=== This is the CP2K Performance-Test === Updating 4c9fa1087..0f4704847 Fast-forward .../dft/{hartree-fock.md => hartree-fock/admm.md} | 2 +- docs/methods/dft/hartree-fock/index.md | 11 + docs/methods/dft/hartree-fock/ri_gamma.md | 282 ++++++++++++ docs/methods/dft/hartree-fock/ri_kpoints.md | 236 ++++++++++ docs/methods/dft/index.md | 3 +- docs/methods/dft/ri_hfx.md | 511 --------------------- docs/methods/properties/index.md | 2 +- .../properties/{ => optical}/bethe-salpeter.md | 0 .../properties/{optical.md => optical/index.md} | 2 +- .../{time-dependent-dft.md => optical/tddft.md} | 0 docs/methods/sampling/newton-x.md | 6 +- 11 files changed, 536 insertions(+), 519 deletions(-) rename docs/methods/dft/{hartree-fock.md => hartree-fock/admm.md} (87%) create mode 100644 docs/methods/dft/hartree-fock/index.md create mode 100644 docs/methods/dft/hartree-fock/ri_gamma.md create mode 100644 docs/methods/dft/hartree-fock/ri_kpoints.md delete mode 100644 docs/methods/dft/ri_hfx.md rename docs/methods/properties/{ => optical}/bethe-salpeter.md (100%) rename docs/methods/properties/{optical.md => optical/index.md} (88%) rename docs/methods/properties/{time-dependent-dft.md => optical/tddft.md} (100%) Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 0f470484747789dc0da415bf6a25bd77b7b21d8e ################# 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 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SPGLIB 1.16.2, LIBGRPP 20231215 # # 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: 25.01.2024 # # \ 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-hdf5 --with-libvdwxc --with-pexsi --with-plumed --with-sirius=no; \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 #USE_DEEPMD := 2.2.7 #USE_SIRIUS := 7.5.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS #LIBVDWXC_VER := 0.4.0 #SPFFT_VER := 1.0.6 #SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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_DEEPMD),) USE_DEEPMD := $(strip $(USE_DEEPMD)) DEEPMD_INC := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/include DEEPMD_LIB := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/lib CFLAGS += -I$(DEEPMD_INC) DFLAGS += -D__DEEPMD LIBS += $(DEEPMD_LIB)/libdeepmd.so LIBS += $(DEEPMD_LIB)/libdeepmd_c.so LIBS += $(DEEPMD_LIB)/libdeepmd_cc.so LIBS += $(DEEPMD_LIB)/libdeepmd_dyn_cudart.so LIBS += $(DEEPMD_LIB)/libdeepmd_op.so LIBS += $(DEEPMD_LIB)/libdeepmd_op_cuda.so LIBS += $(DEEPMD_LIB)/libtensorflow_cc.so.2 LIBS += $(DEEPMD_LIB)/libtensorflow_framework.so.2 endif ifneq ($(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_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(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/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/01 job id: 51918905 --- 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/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/02 job id: 51918906 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/03 job id: 51918907 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/04 job id: 51918908 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/05 job id: 51918909 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/06 job id: 51918910 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/07 job id: 51918911 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/08 job id: 51918912 --- 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/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/09 job id: 51918913 --- 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/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/10 job id: 51918914 --- 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/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/11 