6 0.052 0.058 1.702 1.773 make_images 4572 14.5 0.133 0.138 1.727 1.770 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.627 1.698 mp_alltoall_d11v 2130 13.8 1.372 1.591 1.372 1.591 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.501 1.542 transfer_rs2pw 487 10.6 0.005 0.006 1.406 1.497 acc_transpose_blocks_sync 164592 16.5 1.202 1.442 1.202 1.442 fft3d_ps 1201 14.6 0.370 0.480 1.342 1.412 mp_waitany 12084 13.8 1.260 1.410 1.260 1.410 grid_collocate_task_list 119 9.7 1.348 1.409 1.348 1.409 transfer_pw2rs 487 13.2 0.006 0.007 1.366 1.374 fft_wrap_pw1pw2_140 487 13.2 0.140 0.155 1.256 1.329 parallel_gemm_fm 81 9.0 0.000 0.000 1.211 1.216 parallel_gemm_fm_cosma 81 10.0 1.211 1.216 1.211 1.216 acc_transpose_blocks_kernels 54864 16.5 0.250 0.385 0.828 1.126 dbcsr_dot_sd 1205 11.9 0.048 0.060 0.724 1.121 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.147000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.545455, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 487.436288E+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 1031941. 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 0.022 0.035 39.581 39.582 qs_mol_dyn_low 1 2.0 0.003 0.003 39.379 39.390 qs_forces 11 3.9 0.002 0.003 39.322 39.323 qs_energies 11 4.9 0.002 0.002 37.613 37.617 scf_env_do_scf 11 5.9 0.000 0.001 32.480 32.480 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.939 29.939 dbcsr_multiply_generic 2286 12.5 0.100 0.103 22.255 22.607 qs_scf_new_mos 108 7.5 0.001 0.001 21.105 21.337 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.105 21.337 ot_scf_mini 108 9.5 0.002 0.003 20.167 20.336 velocity_verlet 10 3.0 0.001 0.002 18.802 18.803 multiply_cannon 2286 13.5 0.209 0.216 16.984 18.623 multiply_cannon_loop 2286 14.5 1.193 1.253 15.765 17.326 ot_mini 108 10.5 0.001 0.001 12.345 12.579 mp_waitall_1 200699 16.5 5.731 10.834 5.731 10.834 qs_ot_get_derivative 108 11.5 0.001 0.001 9.946 10.115 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.093 9.458 multiply_cannon_multrec 27432 15.5 1.809 4.218 6.136 8.953 rebuild_ks_matrix 119 8.3 0.000 0.000 6.927 7.072 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.927 7.071 dbcsr_mm_accdrv_process 47894 16.0 3.563 5.928 4.246 6.330 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.079 6.212 qs_ot_get_p 119 10.4 0.001 0.001 4.871 5.092 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.767 4.614 mp_sum_l 7287 12.8 2.098 4.143 2.098 4.143 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.019 4.081 apply_single 119 13.6 0.000 0.000 3.019 4.081 init_scf_run 11 5.9 0.000 0.001 3.917 3.917 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.916 3.917 sum_up_and_integrate 119 10.3 0.001 0.001 3.815 3.821 integrate_v_rspace 119 11.3 0.002 0.003 3.801 3.806 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.720 3.753 calculate_rho_elec 119 8.7 0.021 0.024 3.719 3.753 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.140 3.158 make_m2s 4572 13.5 0.052 0.054 2.726 3.034 make_images 4572 14.5 0.206 0.244 2.638 2.949 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.737 2.738 init_scf_loop 11 6.9 0.000 0.000 2.517 2.518 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.382 2.467 cp_fm_diag_elpa 50 13.0 0.000 0.001 2.391 2.391 calculate_first_density_matrix 1 7.0 0.000 0.000 2.372 2.374 cp_fm_redistribute_end 50 14.0 1.207 2.357 1.210 2.360 ot_diis_step 108 11.5 0.011 0.011 2.344 2.345 cp_fm_diag_elpa_base 50 14.0 1.116 2.262 1.145 2.302 multiply_cannon_sync_h2d 27432 15.5 1.678 2.238 1.678 2.238 calculate_dm_sparse 119 9.5 0.000 0.001 2.084 2.161 density_rs2pw 119 9.7 0.004 0.004 2.018 2.080 acc_transpose_blocks 27432 15.5 0.114 0.119 1.618 1.951 pw_transfer 1439 11.6 0.065 0.069 1.912 1.947 grid_integrate_task_list 119 12.3 1.833 1.934 1.833 1.934 jit_kernel_multiply 10 16.4 0.625 1.927 0.625 1.927 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.907 1.909 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.821 1.859 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.685 1.725 make_images_data 4572 15.5 0.047 0.054 1.278 1.641 prepare_preconditioner 11 7.9 0.000 0.000 1.607 1.633 make_preconditioner 11 8.9 0.000 0.000 1.607 1.633 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.508 1.566 potential_pw2rs 119 12.3 0.006 0.006 1.550 1.561 fft3d_ps 1201 14.6 0.518 0.573 1.457 1.489 wfi_extrapolate 11 7.9 0.001 0.001 1.485 1.485 hybrid_alltoall_any 4725 16.4 0.054 0.115 1.136 1.479 mp_alltoall_d11v 2130 13.8 1.258 1.462 1.258 1.462 fft_wrap_pw1pw2_140 487 13.2 0.160 0.170 1.415 1.450 grid_collocate_task_list 119 9.7 1.286 1.330 1.286 1.330 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.219 1.264 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.255 1.264 transfer_rs2pw 487 10.6 0.005 0.006 1.157 1.235 mp_sum_d 4135 12.0 0.579 1.051 0.579 1.051 mp_allgather_i34 2286 14.5 0.640 1.046 0.640 1.046 transfer_pw2rs 487 13.2 0.004 0.005 0.987 0.995 qs_energies_init_hamiltonians 11 5.9 0.002 0.018 0.951 0.951 acc_transpose_blocks_sync 82296 16.5 0.816 0.941 0.816 0.941 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.907 0.921 acc_transpose_blocks_kernels 27432 16.5 0.190 0.278 0.661 0.871 make_images_sizes 4572 15.5 0.005 0.005 0.579 0.829 mp_alltoall_i44 4572 16.5 0.575 0.825 0.575 0.825 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.582000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.181818, yerr=1.585054 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 520.900608E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.125 0.254 38.402 38.438 qs_mol_dyn_low 1 2.0 0.005 0.016 37.581 37.589 qs_forces 11 3.9 0.002 0.003 37.454 37.466 qs_energies 11 4.9 0.006 0.016 35.669 35.682 scf_env_do_scf 11 5.9 0.007 0.028 29.982 29.982 scf_env_do_scf_inner_loop 108 6.5 0.005 0.026 26.799 26.800 dbcsr_multiply_generic 2286 12.5 0.117 0.229 18.442 18.555 velocity_verlet 10 3.0 0.001 0.002 17.892 17.893 qs_scf_new_mos 108 7.5 0.001 0.001 17.355 17.373 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.355 17.372 ot_scf_mini 108 9.5 0.002 0.002 16.534 16.544 multiply_cannon 2286 13.5 0.194 0.201 13.994 14.815 multiply_cannon_loop 2286 14.5 0.866 0.912 13.074 13.951 ot_mini 108 10.5 0.001 0.001 9.942 9.952 qs_ot_get_derivative 108 11.5 0.001 0.001 8.401 8.411 multiply_cannon_multrec 18288 15.5 1.890 2.817 6.973 7.197 rebuild_ks_matrix 119 8.3 0.000 0.000 7.171 7.185 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.171 7.185 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.286 6.303 dbcsr_mm_accdrv_process 38222 16.0 4.932 5.722 4.990 5.787 mp_waitall_1 158411 16.6 3.876 4.672 3.876 4.672 sum_up_and_integrate 119 10.3 0.001 0.001 4.592 4.605 integrate_v_rspace 119 11.3 0.002 0.003 4.579 4.594 init_scf_run 11 5.9 0.000 0.001 4.276 4.277 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.276 4.277 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.248 4.256 calculate_rho_elec 119 8.7 0.031 0.031 4.247 4.256 qs_ot_get_p 119 10.4 0.001 0.001 4.209 4.232 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.124 3.727 init_scf_loop 11 6.9 0.006 0.028 3.153 3.161 make_m2s 4572 13.5 0.044 0.046 2.781 2.996 make_images 4572 14.5 0.210 0.303 2.696 2.909 density_rs2pw 119 9.7 0.004 0.004 2.750 2.846 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.800 2.809 calculate_first_density_matrix 1 7.0 0.002 0.006 2.753 2.756 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.444 2.446 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.056 2.381 apply_single 119 13.6 0.000 0.000 2.055 2.380 potential_pw2rs 119 12.3 0.007 0.009 2.322 2.360 pw_transfer 1439 11.6 0.065 0.068 2.243 2.285 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.474 2.243 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.194 2.199 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.150 2.193 prepare_preconditioner 11 7.9 0.000 0.000 2.173 2.176 make_preconditioner 11 8.9 0.000 0.001 2.173 2.176 cp_fm_diag_elpa 50 13.0 0.000 0.001 2.089 2.097 cp_fm_diag_elpa_base 50 14.0 2.060 2.074 2.087 2.095 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.991 2.094 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 2.024 2.031 calculate_dm_sparse 119 9.5 0.000 0.000 1.917 1.927 transfer_rs2pw 487 10.6 0.005 0.005 1.785 1.910 grid_integrate_task_list 119 12.3 1.786 1.896 1.786 1.896 mp_sum_l 7287 12.8 1.409 1.822 1.409 1.822 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 1.793 1.816 acc_transpose_blocks 18288 15.5 0.080 0.081 1.670 1.793 fft3d_ps 1201 14.6 0.531 0.549 1.720 1.764 fft_wrap_pw1pw2_140 487 13.2 0.210 0.217 1.603 1.645 transfer_pw2rs 487 13.2 0.004 0.004 1.580 1.610 ot_diis_step 108 11.5 0.011 0.011 1.520 1.521 make_images_data 4572 15.5 0.049 0.070 1.133 1.472 wfi_extrapolate 11 7.9 0.001 0.001 1.464 1.464 hybrid_alltoall_any 4725 16.4 0.060 0.117 0.930 1.319 grid_collocate_task_list 119 9.7 1.233 1.319 1.233 1.319 make_images_sizes 4572 15.5 0.005 0.005 0.876 1.242 mp_alltoall_i44 4572 16.5 0.871 1.238 0.871 1.238 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.207 1.212 mp_alltoall_d11v 2130 13.8 0.969 1.136 0.969 1.136 multiply_cannon_sync_h2d 18288 15.5 1.011 1.128 1.011 1.128 mp_alltoall_z22v 1201 16.6 1.006 1.101 1.006 1.101 qs_energies_init_hamiltonians 11 5.9 0.003 0.008 1.039 1.051 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.022 1.038 cp_fm_cholesky_invert 11 10.9 1.031 1.036 1.031 1.036 mp_sendrecv_dv 11067 12.7 0.972 0.998 0.972 0.998 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.973 0.975 dbcsr_complete_redistribute 329 12.2 0.156 0.274 0.808 0.943 acc_transpose_blocks_sync 54864 16.5 0.752 0.866 0.752 0.866 acc_transpose_blocks_kernels 18288 16.5 0.219 0.227 0.817 0.835 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=38.438000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.636364, yerr=0.642824 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 555.216896E+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 0.070 0.183 42.165 42.166 qs_mol_dyn_low 1 2.0 0.003 0.004 41.602 41.610 qs_forces 11 3.9 0.002 0.002 41.543 41.544 qs_energies 11 4.9 0.003 0.006 39.725 39.735 scf_env_do_scf 11 5.9 0.001 0.001 34.054 34.055 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 29.943 29.943 dbcsr_multiply_generic 2286 12.5 0.100 0.104 21.504 21.733 velocity_verlet 10 3.0 0.002 0.003 21.196 21.198 qs_scf_new_mos 108 7.5 0.001 0.001 20.295 20.357 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.295 20.356 ot_scf_mini 108 9.5 0.002 0.003 19.210 19.262 multiply_cannon 2286 13.5 0.219 0.226 16.706 17.252 multiply_cannon_loop 2286 14.5 1.566 1.631 15.585 15.978 ot_mini 108 10.5 0.001 0.001 11.620 11.682 qs_ot_get_derivative 108 11.5 0.001 0.001 9.670 9.722 multiply_cannon_multrec 27432 15.5 2.461 3.117 9.015 9.275 dbcsr_mm_accdrv_process 47916 15.9 6.020 7.658 6.451 7.778 rebuild_ks_matrix 119 8.3 0.000 0.000 7.243 7.295 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.242 7.294 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.434 6.477 qs_ot_get_p 119 10.4 0.001 0.001 4.540 4.617 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.271 4.312 calculate_rho_elec 119 8.7 0.040 0.046 4.270 4.311 sum_up_and_integrate 119 10.3 0.001 0.001 4.294 4.303 integrate_v_rspace 119 11.3 0.003 0.003 4.283 4.291 init_scf_run 11 5.9 0.000 0.001 4.136 4.136 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.135 4.136 init_scf_loop 11 6.9 0.001 0.005 4.079 4.080 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.474 3.872 mp_waitall_1 137007 16.6 3.123 3.580 3.123 3.580 make_m2s 4572 13.5 0.054 0.056 3.173 3.300 make_images 4572 14.5 0.276 0.336 3.065 3.191 prepare_preconditioner 11 7.9 0.000 0.000 3.147 3.154 make_preconditioner 11 8.9 0.000 0.001 3.147 3.154 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.728 3.075 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.752 2.778 density_rs2pw 119 9.7 0.004 0.004 2.682 2.770 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.752 2.765 acc_transpose_blocks 27432 15.5 0.120 