job id: 51918915 --- 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/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/12 job id: 51918916 --- 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/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/13 job id: 51918917 --- 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/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/14 job id: 51918918 --- 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/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/15 job id: 51918919 --- 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/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/16 job id: 51918920 --- 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/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/17 job id: 51918921 --- 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/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/18 job id: 51918922 --- 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/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/19 job id: 51918923 --- 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/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/20 job id: 51918924 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/21 job id: 51918925 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/22 job id: 51918926 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/23 job id: 51918927 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/24 job id: 51918928 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/25 job id: 51918929 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/26 job id: 51918930 --- 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] = 20 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/27 job id: 51918931 --- 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/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 172669. 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 0.053 0.054 139.750 139.751 farming_run 1 2.0 138.690 138.699 139.687 139.688 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.533813E+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 0.005 0.026 120.082 120.082 qs_energies 1 2.0 0.001 0.001 119.893 119.897 mp2_main 1 3.0 0.000 0.000 117.845 117.849 mp2_gpw_main 1 4.0 0.037 0.046 116.850 116.854 mp2_ri_gpw_compute_in 1 5.0 0.174 0.176 97.959 98.186 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.004 57.406 57.633 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.170 43.854 48.839 get_2c_integrals 1 6.0 0.008 0.009 39.438 40.379 integrate_v_rspace 273 8.0 0.439 0.454 25.587 30.095 fft_wrap_pw1pw2 5465 10.4 0.067 0.070 28.030 28.598 grid_integrate_task_list 273 9.0 20.842 26.284 20.842 26.284 fft_wrap_pw1pw2_100 2178 11.4 1.294 1.430 25.481 26.037 compute_2c_integrals 1 7.0 0.002 0.002 21.334 21.334 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 20.828 21.072 mp2_eri_2c_integrate_gpw 1 9.0 2.394 2.423 20.826 21.071 rpa_ri_compute_en 1 5.0 0.020 0.029 18.776 19.107 cp_fm_cholesky_decompose 12 8.2 18.030 18.956 18.030 18.956 cholesky_decomp 1 7.0 0.000 0.000 16.945 17.877 fft3d_s 5443 12.4 16.913 17.292 16.935 17.314 ao_to_mo_and_store_B_mult_1 272 7.0 10.768 15.319 10.768 15.319 calculate_wavefunction 272 8.0 5.461 5.601 12.909 13.828 calc_potential_gpw 544 9.5 0.004 0.004 10.840 11.645 rpa_num_int 1 6.0 0.000 0.003 10.632 10.633 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.585 10.587 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 9.787 10.505 potential_pw2rs 545 10.0 0.106 0.108 8.504 10.353 calc_mat_Q 8 8.0 0.000 0.000 9.458 9.540 contract_S_to_Q 8 9.0 0.000 0.000 8.882 8.963 collocate_single_gaussian 272 10.0 0.040 0.042 7.836 8.624 parallel_gemm_fm 14 9.1 0.000 0.000 8.465 8.573 parallel_gemm_fm_cosma 14 10.1 8.465 8.573 8.465 8.573 create_integ_mat 1 6.0 0.022 0.029 7.819 7.820 array2fm 1 7.0 0.000 0.000 6.815 7.232 pw_gather_s 2722 12.2 3.946 4.949 3.946 4.949 pw_poisson_solve 545 10.5 0.010 0.011 4.676 4.898 pw_scatter_s 2720 12.7 4.431 4.844 4.431 4.844 pw_poisson_set 548 11.5 0.017 0.018 3.363 3.571 array2fm_buffer_send 1 8.0 3.024 3.176 3.024 3.176 pw_copy 4911 11.6 2.632 2.740 2.632 2.740 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=116.853993, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2777.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 200775. 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 0.026 0.035 532.484 532.486 farming_run 1 2.0 529.543 529.551 532.442 532.445 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.272066E+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 21970578. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 325.638 325.638 qs_energies 1 2.0 0.001 0.008 325.428 325.428 mp2_main 1 3.0 0.001 0.012 209.671 209.672 mp2_gpw_main 1 4.0 0.048 0.084 208.732 208.739 mp2_ri_gpw_compute_en 1 5.0 0.038 0.042 128.508 130.016 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.776 1.975 126.834 126.839 scf_env_do_scf 1 3.0 0.000 0.000 115.470 115.470 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.565 114.572 rebuild_ks_matrix 4 6.0 0.000 0.000 114.564 114.571 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.060 114.564 114.571 hfx_ks_matrix 4 8.0 0.001 0.001 114.211 114.214 integrate_four_center 4 9.0 0.153 0.470 114.210 114.213 mp2_ri_gpw_compute_en_expansio 172 7.0 0.464 0.499 111.051 112.081 local_gemm 172 8.0 110.588 111.613 110.588 111.613 integrate_four_center_main 4 10.0 0.126 0.716 102.133 105.930 integrate_four_center_bin 261 11.0 102.006 105.352 102.006 105.352 init_scf_loop 1 4.0 0.000 0.000 97.472 97.473 mp2_ri_gpw_compute_in 1 5.0 0.072 0.128 80.050 81.102 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.006 56.755 57.809 mp2_eri_3c_integrate_gpw 91 7.0 0.146 0.161 44.593 49.315 integrate_v_rspace 95 8.0 0.397 0.566 28.228 32.765 fft_wrap_pw1pw2 1868 10.4 0.029 0.035 30.418 31.170 ao_to_mo_and_store_B_mult_1 91 7.0 10.489 29.118 10.489 29.118 fft_wrap_pw1pw2_100 730 11.4 0.633 0.763 28.115 28.857 grid_integrate_task_list 95 9.0 23.520 28.287 23.520 28.287 get_2c_integrals 1 6.0 0.001 0.011 23.181 23.223 compute_2c_integrals 1 7.0 0.004 0.015 22.161 22.172 compute_2c_integrals_loop_lm 1 8.0 0.002 0.012 21.679 22.027 mp2_eri_2c_integrate_gpw 1 9.0 1.728 1.816 21.676 22.026 fft3d_s 1823 12.4 19.203 19.994 19.216 20.008 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.996 17.997 calc_potential_gpw 182 9.5 0.002 0.002 12.475 13.239 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.007 10.724 11.426 calculate_wavefunction 91 8.0 2.028 2.073 10.040 10.862 potential_pw2rs 186 10.0 0.033 0.034 8.961 10.409 mp2_ri_gpw_compute_en_comm 22 7.0 0.499 0.519 7.891 9.078 collocate_single_gaussian 91 10.0 0.017 0.021 8.238 8.707 mp_sendrecv_dm3 2068 8.0 5.929 7.130 5.929 7.130 integrate_four_center_load 4 10.0 0.000 0.000 6.770 6.774 hfx_load_balance 1 11.0 0.000 0.000 6.770 6.774 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=208.729878, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1525.