0.124 2.475 2.554 calculate_first_density_matrix 1 7.0 0.001 0.003 2.416 2.418 pw_transfer 1439 11.6 0.065 0.069 2.322 2.362 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.168 2.295 apply_single 119 13.6 0.000 0.000 2.167 2.295 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.284 2.285 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.229 2.272 calculate_dm_sparse 119 9.5 0.000 0.000 2.202 2.264 potential_pw2rs 119 12.3 0.009 0.009 2.041 2.055 cp_fm_diag_elpa 50 13.0 0.001 0.002 1.940 1.948 cp_fm_diag_elpa_base 50 14.0 1.896 1.914 1.936 1.945 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.930 1.932 grid_integrate_task_list 119 12.3 1.830 1.904 1.830 1.904 ot_diis_step 108 11.5 0.012 0.012 1.902 1.902 transfer_rs2pw 487 10.6 0.004 0.005 1.670 1.791 fft3d_ps 1201 14.6 0.558 0.615 1.748 1.783 fft_wrap_pw1pw2_140 487 13.2 0.245 0.258 1.736 1.780 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.682 1.693 wfi_extrapolate 11 7.9 0.001 0.001 1.653 1.653 acc_transpose_blocks_sync 82296 16.5 1.496 1.578 1.496 1.578 make_images_data 4572 15.5 0.048 0.051 1.280 1.530 mp_sum_l 7287 12.8 1.240 1.490 1.240 1.490 multiply_cannon_metrocomm3 27432 15.5 0.041 0.042 0.916 1.422 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.402 1.419 hybrid_alltoall_any 4725 16.4 0.066 0.155 1.047 1.382 transfer_pw2rs 487 13.2 0.004 0.004 1.346 1.360 grid_collocate_task_list 119 9.7 1.250 1.354 1.250 1.354 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.265 1.278 dbcsr_complete_redistribute 329 12.2 0.122 0.164 0.958 1.263 cp_fm_upper_to_full 72 14.2 0.849 1.220 0.849 1.220 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.213 1.215 mp_alltoall_d11v 2130 13.8 1.001 1.151 1.001 1.151 mp_alltoall_z22v 1201 16.6 1.072 1.107 1.072 1.107 cp_fm_cholesky_invert 11 10.9 1.094 1.101 1.094 1.101 jit_kernel_multiply 6 16.2 0.365 1.088 0.365 1.088 make_images_sizes 4572 15.5 0.005 0.005 0.745 1.024 mp_alltoall_i44 4572 16.5 0.740 1.020 0.740 1.020 copy_fm_to_dbcsr 176 11.2 0.001 0.002 0.674 0.973 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.901 0.908 mp_sendrecv_dv 8211 12.7 0.886 0.908 0.886 0.908 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.799 0.880 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=42.166000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.727273, yerr=2.525899 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 625.537024E+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 0.111 0.155 35.579 35.595 qs_mol_dyn_low 1 2.0 0.005 0.032 34.931 34.956 qs_forces 11 3.9 0.015 0.018 34.331 34.335 qs_energies 11 4.9 0.002 0.002 32.487 32.494 scf_env_do_scf 11 5.9 0.000 0.001 27.205 27.205 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 23.943 23.944 velocity_verlet 10 3.0 0.002 0.002 18.329 18.366 dbcsr_multiply_generic 2286 12.5 0.093 0.097 15.384 15.545 qs_scf_new_mos 108 7.5 0.001 0.001 14.779 14.821 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.778 14.821 ot_scf_mini 108 9.5 0.002 0.002 13.937 13.974 multiply_cannon 2286 13.5 0.226 0.230 11.569 12.308 multiply_cannon_loop 2286 14.5 0.645 0.662 10.332 10.640 ot_mini 108 10.5 0.001 0.001 7.966 8.016 rebuild_ks_matrix 119 8.3 0.000 0.000 6.670 6.738 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.670 6.738 qs_ot_get_derivative 108 11.5 0.001 0.001 6.481 6.517 multiply_cannon_multrec 9144 15.5 1.740 2.037 6.013 6.213 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.938 5.998 dbcsr_mm_accdrv_process 12550 15.8 3.306 4.195 4.162 4.238 mp_waitall_1 115863 16.7 2.977 4.211 2.977 4.211 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.193 4.208 calculate_rho_elec 119 8.7 0.060 0.061 4.192 4.208 sum_up_and_integrate 119 10.3 0.001 0.001 4.127 4.133 integrate_v_rspace 119 11.3 0.003 0.003 4.117 4.122 qs_ot_get_p 119 10.4 0.001 0.001 3.737 3.790 init_scf_run 11 5.9 0.000 0.001 3.647 3.647 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.647 3.647 init_scf_loop 11 6.9 0.000 0.000 3.231 3.232 make_m2s 4572 13.5 0.034 0.035 2.614 2.906 make_images 4572 14.5 0.271 0.302 2.525 2.817 density_rs2pw 119 9.7 0.003 0.004 2.466 2.535 pw_transfer 1439 11.6 0.065 0.068 2.390 2.405 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.396 2.402 prepare_preconditioner 11 7.9 0.000 0.000 2.387 2.394 make_preconditioner 11 8.9 0.000 0.000 2.387 2.394 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.297 2.314 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.220 2.279 calculate_first_density_matrix 1 7.0 0.000 0.000 2.195 2.197 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.154 2.181 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.100 2.102 grid_integrate_task_list 119 12.3 1.926 1.970 1.926 1.970 calculate_dm_sparse 119 9.5 0.000 0.000 1.905 1.938 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.790 1.809 potential_pw2rs 119 12.3 0.010 0.011 1.795 1.801 fft_wrap_pw1pw2_140 487 13.2 0.322 0.332 1.773 1.789 cp_fm_diag_elpa 50 13.0 0.001 0.002 1.745 1.753 cp_fm_diag_elpa_base 50 14.0 1.712 1.734 1.742 1.750 jit_kernel_multiply 10 15.3 0.817 1.742 0.817 1.742 fft3d_ps 1201 14.6 0.567 0.580 1.711 1.724 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.647 1.650 make_images_data 4572 15.5 0.042 0.046 1.208 1.620 hybrid_alltoall_any 4725 16.4 0.065 0.175 1.130 1.549 acc_transpose_blocks 9144 15.5 0.042 0.043 1.433 1.462 ot_diis_step 108 11.5 0.013 0.013 1.444 1.444 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.410 1.429 transfer_rs2pw 487 10.6 0.004 0.004 1.343 1.420 grid_collocate_task_list 119 9.7 1.346 1.407 1.346 1.407 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.355 1.406 apply_single 119 13.6 0.000 0.000 1.355 1.406 wfi_extrapolate 11 7.9 0.001 0.001 1.385 1.385 qs_energies_init_hamiltonians 11 5.9 0.030 0.051 1.329 1.335 cp_fm_cholesky_invert 11 10.9 1.276 1.280 1.276 1.280 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.581 1.247 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.190 1.197 mp_alltoall_d11v 2130 13.8 1.097 1.182 1.097 1.182 mp_allgather_i34 2286 14.5 0.546 1.176 0.546 1.176 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.809 1.140 mp_sum_l 7287 12.8 0.781 1.126 0.781 1.126 transfer_pw2rs 487 13.2 0.003 0.004 1.092 1.100 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.050 1.071 mp_alltoall_z22v 1201 16.6 1.016 1.054 1.016 1.054 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.931 make_images_sizes 4572 15.5 0.005 0.005 0.620 0.930 mp_alltoall_i44 4572 16.5 0.615 0.925 0.615 0.925 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.856 0.864 qs_env_update_s_mstruct 11 6.9 0.001 0.002 0.727 0.775 dbcsr_complete_redistribute 329 12.2 0.159 0.177 0.711 0.750 acc_transpose_blocks_sync 27432 16.5 0.721 0.750 0.721 0.750 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=35.595000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=591.181818, yerr=7.661215 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 777.195520E+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 0.017 0.038 47.997 47.998 qs_mol_dyn_low 1 2.0 0.003 0.003 47.740 47.761 qs_forces 11 3.9 0.002 0.002 47.666 47.666 qs_energies 11 4.9 0.002 0.002 45.551 45.557 scf_env_do_scf 11 5.9 0.001 0.001 39.513 39.514 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 30.769 30.770 velocity_verlet 10 3.0 0.020 0.021 27.198 27.203 dbcsr_multiply_generic 2286 12.5 0.102 0.104 20.963 21.411 qs_scf_new_mos 108 7.5 0.001 0.001 19.859 19.964 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.858 19.963 ot_scf_mini 108 9.5 0.002 0.002 18.559 18.669 multiply_cannon 2286 13.5 0.299 0.305 15.639 17.135 multiply_cannon_loop 2286 14.5 0.870 0.890 14.311 15.939 ot_mini 108 10.5 0.001 0.001 11.092 11.248 qs_ot_get_derivative 108 11.5 0.001 0.001 8.824 8.935 multiply_cannon_multrec 9144 15.5 3.449 4.882 8.798 8.886 init_scf_loop 11 6.9 0.000 0.000 8.714 8.717 rebuild_ks_matrix 119 8.3 0.000 0.000 7.983 8.194 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.982 8.194 prepare_preconditioner 11 7.9 0.000 0.000 7.636 7.654 make_preconditioner 11 8.9 0.000 0.000 7.636 7.654 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.156 7.513 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.181 7.373 dbcsr_mm_accdrv_process 12550 15.8 4.682 6.535 5.210 6.567 mp_waitall_1 94719 16.7 3.462 5.424 3.462 5.424 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.954 5.010 calculate_rho_elec 119 8.7 0.118 0.121 4.953 5.009 cp_fm_upper_to_full 72 14.2 3.156 4.570 3.156 4.570 sum_up_and_integrate 119 10.3 0.001 0.001 4.413 4.419 integrate_v_rspace 119 11.3 0.003 0.003 4.403 4.409 qs_ot_get_p 119 10.4 0.001 0.001 4.213 4.394 init_scf_run 11 5.9 0.000 0.001 3.920 3.920 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.920 3.920 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.202 3.627 make_m2s 4572 13.5 0.038 0.038 3.205 3.372 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.812 3.351 make_images 4572 14.5 0.356 0.389 3.084 3.251 pw_transfer 1439 11.6 0.068 0.069 3.233 3.246 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.135 3.148 dbcsr_complete_redistribute 329 12.2 0.285 0.291 2.149 2.985 density_rs2pw 119 9.7 0.004 0.004 2.763 2.773 mp_sum_l 7287 12.8 1.570 2.629 1.570 2.629 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.770 2.586 fft_wrap_pw1pw2_140 487 13.2 0.577 0.581 2.564 2.576 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.335 2.571 apply_single 119 13.6 0.000 0.000 2.335 2.571 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.496 2.504 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.437 2.492 calculate_dm_sparse 119 9.5 0.000 0.000 2.434 2.481 mp_alltoall_i22 627 13.8 1.483 2.367 1.483 2.367 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.475 2.288 fft3d_ps 1201 14.6 0.597 0.607 2.230 2.240 calculate_first_density_matrix 1 7.0 0.000 0.000 2.161 2.214 ot_diis_step 108 11.5 0.014 0.014 2.199 2.201 grid_integrate_task_list 119 12.3 2.172 2.184 2.172 2.184 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.098 2.100 cp_fm_cholesky_invert 11 10.9 2.030 2.035 2.030 2.035 hybrid_alltoall_any 4725 16.4 0.090 0.150 1.391 1.941 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.902 1.903 make_images_data 4572 15.5 0.046 0.049 1.422 1.894 acc_transpose_blocks 9144 15.5 0.044 0.045 1.815 1.860 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.799 1.800 potential_pw2rs 119 12.3 0.014 0.014 1.756 1.758 cp_fm_diag_elpa 50 13.0 0.001 0.001 1.740 1.740 cp_fm_diag_elpa_base 50 14.0 1.588 1.648 1.737 1.737 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.597 1.655 wfi_extrapolate 11 7.9 0.001 0.001 1.638 1.638 mp_alltoall_d11v 2130 13.8 1.544 1.599 1.544 1.599 grid_collocate_task_list 119 9.7 1.585 1.596 1.585 1.596 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.492 1.547 mp_alltoall_z22v 1201 16.6 1.495 1.521 1.495 1.521 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.316 1.361 qs_env_update_s_mstruct 11 6.9 0.014 0.028 1.115 1.145 acc_transpose_blocks_sync 27432 16.5 1.105 1.141 1.105 1.141 transfer_rs2pw 487 10.6 0.005 0.005 1.104 1.132 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.993 1.038 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.015 1.035 make_images_sizes 4572 15.5 0.005 0.005 0.644 1.016 mp_alltoall_i44 4572 16.5 0.640 1.012 0.640 1.012 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=47.998000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=730.636364, yerr=14.201618 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 501.264384E+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 1383689. 