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 447.672320E+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 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 0.217 0.246 58.777 58.780 qs_mol_dyn_low 1 2.0 0.044 0.155 57.499 57.507 qs_forces 11 3.9 0.003 0.008 57.374 57.416 qs_energies 11 4.9 0.046 0.060 55.788 55.843 scf_env_do_scf 11 5.9 0.000 0.001 47.044 47.044 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 44.907 44.908 dbcsr_multiply_generic 2286 12.5 0.094 0.099 36.094 36.672 qs_scf_new_mos 108 7.5 0.000 0.001 34.404 34.720 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.404 34.720 ot_scf_mini 108 9.5 0.002 0.002 32.695 32.903 multiply_cannon 2286 13.5 0.177 0.184 28.294 29.633 multiply_cannon_loop 2286 14.5 1.805 1.901 27.580 28.984 velocity_verlet 10 3.0 0.009 0.029 27.245 27.246 ot_mini 108 10.5 0.001 0.001 19.427 19.690 qs_ot_get_derivative 108 11.5 0.001 0.001 16.467 16.660 mp_waitall_1 245248 16.5 8.491 14.911 8.491 14.911 multiply_cannon_multrec 54864 15.5 3.644 5.712 9.781 13.197 multiply_cannon_metrocomm3 54864 15.5 0.072 0.079 5.920 13.090 qs_ot_get_p 119 10.4 0.001 0.001 8.324 8.615 rebuild_ks_matrix 119 8.3 0.000 0.000 8.131 8.337 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.022 8.130 8.337 dbcsr_mm_accdrv_process 76910 16.1 1.835 2.880 6.053 7.689 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.117 7.308 mp_sum_l 7287 12.8 5.687 7.299 5.687 7.299 init_scf_run 11 5.9 0.000 0.021 7.023 7.023 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.022 7.023 multiply_cannon_sync_h2d 54864 15.5 5.162 6.453 5.162 6.453 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.796 6.290 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.381 5.493 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.417 5.459 calculate_dm_sparse 119 9.5 0.000 0.000 5.133 5.288 jit_kernel_multiply 13 15.8 4.156 4.752 4.156 4.752 calculate_first_density_matrix 1 7.0 0.000 0.000 4.707 4.717 sum_up_and_integrate 119 10.3 0.001 0.002 4.647 4.660 integrate_v_rspace 119 11.3 0.002 0.002 4.636 4.650 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.594 4.595 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.431 4.584 calculate_rho_elec 119 8.7 0.011 0.016 4.430 4.584 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.358 4.359 cp_fm_redistribute_end 50 14.0 2.222 4.330 2.230 4.333 cp_fm_diag_elpa_base 50 14.0 2.097 4.211 2.101 4.221 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 3.586 3.613 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.917 3.122 apply_single 119 13.6 0.000 0.000 2.917 3.122 multiply_cannon_metrocomm1 54864 15.5 0.056 0.062 1.803 2.938 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.712 2.784 ot_diis_step 108 11.5 0.006 0.006 2.732 2.732 acc_transpose_blocks 54864 15.5 0.218 0.240 2.197 2.730 density_rs2pw 119 9.7 0.003 0.004 2.353 2.452 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.242 2.244 wfi_extrapolate 11 7.9 0.001 0.001 2.229 2.230 init_scf_loop 11 6.9 0.000 0.000 2.118 2.118 grid_integrate_task_list 119 12.3 2.018 2.111 2.018 2.111 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.925 1.987 potential_pw2rs 119 12.3 0.003 0.004 1.934 1.960 transfer_rs2pw 487 10.6 0.006 0.007 1.738 1.923 mp_sum_d 4139 12.0 1.321 1.882 1.321 1.882 make_m2s 4572 13.5 0.053 0.055 1.811 1.864 fft_wrap_pw1pw2 1201 11.6 0.011 0.013 1.705 1.793 make_images 4572 14.5 0.132 0.138 1.730 1.782 mp_alltoall_d11v 2130 13.8 1.543 1.715 1.543 1.715 mp_waitany 12084 13.8 1.474 1.687 1.474 1.687 fft3d_ps 1201 13.6 0.365 0.469 1.480 1.562 transfer_pw2rs 487 13.2 0.005 0.006 1.430 1.449 acc_transpose_blocks_sync 164592 16.5 1.196 1.436 1.196 1.436 fft_wrap_pw1pw2_140 487 12.2 0.048 0.053 1.333 1.422 grid_collocate_task_list 119 9.7 1.350 1.415 1.350 1.415 qs_energies_init_hamiltonians 11 5.9 0.013 0.016 1.174 1.218 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=58.780000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=426.454545, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 483.287040E+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 303364. 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 0.058 0.064 45.672 45.673 qs_mol_dyn_low 1 2.0 0.003 0.003 44.846 44.915 qs_forces 11 3.9 0.005 0.023 44.783 44.784 qs_energies 11 4.9 0.004 0.005 43.019 43.027 scf_env_do_scf 11 5.9 0.000 0.001 34.538 34.539 scf_env_do_scf_inner_loop 108 6.5 0.073 0.088 31.818 31.818 dbcsr_multiply_generic 2286 12.5 0.102 0.105 25.641 26.157 qs_scf_new_mos 108 7.5 0.001 0.001 22.399 22.613 qs_scf_loop_do_ot 108 8.5 0.001 0.001 22.399 22.613 ot_scf_mini 108 9.5 0.002 0.003 21.415 21.567 multiply_cannon 2286 13.5 0.208 0.214 19.630 21.550 multiply_cannon_loop 2286 14.5 1.182 1.239 18.347 20.512 velocity_verlet 10 3.0 0.002 0.002 19.919 19.920 ot_mini 108 10.5 0.001 0.001 13.087 13.312 multiply_cannon_multrec 27432 15.5 1.819 4.137 8.626 11.530 mp_waitall_1 200699 16.5 6.061 11.269 6.061 11.269 qs_ot_get_derivative 108 11.5 0.001 0.001 10.584 10.736 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.260 9.899 dbcsr_mm_accdrv_process 47894 16.0 4.118 8.100 6.729 8.943 rebuild_ks_matrix 119 8.3 0.000 0.000 7.246 7.424 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.246 7.424 init_scf_run 11 5.9 0.000 0.004 7.036 7.036 scf_env_initial_rho_setup 11 6.9 0.003 0.004 7.036 7.036 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.354 6.516 qs_ot_get_p 119 10.4 0.001 0.002 5.238 5.458 calculate_first_density_matrix 1 7.0 0.000 0.000 5.361 5.364 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.873 4.763 jit_kernel_multiply 10 16.3 2.551 4.580 2.551 4.580 calculate_dm_sparse 119 9.5 0.000 0.001 4.496 4.572 mp_sum_l 7287 12.8 2.576 4.528 2.576 4.528 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.133 4.104 apply_single 119 13.6 0.000 0.000 3.133 4.104 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.002 4.032 calculate_rho_elec 119 8.7 0.021 0.024 4.002 4.031 sum_up_and_integrate 119 10.3 0.001 0.001 3.996 4.001 integrate_v_rspace 119 11.3 0.002 0.003 3.983 3.988 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 3.633 3.643 make_m2s 4572 13.5 0.052 0.053 3.050 3.371 qs_ot_p2m_diag 50 11.0 0.008 0.012 3.318 3.344 make_images 4572 14.5 0.205 0.245 2.960 3.283 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.888 2.889 init_scf_loop 11 6.9 0.000 0.000 2.697 2.698 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.609 2.695 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.498 2.499 cp_fm_redistribute_end 50 14.0 1.263 2.457 1.268 2.461 ot_diis_step 108 11.5 0.011 0.011 2.427 2.428 cp_fm_diag_elpa_base 50 14.0 1.155 2.345 1.186 2.384 density_rs2pw 119 9.7 0.003 0.004 2.250 2.331 multiply_cannon_sync_h2d 27432 15.5 1.659 2.270 1.659 2.270 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 2.039 2.078 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.038 2.040 acc_transpose_blocks 27432 15.5 0.113 0.120 1.625 1.976 grid_integrate_task_list 119 12.3 1.836 1.919 1.836 1.919 make_images_data 4572 15.5 0.047 0.053 1.412 1.904 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.774 1.819 fft3d_ps 1201 13.6 0.519 0.574 1.757 1.792 prepare_preconditioner 11 7.9 0.000 0.000 1.758 1.785 make_preconditioner 11 8.9 0.000 0.000 1.758 1.784 potential_pw2rs 119 12.3 0.006 0.006 1.694 1.705 hybrid_alltoall_any 4725 16.4 0.054 0.117 1.250 1.681 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.605 1.674 wfi_extrapolate 11 7.9 0.001 0.001 1.609 1.609 fft_wrap_pw1pw2_140 487 12.2 0.048 0.049 1.549 1.584 mp_alltoall_d11v 2130 13.8 1.463 1.577 1.463 1.577 transfer_rs2pw 487 10.6 0.005 0.005 1.399 1.498 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.319 1.364 grid_collocate_task_list 119 9.7 1.287 1.344 1.287 1.344 mp_sum_d 4139 12.0 0.689 1.182 0.689 1.182 mp_allgather_i34 2286 14.5 0.707 1.171 0.707 1.171 parallel_gemm_fm 81 9.0 0.000 0.000 1.121 1.128 