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 0.020 0.035 91.479 91.481 qs_mol_dyn_low 1 2.0 0.004 0.015 90.930 90.939 qs_forces 11 3.9 0.003 0.003 90.377 90.378 qs_energies 11 4.9 0.016 0.050 87.334 87.344 scf_env_do_scf 11 5.9 0.001 0.001 77.640 77.643 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 70.696 70.697 dbcsr_multiply_generic 2055 12.4 0.108 0.115 53.318 53.544 qs_scf_new_mos 99 7.5 0.000 0.001 51.447 51.565 qs_scf_loop_do_ot 99 8.5 0.000 0.001 51.447 51.565 ot_scf_mini 99 9.5 0.002 0.003 48.937 49.050 velocity_verlet 10 3.0 0.001 0.002 47.731 47.737 multiply_cannon 2055 13.4 0.185 0.188 43.266 44.079 multiply_cannon_loop 2055 14.4 1.798 1.833 41.909 42.595 ot_mini 99 10.5 0.001 0.001 28.169 28.255 qs_ot_get_derivative 99 11.5 0.001 0.001 21.265 21.335 multiply_cannon_multrec 49320 15.4 11.245 11.996 17.230 18.055 rebuild_ks_matrix 110 8.3 0.000 0.001 15.844 15.957 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 15.843 15.957 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.914 14.019 mp_waitall_1 220248 16.4 12.437 13.553 12.437 13.553 qs_ot_get_p 110 10.4 0.001 0.001 11.869 11.984 multiply_cannon_sync_h2d 49320 15.4 9.386 10.076 9.386 10.076 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.937 8.472 qs_ot_p2m_diag 48 11.0 0.012 0.019 8.215 8.263 sum_up_and_integrate 110 10.3 0.001 0.002 8.061 8.079 integrate_v_rspace 110 11.3 0.003 0.003 8.036 8.059 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.301 7.824 apply_single 110 13.6 0.000 0.001 7.301 7.824 qs_rho_update_rho_low 110 7.6 0.000 0.001 7.616 7.770 calculate_rho_elec 110 8.6 0.021 0.025 7.616 7.770 multiply_cannon_metrocomm3 49320 15.4 0.083 0.087 6.669 7.610 init_scf_run 11 5.9 0.000 0.001 7.413 7.414 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.413 7.414 cp_dbcsr_syevd 48 12.0 0.002 0.003 7.207 7.209 init_scf_loop 11 6.9 0.000 0.002 6.914 6.916 ot_diis_step 99 11.5 0.006 0.006 6.661 6.662 cp_fm_diag_elpa 48 13.0 0.000 0.001 6.387 6.418 cp_fm_diag_elpa_base 48 14.0 6.362 6.390 6.384 6.415 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.106 6.165 dbcsr_mm_accdrv_process 87628 16.1 3.040 3.116 5.856 6.119 make_m2s 4110 13.4 0.061 0.065 5.099 5.262 mp_sum_l 6594 12.7 4.512 5.198 4.512 5.198 make_images 4110 14.4 0.179 0.208 5.001 5.166 density_rs2pw 110 9.6 0.004 0.005 4.508 4.671 prepare_preconditioner 11 7.9 0.000 0.000 4.570 4.585 make_preconditioner 11 8.9 0.000 0.001 4.570 4.585 wfi_extrapolate 11 7.9 0.001 0.001 4.516 4.517 make_full_inverse_cholesky 11 9.9 0.000 0.001 4.335 4.387 pw_transfer 1331 11.6 0.055 0.062 4.121 4.265 multiply_cannon_metrocomm1 49320 15.4 0.065 0.068 2.932 4.256 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.032 4.180 calculate_dm_sparse 110 9.5 0.001 0.001 3.908 4.055 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.732 3.738 potential_pw2rs 110 12.3 0.006 0.006 3.553 3.595 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.538 3.593 fft3d_ps 1111 14.6 0.796 0.890 3.333 3.452 fft_wrap_pw1pw2_140 451 13.1 0.385 0.425 3.257 3.401 grid_integrate_task_list 110 12.3 3.233 3.391 3.233 3.391 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 3.280 3.328 mp_alltoall_d11v 2046 13.8 2.551 2.961 2.551 2.961 transfer_rs2pw 451 10.6 0.005 0.006 2.622 2.798 calculate_first_density_matrix 1 7.0 0.000 0.002 2.762 2.765 acc_transpose_blocks 49320 15.4 0.218 0.224 2.590 2.640 jit_kernel_multiply 13 15.9 2.538 2.558 2.538 2.558 cp_fm_cholesky_invert 11 10.9 2.544 2.550 2.544 2.550 make_images_data 4110 15.4 0.044 0.048 2.205 2.425 transfer_pw2rs 451 13.1 0.006 0.006 2.337 2.374 mp_alltoall_z22v 1111 16.6 2.113 2.321 2.113 2.321 grid_collocate_task_list 110 9.6 2.163 2.305 2.163 2.305 mp_waitany 14300 13.8 1.977 2.218 1.977 2.218 hybrid_alltoall_any 4261 16.3 0.084 0.478 1.875 2.180 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.063 2.079 mp_sum_d 3889 11.9 1.437 2.000 1.437 2.000 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=91.481000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.818182, yerr=2.948595 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 593.473536E+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 3521759. 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 0.049 0.055 78.652 78.654 qs_mol_dyn_low 1 2.0 0.003 0.004 78.282 78.293 qs_forces 11 3.9 0.003 0.003 78.211 78.213 qs_energies 11 4.9 0.003 0.007 74.791 74.796 scf_env_do_scf 11 5.9 0.001 0.002 65.666 65.669 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 56.551 56.552 dbcsr_multiply_generic 2055 12.4 0.114 0.119 42.053 42.192 velocity_verlet 10 3.0 0.002 0.004 39.454 39.455 qs_scf_new_mos 99 7.5 0.001 0.001 39.103 39.216 qs_scf_loop_do_ot 99 8.5 0.001 0.001 39.102 39.216 ot_scf_mini 99 9.5 0.003 0.004 37.364 37.489 multiply_cannon 2055 13.4 0.223 0.244 34.259 35.407 multiply_cannon_loop 2055 14.4 1.165 1.205 32.786 33.611 ot_mini 99 10.5 0.001 0.001 21.941 22.065 multiply_cannon_multrec 24660 15.4 7.029 8.926 15.900 17.928 qs_ot_get_derivative 99 11.5 0.001 0.001 16.091 16.220 rebuild_ks_matrix 110 8.3 0.000 0.000 14.464 14.560 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 14.463 14.559 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.761 12.848 mp_waitall_1 176588 16.5 9.299 12.808 9.299 12.808 dbcsr_mm_accdrv_process 52282 16.1 6.376 8.064 8.700 9.610 multiply_cannon_metrocomm3 24660 15.4 0.073 0.076 5.777 9.477 init_scf_loop 11 6.9 0.001 0.003 9.071 9.072 qs_ot_get_p 110 10.4 0.001 0.001 8.534 8.701 multiply_cannon_sync_h2d 24660 15.4 6.412 7.608 6.412 7.608 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.535 7.258 apply_single 110 13.6 0.000 0.001 6.535 7.258 sum_up_and_integrate 110 10.3 0.002 0.003 7.195 7.203 integrate_v_rspace 110 11.3 0.002 0.003 7.167 7.177 prepare_preconditioner 11 7.9 0.000 0.000 6.922 6.947 make_preconditioner 11 8.9 0.000 0.001 6.922 6.947 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.823 6.831 calculate_rho_elec 110 8.6 0.039 0.047 6.823 6.831 init_scf_run 11 5.9 0.000 0.002 6.641 6.641 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.640 6.641 make_full_inverse_cholesky 11 9.9 0.000 0.001 6.438 6.620 qs_ot_p2m_diag 48 11.0 0.029 0.044 6.242 6.262 ot_diis_step 99 11.5 0.010 0.010 5.800 5.800 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.010 5.739 make_m2s 4110 13.4 0.057 0.060 5.280 5.717 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.713 5.713 make_images 4110 14.4 0.409 0.464 5.170 5.603 cp_fm_diag_elpa 48 13.0 0.001 0.001 4.963 4.977 cp_fm_diag_elpa_base 48 14.0 4.899 4.923 4.959 4.973 pw_transfer 1331 11.6 0.066 0.073 4.092 4.259 density_rs2pw 110 9.6 0.004 0.005 4.038 4.213 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.985 4.154 jit_kernel_multiply 10 16.2 1.961 3.915 1.961 3.915 wfi_extrapolate 11 7.9 0.001 0.001 3.868 3.868 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.704 3.775 grid_integrate_task_list 110 12.3 3.152 3.396 3.152 3.396 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.387 3.389 cp_fm_cholesky_invert 11 10.9 3.332 3.340 3.332 3.340 make_images_data 4110 15.4 0.049 0.053 2.766 3.271 fft3d_ps 1111 14.6 1.111 1.344 3.084 3.269 fft_wrap_pw1pw2_140 451 13.1 0.457 0.473 3.071 3.239 calculate_dm_sparse 110 9.5 0.001 0.001 3.130 3.158 hybrid_alltoall_any 4261 16.3 0.107 0.442 2.367 3.096 potential_pw2rs 110 12.3 0.008 0.009 3.028 3.058 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.989 3.026 calculate_first_density_matrix 1 7.0 0.000 0.003 2.666 2.669 mp_sum_l 6594 12.7 1.890 2.499 1.890 2.499 grid_collocate_task_list 110 9.6 2.170 2.329 2.170 2.329 mp_alltoall_d11v 2046 13.8 2.007 2.313 2.007 2.313 transfer_rs2pw 451 10.6 0.006 0.007 2.090 2.308 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.093 2.113 acc_transpose_blocks 24660 15.4 0.116 0.120 2.005 2.064 cp_fm_cholesky_decompose 22 10.9 1.889 1.895 1.889 1.895 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.890 1.891 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.861 1.872 mp_allgather_i34 2055 14.4 0.774 1.846 0.774 1.846 transfer_pw2rs 451 13.1 0.005 0.007 1.778 1.807 mp_alltoall_z22v 1111 16.6 1.660 1.760 1.660 1.760 dbcsr_complete_redistribute 325 12.2 0.235 0.299 1.365 1.668 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.644 1.654 mp_waitany 10164 13.8 1.350 1.576 1.350 1.576 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=78.654000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.545455, yerr=7.352966 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 667.979776E+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 0.028 0.051 73.927 73.929 qs_mol_dyn_low 1 2.0 0.003 0.005 73.540 73.550 qs_forces 11 3.9 0.003 0.003 73.433 73.435 qs_energies 11 4.9 0.005 0.019 70.079 70.092 scf_env_do_scf 11 5.9 0.001 0.002 61.235 61.236 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 50.182 50.183 velocity_verlet 10 3.0 0.033 0.044 38.541 38.543 dbcsr_multiply_generic 2055 12.4 0.110 0.115 34.835 35.515 qs_scf_new_mos 99 7.5 0.001 0.001 33.818 33.970 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.818 33.969 ot_scf_mini 99 9.5 0.002 0.003 32.388 32.541 multiply_cannon 2055 13.4 0.211 0.222 26.493 29.027 multiply_cannon_loop 2055 14.4 0.824 0.856 24.898 27.071 ot_mini 99 10.5 0.001 0.001 18.908 19.097 mp_waitall_1 139946 16.5 10.113 15.116 10.113 15.116 qs_ot_get_derivative 99 11.5 0.001 0.001 14.080 14.228 rebuild_ks_matrix 110 8.3 0.000 0.000 13.169 13.440 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.168 13.439 multiply_cannon_multrec 16440 15.4 3.526 4.683 11.196 12.806 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.622 11.859 init_scf_loop 11 6.9 0.001 0.005 11.003 11.005 multiply_cannon_metrocomm3 16440 15.4 0.046 0.048 5.841 10.461 prepare_preconditioner 11 7.9 0.000 0.000 9.072 9.103 make_preconditioner 11 8.9 0.000 0.001 9.072 9.103 make_full_inverse_cholesky 11 9.9 0.000 0.001 8.296 8.695 dbcsr_mm_accdrv_process 34862 16.1 6.264 7.937 7.524 8.401 qs_ot_get_p 110 10.4 0.001 0.001 8.111 8.366 sum_up_and_integrate 110 10.3 0.001 0.002 7.079 7.095 integrate_v_rspace 110 11.3 0.003 0.003 7.052 7.068 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.667 6.679 calculate_rho_elec 110 8.6 0.059 0.059 6.666 6.679 init_scf_run 11 5.9 0.000 0.001 6.155 6.155 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.155 6.155 qs_ot_p2m_diag 48 11.0 0.042 0.044 5.733 5.749 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.150 5.734 apply_single 110 13.6 0.000 0.000 5.150 5.734 make_m2s 4110 13.4 0.049 0.051 5.146 5.713 make_images 4110 14.4 0.401 0.521 5.031 5.599 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.235 5.236 ot_diis_step 99 11.5 0.011 0.011 4.718 4.718 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.884 4.528 cp_fm_diag_elpa 48 13.0 0.001 0.001 4.426 4.436 cp_fm_diag_elpa_base 48 14.0 4.344 4.384 4.422 4.432 cp_fm_cholesky_invert 11 10.9 4.064 4.075 4.064 4.075 pw_transfer 1331 11.6 0.066 0.073 4.018 4.033 density_rs2pw 110 9.6 0.004 0.005 3.837 4.028 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.911 3.928 mp_sum_l 6594 12.7 2.583 3.738 2.583 3.738 multiply_cannon_sync_h2d 16440 15.4 3.151 3.724 3.151 3.724 wfi_extrapolate 11 7.9 0.001 0.001 3.489 3.489 make_images_data 4110 15.4 0.046 0.049 2.821 3.465 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.358 3.441 grid_integrate_task_list 110 12.3 3.187 3.366 3.187 3.366 hybrid_alltoall_any 4261 16.3 0.109 0.384 2.510 3.270 fft_wrap_pw1pw2_140 451 13.1 0.578 0.589 3.125 3.148 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.991 2.994 potential_pw2rs 110 12.3 0.011 0.012 2.891 2.901 fft3d_ps 1111 14.6 1.093 1.110 2.854 2.866 calculate_dm_sparse 110 9.5 0.001 0.001 2.732 2.768 multiply_cannon_metrocomm1 16440 15.4 0.029 0.030 1.075 2.668 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.535 2.617 multiply_cannon_metrocomm4 14385 15.4 0.048 0.054 0.914 2.615 calculate_first_density_matrix 1 7.0 0.001 0.003 2.558 2.560 mp_alltoall_d11v 2046 13.8 2.219 2.534 2.219 2.534 mp_irecv_dv 48980 15.7 0.838 2.475 0.838 2.475 dbcsr_complete_redistribute 325 12.2 0.316 0.346 1.849 2.356 grid_collocate_task_list 110 9.6 2.223 2.345 2.223 2.345 jit_kernel_multiply 2 17.0 0.854 2.332 0.854 2.332 cp_fm_cholesky_decompose 22 10.9 2.303 2.326 2.303 2.326 transfer_rs2pw 451 10.6 0.005 0.006 1.965 2.155 mp_allgather_i34 2055 14.4 0.906 2.087 0.906 2.087 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 2.049 2.051 cp_fm_upper_to_full 70 