parallel_gemm_fm_cosma 81 10.0 1.121 1.127 1.121 1.127 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 1.123 1.127 make_images_sizes 4572 15.5 0.005 0.005 0.768 1.110 mp_alltoall_i44 4572 16.5 0.764 1.105 0.764 1.105 transfer_pw2rs 487 13.2 0.004 0.005 1.083 1.094 mp_alltoall_z22v 1201 15.6 0.978 1.070 0.978 1.070 make_basis_sm 11 9.8 0.000 0.000 1.035 1.038 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.953 0.969 acc_transpose_blocks_sync 82296 16.5 0.814 0.939 0.814 0.939 acc_transpose_blocks_kernels 27432 16.5 0.188 0.277 0.672 0.920 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=45.673000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=460.272727, yerr=1.354515 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 515.219456E+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 0.134 0.209 36.531 36.533 qs_mol_dyn_low 1 2.0 0.010 0.036 35.771 35.778 qs_forces 11 3.9 0.023 0.026 35.641 35.644 qs_energies 11 4.9 0.008 0.036 34.049 34.062 scf_env_do_scf 11 5.9 0.001 0.002 25.813 25.813 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 23.003 23.003 dbcsr_multiply_generic 2286 12.5 0.095 0.097 19.058 19.135 multiply_cannon 2286 13.5 0.190 0.193 15.752 16.446 multiply_cannon_loop 2286 14.5 0.857 0.905 14.980 15.680 qs_scf_new_mos 108 7.5 0.001 0.001 15.437 15.452 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.436 15.451 velocity_verlet 10 3.0 0.009 0.030 15.265 15.269 ot_scf_mini 108 9.5 0.002 0.003 14.670 14.688 multiply_cannon_multrec 18288 15.5 1.870 2.800 9.132 9.418 ot_mini 108 10.5 0.001 0.001 9.055 9.079 dbcsr_mm_accdrv_process 38222 16.0 6.961 7.931 7.170 8.064 qs_ot_get_derivative 108 11.5 0.001 0.001 7.567 7.586 init_scf_run 11 5.9 0.000 0.002 6.762 6.763 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.762 6.762 rebuild_ks_matrix 119 8.3 0.000 0.000 6.029 6.042 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.016 6.028 6.042 calculate_first_density_matrix 1 7.0 0.007 0.047 5.459 5.461 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.317 5.330 calculate_dm_sparse 119 9.5 0.000 0.001 4.031 4.043 sum_up_and_integrate 119 10.3 0.001 0.002 3.526 3.531 integrate_v_rspace 119 11.3 0.003 0.003 3.504 3.518 qs_ot_get_p 119 10.4 0.025 0.029 3.435 3.468 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.812 3.443 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 3.363 3.368 mp_waitall_1 158411 16.6 2.565 3.359 2.565 3.359 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.259 3.269 calculate_rho_elec 119 8.7 0.031 0.032 3.258 3.269 jit_kernel_multiply 4 15.3 0.151 2.942 0.151 2.942 init_scf_loop 11 6.9 0.001 0.003 2.793 2.793 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.037 2.342 apply_single 119 13.6 0.000 0.000 2.036 2.342 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.257 2.263 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.387 2.075 cp_dbcsr_syevd 50 12.0 0.003 0.005 1.961 1.961 make_m2s 4572 13.5 0.043 0.045 1.820 1.943 density_rs2pw 119 9.7 0.019 0.077 1.815 1.941 grid_integrate_task_list 119 12.3 1.811 1.918 1.811 1.918 cp_dbcsr_sm_fm_multiply 37 9.5 0.018 0.071 1.885 1.892 prepare_preconditioner 11 7.9 0.000 0.000 1.880 1.882 make_preconditioner 11 8.9 0.000 0.001 1.879 1.882 make_images 4572 14.5 0.192 0.203 1.734 1.857 make_full_inverse_cholesky 11 9.9 0.020 0.128 1.665 1.815 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.710 1.745 acc_transpose_blocks 18288 15.5 0.078 0.081 1.624 1.731 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.715 1.725 cp_fm_diag_elpa_base 50 14.0 1.690 1.705 1.711 1.722 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.709 1.718 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.528 1.536 ot_diis_step 108 11.5 0.011 0.011 1.470 1.471 