14.2 1.449 1.871 1.449 1.871 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.746 1.782 transfer_pw2rs 451 13.1 0.005 0.005 1.743 1.751 copy_fm_to_dbcsr 174 11.2 0.001 0.002 1.225 1.721 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.642 1.687 acc_transpose_blocks 16440 15.4 0.078 0.079 1.600 1.684 mp_waitany 17072 13.8 1.427 1.644 1.427 1.644 mp_alltoall_z22v 1111 16.6 1.546 1.570 1.546 1.570 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.518 1.540 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=73.929000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=632.636364, yerr=9.373781 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 739.053568E+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 0.014 0.034 66.239 66.240 qs_mol_dyn_low 1 2.0 0.003 0.003 65.962 65.972 qs_forces 11 3.9 0.003 0.003 65.875 65.876 qs_energies 11 4.9 0.002 0.003 62.473 62.477 scf_env_do_scf 11 5.9 0.000 0.001 54.112 54.114 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.398 42.399 velocity_verlet 10 3.0 0.005 0.006 37.537 37.539 dbcsr_multiply_generic 2055 12.4 0.112 0.120 30.867 31.082 qs_scf_new_mos 99 7.5 0.001 0.001 27.820 27.932 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.819 27.932 ot_scf_mini 99 9.5 0.002 0.003 26.122 26.215 multiply_cannon 2055 13.4 0.237 0.258 23.756 24.699 multiply_cannon_loop 2055 14.4 1.412 1.490 22.265 22.931 ot_mini 99 10.5 0.001 0.001 14.974 15.079 multiply_cannon_multrec 24660 15.4 4.072 6.879 13.094 14.103 rebuild_ks_matrix 110 8.3 0.000 0.000 11.927 12.045 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.926 12.044 init_scf_loop 11 6.9 0.000 0.000 11.672 11.672 qs_ot_get_derivative 99 11.5 0.001 0.001 10.760 10.855 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.530 10.632 prepare_preconditioner 11 7.9 0.000 0.000 9.947 9.961 make_preconditioner 11 8.9 0.000 0.000 9.946 9.961 dbcsr_mm_accdrv_process 52304 16.0 7.772 9.183 8.863 9.795 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.166 9.620 mp_waitall_1 121746 16.5 4.197 6.327 4.197 6.327 sum_up_and_integrate 110 10.3 0.001 0.002 6.235 6.247 integrate_v_rspace 110 11.3 0.003 0.003 6.209 6.220 qs_ot_get_p 110 10.4 0.001 0.001 6.008 6.134 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.841 5.853 calculate_rho_elec 110 8.6 0.078 0.081 5.840 5.853 make_m2s 4110 13.4 0.059 0.061 5.405 5.790 make_images 4110 14.4 0.582 0.705 5.265 5.648 init_scf_run 11 5.9 0.000 0.001 5.621 5.621 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.620 5.621 cp_fm_upper_to_full 70 14.2 3.331 4.829 3.331 4.829 ot_diis_step 99 11.5 0.011 0.011 4.170 4.170 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.072 4.153 apply_single 110 13.6 0.000 0.000 4.072 4.152 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.080 4.099 dbcsr_complete_redistribute 325 12.2 0.414 0.453 2.667 3.794 pw_transfer 1331 11.6 0.065 0.073 3.673 3.707 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.562 3.608 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.566 3.602 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.510 3.511 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.406 3.454 grid_integrate_task_list 110 12.3 3.257 3.451 3.257 3.451 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.129 3.248 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.457 3.248 density_rs2pw 110 9.6 0.004 0.004 2.967 3.127 fft_wrap_pw1pw2_140 451 13.1 0.610 0.630 3.050 3.086 calculate_dm_sparse 110 9.5 0.001 0.001 3.036 3.077 wfi_extrapolate 11 7.9 0.001 0.001 3.000 3.000 make_images_data 4110 15.4 0.049 0.052 2.621 2.996 hybrid_alltoall_any 4261 16.3 0.123 0.466 2.274 2.981 cp_fm_diag_elpa 48 13.0 0.001 0.001 2.948 2.958 cp_fm_diag_elpa_base 48 14.0 2.794 2.851 2.944 2.955 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.771 2.876 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.849 2.851 mp_alltoall_i22 605 13.7 1.646 2.790 1.646 2.790 cp_fm_cholesky_invert 11 10.9 2.744 2.752 2.744 2.752 multiply_cannon_sync_h2d 24660 15.4 2.343 2.605 2.343 2.605 acc_transpose_blocks 24660 15.4 0.115 0.118 2.432 2.565 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.493 2.535 calculate_first_density_matrix 1 7.0 0.000 0.000 2.519 2.522 fft3d_ps 1111 14.6 1.086 1.117 2.447 2.468 grid_collocate_task_list 110 9.6 2.262 2.368 2.262 2.368 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.237 2.237 mp_alltoall_d11v 2046 13.8 1.829 2.069 1.829 2.069 potential_pw2rs 110 12.3 0.012 0.013 2.051 2.057 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.820 1.860 cp_fm_cholesky_decompose 22 10.9 1.683 1.733 1.683 1.733 mp_sum_l 6594 12.7 1.033 1.728 1.033 1.728 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.611 1.706 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.639 1.654 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.584 1.602 acc_transpose_blocks_sync 73980 16.4 1.402 1.530 1.402 1.530 mp_allgather_i34 2055 14.4 0.575 1.507 0.575 1.507 multiply_cannon_metrocomm4 20550 15.4 0.062 0.065 0.874 1.479 transfer_rs2pw 451 10.6 0.005 0.005 1.200 1.381 mp_irecv_dv 62702 16.1 0.768 1.348 0.768 1.348 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.240000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=699.000000, yerr=10.608744 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 869.392384E+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 0.026 0.034 60.917 60.919 qs_mol_dyn_low 1 2.0 0.003 0.003 60.621 60.631 qs_forces 11 3.9 0.003 0.004 60.533 60.534 qs_energies 11 4.9 0.001 0.002 56.796 56.801 scf_env_do_scf 11 5.9 0.000 0.001 48.058 48.058 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 38.819 38.820 velocity_verlet 10 3.0 0.002 0.003 34.240 34.244 dbcsr_multiply_generic 2055 12.4 0.105 0.109 24.953 25.213 qs_scf_new_mos 99 7.5 0.001 0.001 23.390 23.459 qs_scf_loop_do_ot 99 8.5 0.001 0.001 23.390 23.458 ot_scf_mini 99 9.5 0.002 0.002 22.054 22.091 multiply_cannon 2055 13.4 0.238 0.253 18.451 20.250 multiply_cannon_loop 2055 14.4 0.605 0.632 17.052 17.655 rebuild_ks_matrix 110 8.3 0.000 0.000 12.211 12.239 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.211 12.239 ot_mini 99 10.5 0.001 0.001 11.883 11.908 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.854 10.884 init_scf_loop 11 6.9 0.000 0.000 9.184 9.185 multiply_cannon_multrec 8220 15.4 3.212 4.352 7.784 8.731 mp_waitall_1 103326 16.6 6.891 8.541 6.891 8.541 qs_ot_get_derivative 99 11.5 0.001 0.001 8.135 8.174 prepare_preconditioner 11 7.9 0.000 0.000 7.461 7.467 make_preconditioner 11 8.9 0.000 0.000 7.461 7.467 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.994 7.119 sum_up_and_integrate 110 10.3 0.002 0.002 6.675 6.687 integrate_v_rspace 110 11.3 0.003 0.003 6.648 6.661 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.426 6.444 calculate_rho_elec 110 8.6 0.114 0.114 6.426 6.443 qs_ot_get_p 110 10.4 0.001 0.001 5.970 6.005 init_scf_run 11 5.9 0.000 0.001 5.527 5.527 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.527 5.527 dbcsr_mm_accdrv_process 17442 15.9 3.083 4.305 4.435 5.414 make_m2s 4110 13.4 0.038 0.039 4.894 5.230 make_images 4110 14.4 0.652 0.715 4.764 5.098 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 2.908 4.411 qs_ot_p2m_diag 48 11.0 0.081 0.085 4.305 4.314 pw_transfer 1331 11.6 0.065 0.070 4.213 4.234 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.105 4.130 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.898 3.899 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.763 3.799 apply_single 110 13.6 0.000 0.000 3.763 3.799 cp_fm_cholesky_invert 11 10.9 3.767 3.771 3.767 3.771 ot_diis_step 99 11.5 0.012 0.012 3.706 3.707 density_rs2pw 110 9.6 0.004 0.004 3.374 3.587 fft_wrap_pw1pw2_140 451 13.1 0.776 0.789 3.469 3.495 grid_integrate_task_list 110 12.3 3.384 3.486 3.384 3.486 cp_fm_diag_elpa 48 13.0 0.001 0.001 3.246 3.253 cp_fm_diag_elpa_base 48 14.0 3.179 3.210 3.243 3.251 make_images_data 4110 15.4 0.040 0.046 2.615 3.202 hybrid_alltoall_any 4261 16.3 0.201 0.857 2.506 3.163 wfi_extrapolate 11 7.9 0.001 0.001 2.990 2.990 fft3d_ps 1111 14.6 1.137 1.161 2.771 2.793 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.692 2.693 calculate_dm_sparse 110 9.5 0.001 0.001 2.641 2.676 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.639 2.642 grid_collocate_task_list 110 9.6 2.367 2.482 2.367 2.482 multiply_cannon_sync_h2d 8220 15.4 2.371 2.442 2.371 2.442 calculate_first_density_matrix 1 7.0 0.000 0.000 2.430 2.431 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.401 2.430 potential_pw2rs 110 12.3 0.015 0.015 2.370 2.380 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.263 2.283 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.233 2.252 cp_fm_cholesky_decompose 22 10.9 2.146 2.165 2.146 2.165 mp_alltoall_d11v 2046 13.8 1.936 2.118 1.936 2.118 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.785 2.003 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 1.305 1.851 mp_allgather_i34 2055 14.4 0.569 1.772 0.569 1.772 dbcsr_complete_redistribute 325 12.2 0.552 0.575 1.607 1.714 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.687 1.691 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.515 1.627 transfer_rs2pw 451 10.6 0.005 0.005 1.397 1.591 jit_kernel_multiply 8 15.6 1.035 1.541 1.035 1.541 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.520 1.531 mp_alltoall_z22v 1111 16.6 1.399 1.434 1.399 1.434 acc_transpose_blocks 8220 15.4 0.039 0.040 1.392 1.433 mp_waitany 9240 13.8 1.160 1.355 1.160 1.355 qs_create_task_list 11 7.9 0.000 0.000 1.220 1.319 generate_qs_task_list 11 8.9 0.372 0.440 1.220 1.319 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.266 1.289 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=60.919000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=810.636364, yerr=13.593265 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 1.414046E+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 0.020 0.038 95.727 95.728 qs_mol_dyn_low 1 2.0 0.003 0.003 95.379 95.387 qs_forces 11 3.9 0.003 0.003 94.376 94.377 qs_energies 11 4.9 0.002 0.002 90.059 90.065 scf_env_do_scf 11 5.9 0.001 0.001 79.569 79.570 velocity_verlet 10 3.0 0.002 0.003 61.164 61.287 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 48.881 48.883 dbcsr_multiply_generic 2055 12.4 0.120 0.124 31.571 31.649 init_scf_loop 11 6.9 0.000 0.000 30.608 30.611 qs_scf_new_mos 99 7.5 0.001 0.001 30.017 30.093 qs_scf_loop_do_ot 99 8.5 0.001 0.001 30.017 30.093 prepare_preconditioner 11 7.9 0.000 0.000 28.486 28.494 make_preconditioner 11 8.9 0.000 0.000 28.486 28.494 ot_scf_mini 99 9.5 0.002 0.002 28.117 28.191 make_full_inverse_cholesky 11 9.9 0.000 0.000 22.702 27.927 multiply_cannon 2055 13.4 0.337 0.353 23.399 24.654 multiply_cannon_loop 2055 14.4 0.824 0.831 21.330 21.996 cp_fm_upper_to_full 70 14.2 12.608 18.095 12.608 18.095 ot_mini 99 10.5 0.001 0.001 15.594 15.628 rebuild_ks_matrix 110 8.3 0.001 0.001 14.557 14.608 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.556 14.608 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.154 13.197 dbcsr_complete_redistribute 325 12.2 1.023 1.043 7.556 10.799 qs_ot_get_derivative 99 11.5 0.001 0.001 10.678 10.743 mp_waitall_1 84994 16.7 8.816 9.935 8.816 9.935 multiply_cannon_multrec 8220 15.4 4.079 4.242 9.764 9.868 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.414 9.651 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.768 8.974 mp_alltoall_i22 605 13.7 5.409 8.662 5.409 8.662 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.994 8.033 calculate_rho_elec 110 8.6 0.225 0.225 7.993 8.033 sum_up_and_integrate 110 10.3 0.002 0.002 7.457 7.474 integrate_v_rspace 110 11.3 0.003 0.003 7.429 7.446 qs_ot_get_p 110 10.4 0.001 0.001 6.875 6.926 cp_fm_cholesky_invert 11 10.9 6.811 6.816 6.811 6.816 make_m2s 4110 13.4 0.043 0.044 6.020 6.460 make_images 4110 14.4 0.892 0.948 5.831 6.270 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 5.517 6.201 init_scf_run 11 5.9 0.000 0.001 6.122 6.122 scf_env_initial_rho_setup 11 6.9 0.000 0.000 6.122 