mp_sum_l 7287 12.8 1.121 1.459 1.121 1.459 fft3d_ps 1201 13.6 0.535 0.557 1.394 1.431 fft_wrap_pw1pw2_140 487 12.2 0.061 0.064 1.381 1.415 potential_pw2rs 119 12.3 0.007 0.007 1.317 1.330 grid_collocate_task_list 119 9.7 1.236 1.320 1.236 1.320 multiply_cannon_sync_h2d 18288 15.5 1.033 1.235 1.033 1.235 parallel_gemm_fm 81 9.0 0.000 0.000 1.192 1.195 parallel_gemm_fm_cosma 81 10.0 1.192 1.195 1.192 1.195 make_basis_sm 11 9.8 0.000 0.001 1.174 1.181 wfi_extrapolate 11 7.9 0.002 0.007 1.176 1.176 transfer_rs2pw 487 10.6 0.005 0.005 1.055 1.170 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 1.096 1.100 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.955 0.972 make_images_data 4572 15.5 0.047 0.051 0.777 0.914 acc_transpose_blocks_sync 54864 16.5 0.740 0.841 0.740 0.841 transfer_pw2rs 487 13.2 0.004 0.004 0.828 0.831 dbcsr_complete_redistribute 329 12.2 0.106 0.184 0.664 0.825 hybrid_alltoall_any 4725 16.4 0.059 0.115 0.684 0.823 mp_alltoall_d11v 2130 13.8 0.690 0.803 0.690 0.803 mp_alltoall_z22v 1201 15.6 0.710 0.802 0.710 0.802 acc_transpose_blocks_kernels 18288 16.5 0.216 0.223 0.786 0.793 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.782 0.784 cp_fm_cholesky_invert 11 10.9 0.736 0.739 0.736 0.739 build_core_hamiltonian_matrix_ 11 4.9 0.003 0.012 0.675 0.739 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=36.533000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=490.909091, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 553.766912E+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 0.036 0.045 43.474 43.475 qs_mol_dyn_low 1 2.0 0.003 0.003 42.824 43.102 qs_forces 11 3.9 0.002 0.002 41.628 41.629 qs_energies 11 4.9 0.002 0.002 39.892 39.899 scf_env_do_scf 11 5.9 0.000 0.001 31.277 31.279 scf_env_do_scf_inner_loop 108 6.5 0.035 0.050 27.322 27.322 dbcsr_multiply_generic 2286 12.5 0.101 0.104 22.900 23.004 velocity_verlet 10 3.0 0.002 0.002 20.233 20.281 multiply_cannon 2286 13.5 0.216 0.226 18.702 19.115 qs_scf_new_mos 108 7.5 0.001 0.001 18.999 19.073 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.998 19.073 ot_scf_mini 108 9.5 0.002 0.003 17.947 18.001 multiply_cannon_loop 2286 14.5 1.509 1.570 17.638 17.897 multiply_cannon_multrec 27432 15.5 2.579 3.343 11.375 11.780 ot_mini 108 10.5 0.001 0.001 11.024 11.083 dbcsr_mm_accdrv_process 47916 15.9 8.227 10.125 8.696 10.186 qs_ot_get_derivative 108 11.5 0.001 0.001 9.123 9.178 init_scf_run 11 5.9 0.002 0.042 7.250 7.250 scf_env_initial_rho_setup 11 6.9 0.004 0.042 7.248 7.250 rebuild_ks_matrix 119 8.3 0.000 0.000 6.626 6.676 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.626 6.675 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.886 5.931 calculate_first_density_matrix 1 7.0 0.000 0.000 5.788 5.795 calculate_dm_sparse 119 9.5 0.000 0.000 4.261 4.317 qs_ot_get_p 119 10.4 0.056 0.143 4.195 4.268 init_scf_loop 11 6.9 0.000 0.000 3.933 3.933 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.244 3.707 sum_up_and_integrate 119 10.3 0.001 0.001 3.572 3.582 integrate_v_rspace 119 11.3 0.003 0.003 3.560 3.571 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.462 3.506 calculate_rho_elec 119 8.7 0.040 0.046 3.461 3.505 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 3.344 3.354 jit_kernel_multiply 5 16.3 0.402 2.832 0.402 2.832 prepare_preconditioner 11 7.9 0.000 0.000 2.742 2.751 make_preconditioner 11 8.9 0.000 0.000 2.742 2.751 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.341 2.672 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.196 2.670 apply_single 119 13.6 0.000 0.000 2.196 2.669 make_m2s 4572 13.5 0.053 0.056 2.537 2.606 acc_transpose_blocks 27432 15.5 0.116 0.119 2.502 2.602 mp_waitall_1 137007 16.6 1.985 2.551 1.985 2.551 