6.122 dbcsr_mm_accdrv_process 11614 15.7 3.934 4.486 5.537 5.745 pw_transfer 1331 11.6 0.075 0.075 5.731 5.738 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.614 5.622 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.003 5.493 apply_single 110 13.6 0.000 0.000 5.003 5.493 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.988 4.992 ot_diis_step 99 11.5 0.015 0.016 4.862 4.863 fft_wrap_pw1pw2_140 451 13.1 1.282 1.287 4.743 4.756 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.447 4.449 density_rs2pw 110 9.6 0.004 0.004 4.325 4.351 hybrid_alltoall_any 4261 16.3 0.264 0.563 3.091 4.032 make_images_data 4110 15.4 0.045 0.048 3.135 3.947 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.371 3.851 grid_integrate_task_list 110 12.3 3.699 3.774 3.699 3.774 cp_fm_diag_elpa 48 13.0 0.001 0.001 3.736 3.737 cp_fm_diag_elpa_base 48 14.0 3.178 3.380 3.734 3.734 wfi_extrapolate 11 7.9 0.001 0.001 3.696 3.697 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.693 3.694 fft3d_ps 1111 14.6 1.302 1.327 3.664 3.673 calculate_dm_sparse 110 9.5 0.001 0.001 3.285 3.328 multiply_cannon_sync_h2d 8220 15.4 3.129 3.146 3.129 3.146 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.031 3.071 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.044 3.045 potential_pw2rs 110 12.3 0.021 0.021 2.807 2.818 cp_fm_cholesky_decompose 22 10.9 2.739 2.760 2.739 2.760 grid_collocate_task_list 110 9.6 2.686 2.722 2.686 2.722 mp_alltoall_d11v 2046 13.8 2.532 2.600 2.532 2.600 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.456 2.470 calculate_first_density_matrix 1 7.0 0.000 0.000 2.287 2.289 qs_env_update_s_mstruct 11 6.9 0.002 0.002 2.207 2.273 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.099 2.210 mp_alltoall_z22v 1111 16.6 2.109 2.119 2.109 2.119 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.004 2.012 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.929 1.965 mp_allgather_i34 2055 14.4 0.818 1.954 0.818 1.954 qs_create_task_list 11 7.9 0.000 0.000 1.890 1.939 generate_qs_task_list 11 8.9 0.730 0.781 1.890 1.939 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=95.728000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1261.636364, yerr=70.785557 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 630.992896E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2529110. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.028 220.262 220.263 qs_mol_dyn_low 1 2.0 0.003 0.003 219.763 219.777 qs_forces 11 3.9 0.004 0.006 219.671 219.673 qs_energies 11 4.9 0.002 0.003 213.944 213.955 scf_env_do_scf 11 5.9 0.001 0.001 197.036 197.040 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 173.602 173.603 velocity_verlet 10 3.0 0.002 0.003 132.213 132.214 qs_scf_new_mos 117 7.6 0.001 0.001 131.919 132.125 qs_scf_loop_do_ot 117 8.6 0.001 0.001 131.918 132.125 dbcsr_multiply_generic 2507 12.6 0.180 0.182 127.582 128.548 ot_scf_mini 117 9.6 0.003 0.003 125.196 125.372 multiply_cannon 2507 13.6 0.241 0.248 102.193 104.523 multiply_cannon_loop 2507 14.6 2.427 2.483 99.698 101.707 ot_mini 117 10.6 0.001 0.001 68.291 68.488 multiply_cannon_multrec 60168 15.6 31.669 34.647 41.457 43.621 qs_ot_get_derivative 117 11.6 0.001 0.001 43.300 43.502 rebuild_ks_matrix 128 8.3 0.001 0.001 34.819 35.121 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 34.818 35.121 qs_ot_get_p 128 10.4 0.001 0.001 34.567 34.798 mp_waitall_1 267128 16.5 30.687 33.742 30.687 33.742 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.295 31.571 multiply_cannon_sync_h2d 60168 15.6 26.115 28.406 26.115 28.406 qs_ot_p2m_diag 83 11.4 0.079 0.091 27.377 27.431 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.355 25.335 apply_single 128 13.6 0.001 0.001 24.354 25.334 ot_diis_step 117 11.6 0.008 0.008 24.690 24.691 cp_dbcsr_syevd 83 12.4 0.005 0.005 24.582 24.584 init_scf_loop 11 6.9 0.000 0.000 23.354 23.355 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.923 21.086 cp_fm_diag_elpa 83 13.4 0.001 0.002 21.043 21.073 cp_fm_diag_elpa_base 83 14.4 20.941 20.990 21.038 21.066 prepare_preconditioner 11 7.9 0.000 0.000 18.610 18.653 make_preconditioner 11 8.9 0.000 0.000 18.610 18.653 multiply_cannon_metrocomm3 60168 15.6 0.117 0.123 16.156 18.385 make_full_inverse_cholesky 11 9.9 0.000 0.000 17.827 18.017 make_m2s 5014 13.6 0.104 0.112 14.870 15.298 sum_up_and_integrate 128 10.3 0.002 0.004 15.142 15.158 make_images 5014 14.6 0.399 0.417 14.686 15.128 integrate_v_rspace 128 11.3 0.004 0.004 15.083 15.100 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.475 14.599 calculate_rho_elec 128 8.7 0.045 0.064 14.474 14.598 init_scf_run 11 5.9 0.000 0.001 12.646 12.647 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.646 12.647 cp_fm_cholesky_invert 11 10.9 10.838 10.847 10.838 10.847 mp_sum_l 7950 12.9 9.255 10.351 9.255 10.351 dbcsr_mm_accdrv_process 124484 16.2 4.723 4.838 9.351 9.892 wfi_extrapolate 11 7.9 0.001 0.001 9.448 9.448 density_rs2pw 128 9.7 0.006 0.007 8.152 8.922 calculate_dm_sparse 128 9.5 0.001 0.001 8.538 8.649 multiply_cannon_metrocomm1 60168 15.6 0.093 0.097 6.780 8.626 pw_transfer 1547 11.6 0.075 0.092 8.343 8.601 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.361 8.523 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.139 8.396 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.189 8.299 make_images_data 5014 15.6 0.068 0.074 7.264 8.208 grid_integrate_task_list 128 12.3 7.026 7.542 7.026 7.542 hybrid_alltoall_any 5200 16.5 0.295 2.259 6.287 7.402 fft_wrap_pw1pw2_140 523 13.2 1.113 1.169 6.680 6.924 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.690 6.701 fft3d_ps 1291 14.7 2.195 2.846 6.158 6.573 mp_alltoall_d11v 2415 14.1 4.516 5.943 4.516 5.943 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.790 5.868 potential_pw2rs 128 12.3 0.009 0.010 5.602 5.674 cp_fm_cholesky_decompose 22 10.9 5.498 5.511 5.498 5.511 grid_collocate_task_list 128 9.7 4.841 5.295 4.841 5.295 transfer_rs2pw 523 10.6 0.008 0.009 4.040 4.818 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=220.263000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.272727, yerr=6.239040 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 837.550080E+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 6395397. 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 0.028 0.104 191.583 191.586 qs_mol_dyn_low 1 2.0 0.005 0.022 191.046 191.061 qs_forces 11 3.9 0.010 0.056 190.897 190.901 qs_energies 11 4.9 0.004 0.019 184.080 184.093 scf_env_do_scf 11 5.9 0.001 0.002 167.593 167.603 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 134.503 134.505 velocity_verlet 10 3.0 0.002 0.003 120.784 120.786 dbcsr_multiply_generic 2527 12.6 0.190 0.196 98.019 99.290 qs_scf_new_mos 118 7.6 0.001 0.001 95.615 96.204 qs_scf_loop_do_ot 118 8.6 0.001 0.001 95.614 96.203 ot_scf_mini 118 9.6 0.004 0.005 90.768 91.418 multiply_cannon 2527 13.6 0.505 0.566 77.632 82.153 multiply_cannon_loop 2527 14.6 1.588 1.661 74.652 77.292 ot_mini 118 10.6 0.001 0.001 51.001 51.652 mp_waitall_1 216438 16.6 24.725 38.731 24.725 38.731 multiply_cannon_multrec 30324 15.6 20.965 25.443 31.650 36.533 rebuild_ks_matrix 129 8.3 0.001 0.001 32.498 33.152 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 32.498 33.151 init_scf_loop 11 6.9 0.001 0.006 33.001 33.002 multiply_cannon_metrocomm3 30324 15.6 0.098 0.103 16.247 30.245 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.242 29.836 qs_ot_get_derivative 118 11.6 0.001 0.002 28.722 29.357 prepare_preconditioner 11 7.9 0.000 0.001 28.651 28.729 make_preconditioner 11 8.9 0.000 0.002 28.651 28.729 make_full_inverse_cholesky 11 9.9 0.000 0.001 27.330 27.902 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.334 23.448 apply_single 129 13.6 0.001 0.001 22.333 23.448 qs_ot_get_p 129 10.4 0.001 0.001 21.794 22.518 ot_diis_step 118 11.6 0.014 0.015 22.099 22.101 multiply_cannon_sync_h2d 30324 15.6 17.944 19.825 17.944 19.825 qs_ot_p2m_diag 83 11.4 0.189 0.217 16.820 16.857 cp_fm_cholesky_invert 11 10.9 16.769 16.781 16.769 16.781 cp_dbcsr_syevd 83 12.4 0.006 0.006 15.597 15.597 make_m2s 5054 13.6 0.091 0.096 14.333 15.579 make_images 5054 14.6 1.196 1.402 14.120 15.363 sum_up_and_integrate 129 10.3 0.002 0.004 14.213 14.250 integrate_v_rspace 129 11.3 0.004 0.027 14.153 14.192 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.530 13.565 calculate_rho_elec 129 8.7 0.088 0.104 13.529 13.565 cp_fm_diag_elpa 83 13.4 0.001 0.001 12.401 12.437 cp_fm_diag_elpa_base 83 14.4 12.145 12.232 12.394 12.427 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.203 11.641 init_scf_run 11 5.9 0.000 0.001 11.531 11.533 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.531 11.533 multiply_cannon_metrocomm4 27797 15.6 0.106 0.119 3.975 11.224 dbcsr_mm_accdrv_process 62734 16.2 5.579 6.425 10.130 10.973 mp_irecv_dv 70031 16.3 3.765 10.802 3.765 10.802 make_images_data 5054 15.6 0.068 0.077 8.414 9.932 hybrid_alltoall_any 5240 16.5 0.356 1.546 7.157 9.404 pw_transfer 1559 11.6 0.085 0.096 8.698 8.767 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.473 8.547 wfi_extrapolate 11 7.9 0.001 0.001 8.373 8.373 density_rs2pw 129 9.7 0.006 0.007 7.157 7.581 fft_wrap_pw1pw2_140 527 13.2 1.226 1.242 7.497 7.580 grid_integrate_task_list 129 12.3 7.228 7.557 7.228 7.557 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.742 7.442 cp_fm_cholesky_decompose 22 10.9 7.013 7.088 7.013 7.088 calculate_dm_sparse 129 9.5 0.001 0.001 6.519 6.665 mp_sum_l 8010 12.9 4.292 6.454 4.292 6.454 fft3d_ps 1301 14.7 2.859 3.043 6.010 6.080 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.057 6.069 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.446 5.509 grid_collocate_task_list 129 9.7 5.047 5.387 5.047 5.387 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.149 5.334 mp_alltoall_d11v 2423 14.1 4.258 4.862 4.258 4.862 potential_pw2rs 129 12.3 0.016 0.019 4.605 4.628 mp_allgather_i34 2527 14.6 1.493 4.540 1.493 4.540 mp_sum_d 4499 12.1 2.705 4.087 2.705 4.087 dbcsr_complete_redistribute 395 12.7 0.776 0.852 3.150 4.000 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.586000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=798.363636, yerr=2.012359 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+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 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 964.681728E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58351. MP_Allreduce 11057 1000. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.059 191.669 191.676 qs_mol_dyn_low 1 2.0 0.003 0.004 191.107 191.121 qs_forces 11 3.9 0.004 0.004 191.011 191.014 qs_energies 11 4.9 0.007 0.034 184.157 184.182 scf_env_do_scf 11 5.9 0.001 0.004 167.664 167.666 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 128.515 128.519 velocity_verlet 10 3.0 0.003 0.006 122.362 122.365 qs_scf_new_mos 117 7.6 0.001 0.001 89.307 89.674 qs_scf_loop_do_ot 117 8.6 0.001 0.001 89.306 89.673 dbcsr_multiply_generic 2507 12.6 0.181 0.188 86.269 87.627 ot_scf_mini 117 9.6 0.004 0.004 84.936 85.362 multiply_cannon 2507 13.6 0.510 0.525 64.127 68.730 multiply_cannon_loop 2507 14.6 1.136 1.211 60.349 64.230 ot_mini 117 10.6 0.001 0.001 45.819 46.267 init_scf_loop 11 6.9 0.001 0.008 39.037 39.040 mp_waitall_1 170520 16.6 26.833 38.310 26.833 38.310 prepare_preconditioner 11 7.9 0.000 0.001 34.852 34.904 make_preconditioner 11 8.9 0.000 0.001 34.852 34.904 make_full_inverse_cholesky 11 9.9 0.000 0.001 32.411 33.880 rebuild_ks_matrix 128 8.3 0.001 0.001 31.432 31.994 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 31.432 31.993 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.269 28.774 multiply_cannon_metrocomm3 20056 15.6 0.063 0.066 16.181 27.264 qs_ot_get_derivative 117 11.6 0.002 0.002 26.055 26.489 multiply_cannon_multrec 20056 15.6 12.897 15.675 22.318 25.170 qs_ot_get_p 128 10.4 0.001 0.002 23.940 24.478 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.134 21.394 apply_single 128 13.6 0.001 0.001 20.133 21.394 ot_diis_step 117 11.6 0.018 0.018 19.610 19.610 qs_ot_p2m_diag 83 11.4 0.266 0.273 19.130 19.190 cp_dbcsr_syevd 83 12.4 0.005 0.006 17.806 17.808 make_m2s 5014 13.6 0.079 0.084 15.650 17.187 cp_fm_cholesky_invert 11 10.9 16.991 17.007 16.991 17.007 make_images 5014 14.6 1.165 1.258 15.418 16.953 multiply_cannon_sync_h2d 20056 15.6 13.490 15.046 13.490 15.046 sum_up_and_integrate 128 10.3 0.002 0.003 14.925 14.964 integrate_v_rspace 128 11.3 0.003 0.004 14.866 14.904 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.845 14.903 calculate_rho_elec 128 8.7 0.130 0.145 14.844 14.902 cp_fm_diag_elpa 83 13.4 0.002 0.005 14.142 14.165 cp_fm_diag_elpa_base 83 14.4 13.671 13.879 14.133 14.157 make_images_data 5014 15.6 0.062 0.072 9.285 11.437 init_scf_run 11 5.9 0.000 0.001 11.241 11.241 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.240 11.241 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.296 10.613 hybrid_alltoall_any 5200 16.5 0.451 2.046 8.022 10.552 multiply_cannon_metrocomm4 17549 15.6 0.066 0.076 3.523 9.657 pw_transfer 1547 11.6 0.085 0.102 9.410 9.552 mp_irecv_dv 50230 16.2 3.393 9.395 3.393 9.395 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.187 9.335 dbcsr_mm_accdrv_process 41502 16.2 5.735 6.043 8.894 9.043 cp_fm_cholesky_decompose 22 10.9 8.626 8.693 8.626 8.693 density_rs2pw 128 9.7 0.006 0.006 8.057 8.300 fft_wrap_pw1pw2_140 523 13.2 1.299 1.335 7.854 8.018 wfi_extrapolate 11 7.9 0.001 0.001 7.981 7.982 grid_integrate_task_list 128 12.3 7.319 7.893 7.319 7.893 cp_fm_upper_to_full 105 14.8 5.804 7.433 5.804 7.433 mp_sum_l 7950 12.9 4.681 6.977 4.681 6.977 dbcsr_complete_redistribute 395 12.7 1.176 1.212 4.925 6.758 fft3d_ps 1291 14.7 2.763 2.992 6.565 6.673 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.280 6.289 calculate_dm_sparse 128 9.5 0.001 0.001 6.037 6.129 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.656 5.661 grid_collocate_task_list 128 9.7 5.151 5.536 5.151 5.536 mp_alltoall_d11v 2415 14.1 4.791 5.526 4.791 5.526 mp_allgather_i34 2507 14.6 2.269 5.473 2.269 5.473 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.632 5.464 potential_pw2rs 128 12.3 0.020 0.022 5.259 5.296 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.642 4.796 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.302 4.334 transfer_fm_to_dbcsr 11 9.9 0.018 0.024 2.421 4.237 mp_alltoall_i22 716 14.1 2.101 4.048 2.101 4.048 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=191.676000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=915.181818, yerr=7.825578 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.153790E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379345520 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58348. MP_Allreduce 11057 1083. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.115 203.512 203.514 qs_mol_dyn_low 1 2.0 0.003 0.003 203.012 203.026 qs_forces 11 3.9 0.004 0.004 202.892 202.907 qs_energies 11 4.9 0.003 0.007 195.675 195.687 scf_env_do_scf 11 5.9 0.001 0.002 177.527 177.541 velocity_verlet 10 3.0 0.002 0.002 133.812 133.815 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 126.429 126.431 qs_scf_new_mos 117 7.6 0.001 0.001 88.614 88.953 qs_scf_loop_do_ot 117 8.6 0.001 0.001 88.613 88.952 ot_scf_mini 117 9.6 0.003 0.004 84.004 84.332 dbcsr_multiply_generic 2507 12.6 0.187 0.193 83.255 84.042 multiply_cannon 2507 13.6 0.544 0.571 57.036 60.320 multiply_cannon_loop 2507 14.6 1.857 1.920 53.166 54.976 init_scf_loop 11 6.9 0.001 0.004 50.958 50.959 prepare_preconditioner 11 7.9 0.000 0.000 46.761 46.795 make_preconditioner 11 8.9 0.000 0.002 46.761 46.795 make_full_inverse_cholesky 11 9.9 0.011 0.023 40.218 45.448 ot_mini 117 10.6 0.001 0.001 44.625 44.934 multiply_cannon_multrec 30084 15.6 13.641 18.766 26.803 31.502 rebuild_ks_matrix 128 8.3 0.001 0.001 30.129 30.442 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 30.129 30.442 mp_waitall_1 147882 16.7 18.743 27.670 18.743 27.670 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.196 27.478 qs_ot_get_p 128 10.4 0.001 0.001 25.293 25.655 qs_ot_get_derivative 117 11.6 0.002 0.002 24.888 25.219 make_m2s 5014 13.6 0.095 0.099 21.652 22.767 make_images 5014 14.6 1.968 2.267 21.351 22.465 qs_ot_p2m_diag 83 11.4 0.343 0.390 20.725 20.776 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.112 19.761 apply_single 128 13.6 0.001 0.001 19.112 19.760 ot_diis_step 117 11.6 0.017 0.018 19.609 19.611 cp_dbcsr_syevd 83 12.4 0.005 0.006 19.182 19.185 cp_fm_cholesky_invert 11 10.9 18.588 18.598 18.588 18.598 cp_fm_upper_to_full 105 14.8 11.857 17.538 11.857 17.538 cp_fm_diag_elpa 83 13.4 0.001 0.002 15.463 15.477 cp_fm_diag_elpa_base 83 14.4 14.406 14.756 15.454 15.467 multiply_cannon_metrocomm3 30084 15.6 0.049 0.052 6.358 14.920 sum_up_and_integrate 128 10.3 0.002 0.003 14.698 14.730 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.685 14.726 calculate_rho_elec 128 8.7 0.173 0.189 14.685 14.725 integrate_v_rspace 128 11.3 0.003 0.004 14.638 14.672 make_images_data 5014 15.6 0.066 0.073 11.683 13.781 dbcsr_complete_redistribute 395 12.7 1.522 1.665 9.607 13.587 dbcsr_mm_accdrv_process 62264 16.2 8.593 9.447 12.735 13.389 hybrid_alltoall_any 5200 16.5 0.532 2.218 10.208 12.554 init_scf_run 11 5.9 0.000 0.001 12.048 12.049 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.047 12.049 copy_fm_to_dbcsr 209 11.7 0.002 0.002 8.072 12.012 multiply_cannon_sync_h2d 30084 15.6 10.419 11.258 10.419 11.258 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 10.497 10.736 transfer_fm_to_dbcsr 11 9.9 0.001 0.012 6.521 10.413 mp_alltoall_i22 716 14.1 5.799 9.841 5.799 9.841 pw_transfer 1547 11.6 0.086 0.101 9.335 9.419 cp_fm_cholesky_decompose 22 10.9 9.155 9.248 9.155 9.248 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.109 9.197 wfi_extrapolate 11 7.9 0.001 0.001 8.137 8.137 fft_wrap_pw1pw2_140 523 13.2 1.445 1.490 7.929 8.014 density_rs2pw 128 9.7 0.006 0.006 7.680 7.995 grid_integrate_task_list 128 12.3 7.557 7.920 7.557 7.920 calculate_dm_sparse 128 9.5 0.001 0.001 6.964 7.036 multiply_cannon_metrocomm4 25070 15.6 0.084 0.093 2.819 6.402 fft3d_ps 1291 14.7 2.842 2.897 6.287 6.333 mp_irecv_dv 76098 16.2 2.664 6.138 2.664 6.138 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.740 5.795 mp_alltoall_d11v 2415 14.1 5.290 5.714 5.290 5.714 grid_collocate_task_list 128 9.7 5.289 5.541 5.289 5.541 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.845 4.955 potential_pw2rs 128 12.3 0.023 0.023 4.915 4.943 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.513 4.568 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.481 4.483 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.355 4.425 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=203.514000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1093.545455, yerr=17.116519 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796578E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500608 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.539695E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108623056 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58187. MP_Allreduce 11085 1167. MP_Sync 86 MP_Alltoall 1700 18828162. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.035 190.249 190.251 qs_mol_dyn_low 1 2.0 0.003 0.003 189.837 189.850 qs_forces 11 3.9 0.004 0.004 188.941 188.943 qs_energies 11 4.9 0.002 0.002 181.341 181.348 scf_env_do_scf 11 5.9 0.001 0.001 163.265 163.276 velocity_verlet 10 3.0 0.002 0.003 124.162 124.216 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 121.037 121.039 qs_scf_new_mos 116 7.6 0.001 0.001 82.705 82.853 qs_scf_loop_do_ot 116 8.6 0.001 0.001 82.704 82.852 ot_scf_mini 116 9.6 0.003 0.004 78.175 78.244 dbcsr_multiply_generic 2485 12.5 0.180 0.191 74.596 75.161 multiply_cannon 2485 13.5 0.556 0.582 54.480 58.515 multiply_cannon_loop 2485 14.5 0.805 0.835 51.050 52.397 init_scf_loop 11 6.9 0.000 0.000 42.063 42.065 ot_mini 116 10.6 0.001 0.001 39.826 39.923 prepare_preconditioner 11 7.9 0.000 0.000 38.095 38.122 make_preconditioner 11 8.9 0.000 0.000 38.095 38.122 make_full_inverse_cholesky 11 9.9 0.015 0.028 35.780 36.129 mp_waitall_1 124680 16.7 26.567 33.932 26.567 33.932 rebuild_ks_matrix 127 8.3 0.001 0.001 29.880 29.998 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 29.879 29.997 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.204 27.312 qs_ot_get_p 127 10.4 0.001 0.001 25.377 25.490 qs_ot_get_derivative 116 11.6 0.001 0.002 22.771 22.837 cp_fm_cholesky_invert 11 10.9 22.593 22.601 22.593 22.601 qs_ot_p2m_diag 82 11.4 0.490 0.496 20.945 20.963 multiply_cannon_multrec 9940 15.5 10.295 14.730 17.881 20.937 multiply_cannon_metrocomm3 9940 15.5 0.025 0.027 12.647 20.374 cp_dbcsr_syevd 82 12.4 0.005 0.006 19.498 19.500 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.048 17.315 apply_single 127 13.6 0.001 0.001 17.048 17.314 ot_diis_step 116 11.6 0.020 0.020 16.937 16.937 make_m2s 4970 13.5 0.065 0.070 15.786 16.689 make_images 4970 14.5 2.167 2.583 15.484 16.393 cp_fm_diag_elpa 82 13.4 0.001 0.001 15.879 15.887 cp_fm_diag_elpa_base 82 14.4 15.602 15.711 15.871 15.880 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.084 15.125 calculate_rho_elec 127 8.7 0.254 0.265 15.083 15.124 sum_up_and_integrate 127 10.3 0.002 0.002 14.998 15.042 integrate_v_rspace 127 11.3 0.004 0.004 14.938 14.985 init_scf_run 11 5.9 0.000 0.001 11.081 11.081 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.081 11.081 multiply_cannon_sync_h2d 9940 15.5 10.705 11.064 10.705 11.064 make_images_data 4970 15.5 0.055 0.065 8.986 10.523 hybrid_alltoall_any 5155 16.4 0.856 3.746 8.655 10.047 pw_transfer 1535 11.6 0.084 0.092 9.908 9.945 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 9.687 9.729 cp_fm_cholesky_decompose 22 10.9 9.569 9.695 9.569 9.695 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 9.079 9.132 fft_wrap_pw1pw2_140 519 13.2 1.792 1.829 8.356 8.406 grid_integrate_task_list 127 12.3 7.769 8.066 7.769 8.066 density_rs2pw 127 9.7 0.006 0.006 7.621 8.031 multiply_cannon_metrocomm1 9940 15.5 0.030 0.031 5.247 7.997 dbcsr_mm_accdrv_process 20590 16.1 3.301 4.505 7.242 7.940 wfi_extrapolate 11 7.9 0.001 0.001 7.900 7.900 fft3d_ps 1281 14.7 2.752 2.825 6.437 6.472 dbcsr_complete_redistribute 393 12.7 2.104 2.153 5.874 6.341 calculate_dm_sparse 127 9.5 0.001 0.001 6.218 6.326 grid_collocate_task_list 127 9.7 5.524 5.824 5.524 5.824 mp_alltoall_d11v 2401 14.1 5.474 5.699 5.474 5.699 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.307 5.319 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.282 5.282 mp_allgather_i34 2485 14.5 1.631 5.122 1.631 5.122 potential_pw2rs 127 12.3 0.026 0.026 5.001 5.019 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.787 4.181 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.079 4.133 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.999 4.072 multiply_cannon_metrocomm4 7455 15.5 0.026 0.028 1.800 4.043 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.945 4.006 mp_irecv_dv 28618 15.9 1.762 3.959 1.762 3.959 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.583 3.913 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.762 3.838 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=190.251000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1452.090909, yerr=14.475399 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 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 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 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 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 94.243758E+12 0.0% 0.0% 100.0% flops max/rank 11.787404E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547744 0.0% 0.0% 100.0% number of processed stacks 1982128 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.0 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 3.184288E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101360 MPI messages size (bytes): total size 1.145137E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.297725E+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 45848 35601252352 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58558. MP_Allreduce 11184 1500. MP_Sync 88 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 0.037 0.062 312.629 312.633 qs_mol_dyn_low 1 2.0 0.003 0.003 312.016 312.046 qs_forces 11 3.9 0.004 0.005 311.620 311.627 qs_energies 11 4.9 0.006 0.006 302.568 302.583 scf_env_do_scf 11 5.9 0.001 0.001 279.019 279.025 velocity_verlet 10 3.0 0.070 0.079 225.557 225.598 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 147.013 147.015 init_scf_loop 11 6.9 0.000 0.000 131.708 131.710 prepare_preconditioner 11 7.9 0.000 0.000 126.650 126.681 make_preconditioner 11 8.9 0.000 0.000 126.650 126.681 make_full_inverse_cholesky 11 9.9 0.038 0.039 102.126 123.666 qs_scf_new_mos 118 7.6 0.001 0.001 99.920 100.044 qs_scf_loop_do_ot 118 8.6 0.001 0.001 99.919 100.043 ot_scf_mini 118 9.6 0.004 0.004 94.889 94.968 dbcsr_multiply_generic 2534 12.6 0.213 0.224 85.968 86.101 cp_fm_upper_to_full 106 14.8 52.903 76.444 52.903 76.444 multiply_cannon 2534 13.6 0.692 0.728 60.316 62.905 multiply_cannon_loop 2534 14.6 1.071 1.090 56.210 56.337 ot_mini 118 10.6 0.001 0.001 47.140 47.266 dbcsr_complete_redistribute 397 12.7 4.009 4.049 30.978 44.152 copy_fm_to_dbcsr 210 11.7 0.001 0.002 27.218 40.367 cp_fm_cholesky_invert 11 10.9 38.256 38.262 38.256 38.262 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 24.481 37.499 rebuild_ks_matrix 129 8.3 0.001 0.001 36.411 36.591 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 36.411 36.591 mp_alltoall_i22 720 14.1 22.168 35.242 22.168 35.242 mp_waitall_1 104780 16.8 30.467 34.144 30.467 34.144 qs_ks_update_qs_env 129 7.6 0.001 0.001 33.766 33.939 qs_ot_get_p 129 10.4 0.001 0.001 31.876 32.011 qs_ot_p2m_diag 84 11.4 0.890 0.895 27.254 27.266 qs_ot_get_derivative 118 11.6 0.002 0.002 26.976 27.058 cp_dbcsr_syevd 84 12.4 0.006 0.006 25.307 25.309 make_m2s 5068 13.6 0.077 0.078 20.372 21.760 cp_fm_diag_elpa 84 13.4 0.013 0.026 21.642 21.643 cp_fm_diag_elpa_base 84 14.4 16.935 18.692 21.612 21.612 make_images 5068 14.6 3.113 3.310 19.887 21.272 ot_diis_step 118 11.6 0.022 0.023 20.085 20.085 multiply_cannon_metrocomm3 10136 15.6 0.025 0.027 19.727 19.848 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.153 19.491 apply_single 129 13.6 0.001 0.001 19.153 19.491 multiply_cannon_multrec 10136 15.6 10.473 12.356 18.334 18.396 qs_rho_update_rho_low 129 7.7 0.001 0.001 18.200 18.215 calculate_rho_elec 129 8.7 0.484 0.485 18.200 18.214 sum_up_and_integrate 129 10.3 0.002 0.002 17.802 17.887 integrate_v_rspace 129 11.3 0.004 0.004 17.739 17.824 multiply_cannon_sync_h2d 10136 15.6 14.403 14.419 14.403 14.419 hybrid_alltoall_any 5255 16.5 1.322 3.087 11.779 14.208 make_images_data 5068 15.6 0.065 0.070 11.715 14.112 pw_transfer 1559 11.6 0.096 0.097 13.143 13.162 init_scf_run 11 5.9 0.000 0.001 12.962 12.964 scf_env_initial_rho_setup 11 6.9 0.000 0.000 12.962 12.963 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 12.902 12.922 fft_wrap_pw1pw2_140 527 13.2 3.069 3.140 11.334 11.368 cp_fm_cholesky_decompose 22 10.9 10.897 10.920 10.897 10.920 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 10.789 10.830 wfi_extrapolate 11 7.9 0.001 0.001 9.755 9.756 dbcsr_mm_accdrv_process 20954 16.1 4.332 6.118 7.606 9.561 grid_integrate_task_list 129 12.3 8.642 8.871 8.642 8.871 density_rs2pw 129 9.7 0.005 0.006 8.765 8.820 mp_alltoall_d11v 2429 14.1 8.641 8.781 8.641 8.781 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.231 8.232 fft3d_ps 1301 14.7 2.887 2.946 8.112 8.158 calculate_dm_sparse 129 9.5 0.001 0.001 6.931 6.958 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.725 6.763 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.574 6.711 grid_collocate_task_list 129 9.7 6.478 6.575 6.478 6.575 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=312.633000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2867.909091, yerr=157.381098 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 1.259622E+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 0.033 0.066 85.062 85.063 qs_energies 1 2.0 0.000 0.000 84.552 84.561 ls_scf 1 3.0 0.000 0.000 83.655 83.664 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.553 72.762 multiply_cannon 111 7.7 0.017 0.020 55.895 56.970 multiply_cannon_loop 111 8.7 0.225 0.243 52.434 53.775 ls_scf_main 1 4.0 0.000 0.000 52.343 52.344 density_matrix_trs4 2 5.0 0.002 0.003 46.753 46.832 ls_scf_init_scf 1 4.0 0.000 0.000 28.260 28.261 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.079 27.134 mp_waitall_1 11031 10.9 22.152 25.381 22.152 25.381 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.977 24.997 multiply_cannon_multrec 2664 9.7 8.181 8.959 15.621 17.426 multiply_cannon_sync_h2d 2664 9.7 13.666 15.684 13.666 15.684 make_m2s 222 7.7 0.008 0.011 13.037 13.514 make_images 222 8.7 0.098 0.108 13.015 13.494 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.502 11.806 make_images_data 222 9.7 0.004 0.005 7.594 8.172 dbcsr_mm_accdrv_process 4760 10.4 0.587 0.691 7.057 8.034 hybrid_alltoall_any 227 10.6 0.216 1.842 6.499 7.786 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.388 7.213 dbcsr_mm_accdrv_process_sort 4760 11.4 6.271 7.182 6.271 7.182 calculate_norms 4752 9.8 5.542 6.256 5.542 6.256 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.048 5.219 mp_sum_l 887 5.1 3.079 5.142 3.079 5.142 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.050 4.035 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.280 4.032 mp_irecv_dv 6231 10.9 2.032 4.007 2.032 4.007 make_images_sizes 222 9.7 0.000 0.000 0.775 3.782 mp_alltoall_i44 222 10.7 0.775 3.782 0.775 3.782 arnoldi_extremal 4 6.8 0.000 0.000 3.223 3.254 arnoldi_normal_ev 4 7.8 0.001 0.002 3.223 3.254 build_subspace 16 8.4 0.009 0.012 3.128 3.131 ls_scf_post 1 4.0 0.000 0.000 3.052 3.061 ls_scf_store_result 1 5.0 0.000 0.000 2.865 2.903 dbcsr_special_finalize 555 9.7 0.005 0.006 2.315 2.900 dbcsr_merge_single_wm 555 10.7 0.466 0.584 2.307 2.891 make_images_pack 222 9.7 2.207 2.624 2.209 2.626 dbcsr_sort_data 658 11.4 2.099 2.599 2.099 2.599 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.320 2.559 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.460 2.072 2.461 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.321 2.411 buffer_matrices_ensure_size 222 8.7 1.763 2.059 1.763 2.059 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.817 1.820 rebuild_ks_matrix 3 7.3 0.000 0.000 1.807 1.811 qs_ks_build_kohn_sham_matrix 3 8.3 0.026 0.052 1.807 1.811 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.063000, 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/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 2.148168E+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 0.080 0.238 90.507 90.508 qs_energies 1 2.0 0.001 0.010 89.773 89.821 ls_scf 1 3.0 0.000 0.001 88.427 88.476 dbcsr_multiply_generic 111 6.7 0.016 0.022 74.060 74.376 multiply_cannon 111 7.7 0.027 0.039 52.641 56.761 ls_scf_main 1 4.0 0.001 0.017 54.334 54.340 multiply_cannon_loop 111 8.7 0.135 0.149 49.900 53.111 density_matrix_trs4 2 5.0 0.003 0.017 48.567 48.790 ls_scf_init_scf 1 4.0 0.000 0.002 30.390 30.406 mp_waitall_1 9105 10.9 20.640 30.264 20.640 30.264 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.807 28.895 multiply_cannon_multrec 1332 9.7 13.327 17.424 22.675 27.934 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.380 26.393 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.336 20.854 make_m2s 222 7.7 0.006 0.008 15.066 15.615 make_images 222 8.7 1.373 1.701 15.035 15.585 dbcsr_mm_accdrv_process 4041 10.4 0.330 0.559 8.944 10.527 dbcsr_mm_accdrv_process_sort 4041 11.4 8.455 9.998 8.455 9.998 make_images_data 222 9.7 0.004 0.005 8.748 9.598 hybrid_alltoall_any 227 10.6 0.543 2.547 8.167 9.490 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.188 8.006 mp_sum_l 887 5.1 4.969 7.968 4.969 7.968 mp_irecv_dv 3311 11.0 3.168 7.953 3.168 7.953 calculate_norms 2376 9.8 6.073 6.876 6.073 6.876 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.655 6.222 multiply_cannon_sync_h2d 1332 9.7 4.900 6.165 4.900 6.165 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.991 5.201 arnoldi_extremal 4 6.8 0.000 0.000 4.676 4.695 arnoldi_normal_ev 4 7.8 0.001 0.005 4.676 4.695 build_subspace 16 8.4 0.014 0.021 4.414 4.417 ls_scf_post 1 4.0 0.001 0.010 3.703 3.752 ls_scf_store_result 1 5.0 0.000 0.000 3.411 3.516 dbcsr_matrix_vector_mult 304 9.0 0.010 0.023 3.153 3.401 dbcsr_matrix_vector_mult_local 304 10.0 2.749 3.233 2.751 3.235 mp_allgather_i34 111 8.7 0.731 2.919 0.731 2.919 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.203 2.730 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.605 2.714 dbcsr_data_new 4174 10.1 2.120 2.417 2.120 2.417 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.141 2.158 rebuild_ks_matrix 3 7.3 0.000 0.000 2.128 2.146 qs_ks_build_kohn_sham_matrix 3 8.3 0.008 0.024 2.128 2.146 make_images_pack 222 9.7 1.820 2.107 1.823 2.109 dbcsr_sort_data 436 11.2 1.837 2.078 1.837 2.078 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.508000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1774.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 2.898190E+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 0.033 0.045 91.700 91.701 qs_energies 1 2.0 0.000 0.000 91.116 91.120 ls_scf 1 3.0 0.000 0.001 89.711 89.715 dbcsr_multiply_generic 111 6.7 0.016 0.016 74.356 74.655 ls_scf_main 1 4.0 0.000 0.001 55.822 55.827 multiply_cannon 111 7.7 0.031 0.069 51.631 55.728 multiply_cannon_loop 111 8.7 0.116 0.131 48.927 52.881 density_matrix_trs4 2 5.0 0.002 0.004 49.935 50.056 mp_waitall_1 7281 11.0 22.999 32.922 22.999 32.922 ls_scf_init_scf 1 4.0 0.000 0.001 30.170 30.173 ls_scf_init_matrix_S 1 5.0 0.000 0.002 28.888 28.976 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.539 26.551 multiply_cannon_multrec 888 9.7 12.683 15.513 21.322 24.779 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.637 22.475 make_m2s 222 7.7 0.006 0.007 16.073 16.799 make_images 222 8.7 1.587 1.857 16.035 16.760 make_images_data 222 9.7 0.004 0.004 9.364 10.468 hybrid_alltoall_any 227 10.6 0.640 2.958 9.016 10.110 dbcsr_mm_accdrv_process 3754 10.4 0.297 0.492 8.152 9.406 dbcsr_mm_accdrv_process_sort 3754 11.4 7.719 8.913 7.719 8.913 mp_sum_l 887 5.1 4.988 8.415 4.988 8.415 multiply_cannon_sync_h2d 888 9.7 6.013 7.129 6.013 7.129 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.447 7.102 mp_irecv_dv 2335 11.1 2.432 7.044 2.432 7.044 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.641 6.973 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.788 6.572 arnoldi_extremal 4 6.8 0.000 0.000 5.078 5.093 arnoldi_normal_ev 4 7.8 0.001 0.005 5.078 5.093 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.735 4.962 calculate_norms 1584 9.8 4.395 4.846 4.395 4.846 build_subspace 16 8.4 0.014 0.020 4.771 4.776 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.440 3.771 ls_scf_post 1 4.0 0.000 0.001 3.719 3.723 mp_allgather_i34 111 8.7 0.823 3.610 0.823 3.610 dbcsr_matrix_vector_mult_local 304 10.0 3.027 3.605 3.029 3.607 ls_scf_store_result 1 5.0 0.000 0.000 3.467 3.545 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.861 2.964 dbcsr_data_new 4116 9.9 2.103 2.436 2.103 2.436 dbcsr_sort_data 325 11.1 1.875 2.061 1.875 2.061 make_images_sizes 222 9.7 0.000 0.000 0.871 1.999 mp_alltoall_i44 222 10.7 0.871 1.999 0.871 1.999 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.886 1.888 make_images_pack 222 9.7 1.624 1.880 1.627 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 0.001 0.001 1.868 1.870 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=91.701000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2259.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 3.330462E+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 0.052 0.128 98.541 98.543 qs_energies 1 2.0 0.000 0.000 97.710 97.750 ls_scf 1 3.0 0.002 0.018 96.036 96.045 dbcsr_multiply_generic 111 6.7 0.018 0.021 79.417 79.669 ls_scf_main 1 4.0 0.003 0.023 59.692 59.693 multiply_cannon 111 7.7 0.048 0.130 52.061 56.563 density_matrix_trs4 2 5.0 0.003 0.027 53.508 53.635 multiply_cannon_loop 111 8.7 0.154 0.167 47.007 50.630 ls_scf_init_scf 1 4.0 0.000 0.002 33.061 33.063 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.803 31.871 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.265 29.280 mp_waitall_1 6369 11.0 23.068 29.176 23.068 29.176 multiply_cannon_multrec 1332 9.7 14.148 17.155 22.103 24.892 make_m2s 222 7.7 0.007 0.008 21.308 22.758 make_images 222 8.7 3.157 3.618 21.257 22.710 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.433 17.820 make_images_data 222 9.7 0.004 0.005 11.869 13.377 hybrid_alltoall_any 227 10.6 0.799 3.767 11.267 12.988 dbcsr_mm_accdrv_process 3641 10.4 0.301 0.477 7.596 9.130 dbcsr_mm_accdrv_process_sort 3641 11.4 7.157 8.640 7.157 8.640 mp_sum_l 887 5.1 4.498 8.399 4.498 8.399 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.465 6.633 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.105 6.077 multiply_cannon_sync_h2d 1332 9.7 5.461 6.070 5.461 6.070 mp_irecv_dv 3229 10.9 2.078 6.014 2.078 6.014 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.595 5.623 arnoldi_extremal 4 6.8 0.000 0.000 5.407 5.419 arnoldi_normal_ev 4 7.8 0.001 0.005 5.406 5.419 build_subspace 16 8.4 0.014 0.021 5.070 5.078 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.579 4.733 calculate_norms 2376 9.8 4.173 4.522 4.173 4.522 mp_allgather_i34 111 8.7 2.167 4.484 2.167 4.484 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.666 3.966 dbcsr_matrix_vector_mult_local 304 10.0 3.201 3.702 3.203 3.704 dbcsr_sort_data 658 11.4 3.099 3.467 3.099 3.467 ls_scf_post 1 4.0 0.000 0.001 3.281 3.289 dbcsr_special_finalize 555 9.7 0.006 0.007 2.841 3.253 dbcsr_merge_single_wm 555 10.7 0.540 0.669 2.832 3.244 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.028 3.093 ls_scf_store_result 1 5.0 0.000 0.000 3.021 3.086 dbcsr_data_release 10477 10.7 1.583 2.384 1.583 2.384 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.543000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2723.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 4.732981E+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 0.062 0.073 92.011 92.012 qs_energies 1 2.0 0.000 0.000 91.193 91.198 ls_scf 1 3.0 0.000 0.000 89.231 89.236 dbcsr_multiply_generic 111 6.7 0.017 0.019 70.408 70.588 ls_scf_main 1 4.0 0.000 0.000 56.324 56.325 multiply_cannon 111 7.7 0.082 0.138 51.856 54.744 multiply_cannon_loop 111 8.7 0.088 0.093 49.305 50.737 density_matrix_trs4 2 5.0 0.002 0.003 49.265 49.354 ls_scf_init_scf 1 4.0 0.000 0.000 29.491 29.492 mp_waitall_1 5436 11.0 23.793 28.415 23.793 28.415 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.214 28.245 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.117 26.132 multiply_cannon_multrec 444 9.7 13.662 16.323 20.805 23.078 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 5.925 14.818 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.325 14.811 make_m2s 222 7.7 0.005 0.005 13.724 14.637 make_images 222 8.7 2.045 2.479 13.656 14.566 hybrid_alltoall_any 227 10.6 0.801 3.828 8.132 9.820 make_images_data 222 9.7 0.003 0.004 8.353 9.621 multiply_cannon_sync_h2d 444 9.7 6.744 8.266 6.744 8.266 dbcsr_mm_accdrv_process 3003 10.4 0.371 0.438 6.835 8.028 dbcsr_mm_accdrv_process_sort 3003 11.4 6.464 7.623 6.464 7.623 arnoldi_extremal 4 6.8 0.000 0.000 6.114 6.130 arnoldi_normal_ev 4 7.8 0.001 0.005 6.114 6.130 build_subspace 16 8.4 0.015 0.019 5.720 5.730 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.484 4.720 mp_sum_l 887 5.1 2.833 4.651 2.833 4.651 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.313 4.496 dbcsr_matrix_vector_mult_local 304 10.0 3.706 4.183 3.708 4.185 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.513 3.852 mp_irecv_dv 1241 11.2 1.501 3.821 1.501 3.821 calculate_norms 792 9.8 3.624 3.745 3.624 3.745 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.567 3.656 mp_allgather_i34 111 8.7 1.121 3.599 1.121 3.599 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.024 3.585 make_images_sizes 222 9.7 0.000 0.000 0.974 3.489 mp_alltoall_i44 222 10.7 0.974 3.489 0.974 3.489 ls_scf_post 1 4.0 0.000 0.000 3.415 3.421 ls_scf_store_result 1 5.0 0.000 0.000 3.206 3.239 dbcsr_data_new 4608 9.7 1.804 2.336 1.804 2.336 dbcsr_finalize 304 7.8 0.062 0.077 2.197 2.297 dbcsr_merge_all 275 8.9 0.480 0.531 2.056 2.138 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.059 2.059 rebuild_ks_matrix 3 7.3 0.000 0.000 2.025 2.026 qs_ks_build_kohn_sham_matrix 3 8.3 0.009 0.014 2.025 2.026 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 1.946 1.949 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.012000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3744.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 8.879022E+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 0.081 0.119 105.160 105.161 qs_energies 1 2.0 0.000 0.000 103.760 103.764 ls_scf 1 3.0 0.000 0.000 100.835 100.839 dbcsr_multiply_generic 111 6.7 0.024 0.027 74.442 74.603 ls_scf_main 1 4.0 0.000 0.000 63.380 63.381 density_matrix_trs4 2 5.0 0.002 0.003 54.452 54.519 multiply_cannon 111 7.7 0.131 0.226 48.381 50.306 multiply_cannon_loop 111 8.7 0.098 0.100 45.341 45.982 ls_scf_init_scf 1 4.0 0.000 0.000 33.649 33.650 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.092 32.121 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.349 29.359 mp_waitall_1 4527 11.1 21.465 25.236 21.465 25.236 make_m2s 222 7.7 0.005 0.005 22.395 23.476 make_images 222 8.7 3.583 3.890 22.288 23.368 multiply_cannon_multrec 444 9.7 17.809 18.544 22.492 23.248 hybrid_alltoall_any 227 10.6 1.654 3.606 12.560 15.332 make_images_data 222 9.7 0.003 0.004 12.758 14.917 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.110 10.459 multiply_cannon_sync_h2d 444 9.7 8.801 8.849 8.801 8.849 arnoldi_extremal 4 6.8 0.000 0.000 7.417 7.431 arnoldi_normal_ev 4 7.8 0.003 0.009 7.417 7.431 build_subspace 16 8.4 0.026 0.036 6.865 6.874 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.515 5.659 dbcsr_matrix_vector_mult_local 304 10.0 5.084 5.363 5.087 5.366 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.067 5.163 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.868 5.120 dbcsr_mm_accdrv_process 1814 10.4 0.271 0.353 4.497 4.623 dbcsr_mm_accdrv_process_sort 1814 11.4 4.155 4.291 4.155 4.291 ls_scf_post 1 4.0 0.000 0.000 3.806 3.809 make_images_sizes 222 9.7 0.000 0.000 1.475 3.607 mp_alltoall_i44 222 10.7 1.474 3.607 1.474 3.607 ls_scf_store_result 1 5.0 0.000 0.000 3.519 3.558 mp_allgather_i34 111 8.7 1.137 3.533 1.137 3.533 calculate_norms 792 9.8 3.224 3.265 3.224 3.265 dbcsr_finalize 304 7.8 0.082 0.089 3.083 3.180 dbcsr_merge_all 275 8.9 0.883 0.916 2.859 2.949 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.895 2.895 dbcsr_complete_redistribute 5 7.6 1.435 1.471 2.754 2.876 matrix_ls_to_qs 2 6.0 0.000 0.000 2.387 2.516 dbcsr_sort_data 325 11.1 2.441 2.505 2.441 2.505 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.442 2.444 dbcsr_data_new 6591 9.6 1.877 2.377 1.877 2.377 rebuild_ks_matrix 3 7.3 0.000 0.000 2.374 2.376 qs_ks_build_kohn_sham_matrix 3 8.3 0.028 0.054 2.374 2.376 dbcsr_new_transposed 4 7.5 0.252 0.272 2.287 2.302 dbcsr_frobenius_norm 74 6.6 2.057 2.128 2.201 2.238 dbcsr_add_d 103 6.2 0.000 0.000 2.152 2.234 dbcsr_add_anytype 103 7.2 0.860 0.890 2.152 2.234 dbcsr_data_release 12724 10.6 1.990 2.163 1.990 2.163 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.161000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7026.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ad1d17f038befc10386af0239c8c9a26c496fd9e_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 588.529664E+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 894229043. 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 0.057 0.122 265.973 265.976 qs_mol_dyn_low 1 2.0 0.004 0.036 264.825 264.871 qs_forces 5 3.8 0.005 0.041 264.685 264.690 qs_energies 5 4.8 0.005 0.090 261.405 261.447 scf_env_do_scf 5 5.8 0.000 0.003 247.234 247.239 scf_env_do_scf_inner_loop 105 6.6 0.002 0.013 214.194 214.202 qs_scf_new_mos 105 7.6 0.000 0.001 167.002 167.257 qs_scf_loop_do_ot 105 8.6 0.001 0.001 167.002 167.257 ot_scf_mini 105 9.6 0.003 0.006 156.800 157.113 dbcsr_multiply_generic 1445 12.2 0.125 0.132 136.804 137.650 multiply_cannon 1445 13.2 0.277 0.288 114.924 117.502 velocity_verlet 4 3.0 0.001 0.011 116.231 116.232 multiply_cannon_loop 1445 14.2 2.834 2.995 112.530 115.507 qs_ot_get_p 112 10.4 0.001 0.001 62.789 63.236 ot_mini 105 10.6 0.001 0.007 62.934 63.070 qs_ot_p2m_diag 40 11.0 0.021 0.040 50.464 50.586 mp_waitall_1 488190 16.1 38.450 48.835 38.450 48.835 cp_dbcsr_syevd 40 12.0 0.002 0.002 46.772 46.774 multiply_cannon_multrec 69360 15.2 29.502 34.340 39.490 44.048 qs_ot_get_derivative 55 11.6 0.001 0.001 40.914 41.128 cp_fm_syevd 40 13.0 0.000 0.002 40.498 40.666 multiply_cannon_metrocomm3 69360 15.2 0.202 0.214 26.814 37.712 cp_fm_redistribute_end 40 14.0 17.638 35.171 17.647 35.176 cp_fm_syevd_base 40 14.0 17.511 35.053 17.511 35.053 multiply_cannon_sync_h2d 69360 15.2 28.943 33.059 28.943 33.059 init_scf_loop 7 6.6 0.001 0.029 32.994 32.997 rebuild_ks_matrix 110 8.4 0.000 0.000 31.970 32.162 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.031 31.969 32.161 qs_ks_update_qs_env 112 7.6 0.001 0.001 29.446 29.625 prepare_preconditioner 7 7.6 0.000 0.000 27.410 27.445 make_preconditioner 7 8.6 0.001 0.056 27.410 27.445 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.212 23.481 apply_single 62 13.6 0.000 0.000 23.212 23.481 qs_rho_update_rho_low 110 7.6 0.001 0.015 21.824 22.248 calculate_rho_elec 110 8.6 0.030 0.032 21.823 22.247 ot_new_cg_direction 55 11.6 0.001 0.004 21.193 21.193 make_full_inverse_cholesky 7 9.6 0.000 0.008 19.091 19.150 density_rs2pw 110 9.6 0.005 0.007 15.666 16.132 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.694 15.844 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.629 14.835 mp_sum_l 4764 12.2 13.409 14.459 13.409 14.459 cp_fm_cholesky_invert 7 10.6 13.254 13.278 13.254 13.278 pw_transfer 1645 12.4 0.080 0.102 12.858 13.125 fft_wrap_pw1pw2 1425 13.5 0.012 0.016 12.719 12.990 transfer_rs2pw 445 10.6 0.007 0.008 11.857 12.404 calculate_dm_sparse 110 9.5 0.000 0.001 11.869 12.063 init_scf_run 5 5.8 0.000 0.001 11.274 11.275 scf_env_initial_rho_setup 5 6.8 0.000 0.002 11.273 11.274 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 11.130 11.220 fft_wrap_pw1pw2_240 915 15.0 1.157 1.250 10.972 11.205 qs_vxc_create 110 10.4 0.002 0.016 11.000 11.062 check_diag 80 13.5 8.688 8.970 10.803 10.948 sum_up_and_integrate 60 10.3 0.001 0.011 10.808 10.818 integrate_v_rspace 60 11.3 0.002 0.013 10.790 10.801 dbcsr_mm_accdrv_process 154766 15.8 6.126 6.339 9.857 10.695 fft3d_pb 915 16.0 2.394 2.676 9.149 9.501 make_m2s 2890 13.2 0.079 0.086 8.041 8.621 make_images 2890 14.2 0.239 0.259 7.934 8.514 multiply_cannon_metrocomm1 69360 15.2 0.095 0.103 5.002 8.365 make_full_single_inverse 7 9.6 0.001 0.002 7.974 8.009 xc_rho_set_and_dset_create 110 12.4 0.078 0.146 7.570 7.841 xc_vxc_pw_create 60 11.3 0.039 0.051 7.384 7.446 potential_pw2rs 60 12.3 0.003 0.003 6.750 6.804 xc_pw_derive 510 13.4 0.006 0.007 6.656 6.719 acc_transpose_blocks 69360 15.2 0.355 0.372 6.338 6.656 mp_alltoall_z22v 2340 17.7 6.225 6.649 6.225 6.649 transfer_rs2pw_30 110 11.6 1.281 1.340 5.787 6.475 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 6.319 6.336 mp_sendrecv_dv 168740 12.6 5.847 6.098 5.847 6.098 calculate_first_density_matrix 1 7.0 0.001 0.018 6.016 6.050 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 5.686 5.760 transfer_pw2rs 245 13.2 0.003 0.004 5.545 5.592 mp_alltoall_d11v 1300 13.8 4.369 5.357 4.369 5.357 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=265.976000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: ad1d17f038befc10386af0239c8c9a26c496fd9e Summary: empty Status: OK