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.471 2.498 make_images 4572 14.5 0.272 0.333 2.430 2.498 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.469 2.480 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.048 2.050 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.032 2.033 density_rs2pw 119 9.7 0.003 0.004 1.872 1.982 grid_integrate_task_list 119 12.3 1.823 1.907 1.823 1.907 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.848 1.882 ot_diis_step 108 11.5 0.012 0.013 1.860 1.861 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.844 1.857 mp_sum_l 7287 12.8 1.231 1.801 1.231 1.801 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.759 1.768 cp_fm_diag_elpa_base 50 14.0 1.723 1.739 1.757 1.766 parallel_gemm_fm 81 9.0 0.000 0.000 1.615 1.618 parallel_gemm_fm_cosma 81 10.0 1.614 1.618 1.614 1.618 make_basis_sm 11 9.8 0.000 0.000 1.566 1.569 fft3d_ps 1201 13.6 0.577 0.641 1.509 1.536 fft_wrap_pw1pw2_140 487 12.2 0.071 0.073 1.489 1.530 acc_transpose_blocks_sync 82296 16.5 1.411 1.484 1.411 1.484 md_write_output 11 3.9 0.060 1.435 0.062 1.449 md_output 10 3.0 0.000 0.000 0.063 1.448 wfi_extrapolate 11 7.9 0.001 0.001 1.402 1.402 potential_pw2rs 119 12.3 0.008 0.009 1.367 1.374 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.336 1.352 grid_collocate_task_list 119 9.7 1.248 1.331 1.248 1.331 multiply_cannon_metrocomm3 27432 15.5 0.041 0.043 0.772 1.314 mp_sum_dm 438 4.9 1.142 1.197 1.142 1.197 cp_fm_upper_to_full 72 14.2 0.837 1.188 0.837 1.188 update_particle_set 20 4.0 0.000 0.000 1.115 1.164 transfer_rs2pw 487 10.6 0.004 0.005 1.000 1.115 dbcsr_complete_redistribute 329 12.2 0.114 0.144 0.832 1.110 qs_energies_init_hamiltonians 11 5.9 0.006 0.007 1.082 1.084 make_images_data 4572 15.5 0.047 0.051 0.950 1.062 mp_alltoall_d11v 2130 13.8 0.853 1.012 0.853 1.012 acc_transpose_blocks_kernels 27432 16.5 0.270 0.277 0.946 0.984 hybrid_alltoall_any 4725 16.4 0.066 0.155 0.804 0.953 mp_sum_d 4137 12.0 0.601 0.950 0.601 0.950 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.801 0.877 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.597 0.871 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=43.475000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.181818, yerr=2.248967 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 610.103296E+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 0.029 0.031 33.937 33.938 qs_mol_dyn_low 1 2.0 0.003 0.003 33.397 33.404 qs_forces 11 3.9 0.002 0.002 31.052 31.052 qs_energies 11 4.9 0.001 0.001 29.355 29.357 scf_env_do_scf 11 5.9 0.000 0.001 21.664 21.665 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.183 19.184 velocity_verlet 10 3.0 0.002 0.002 16.747 16.892 dbcsr_multiply_generic 2286 12.5 0.095 0.098 15.335 15.401 multiply_cannon 2286 13.5 0.224 0.230 12.802 13.252 multiply_cannon_loop 2286 14.5 0.643 0.659 11.913 12.053 qs_scf_new_mos 108 7.5 0.001 0.001 11.756 11.786 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.755 11.785 ot_scf_mini 108 9.5 0.002 0.002 11.040 11.065 multiply_cannon_multrec 9144 15.5 1.836 2.063 8.329 8.573 dbcsr_mm_accdrv_process 12550 15.8 5.446 6.451 6.385 6.494 ot_mini 108 10.5 0.001 0.001 6.403 6.434 init_scf_run 11 5.9 0.000 0.003 6.156 6.156 scf_env_initial_rho_setup 11 6.9 0.002 0.003 6.155 6.156 rebuild_ks_matrix 119 8.3 0.000 0.000 5.651 5.671 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.651 5.670 qs_ot_get_derivative 108 11.5 0.001 0.001 5.131 5.157 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.024 5.041 calculate_first_density_matrix 1 7.0 0.000 0.000 5.040 5.041 calculate_dm_sparse 119 9.5 0.000 0.000 3.961 3.977 jit_kernel_multiply 10 15.6 0.900 3.949 0.900 3.949 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 3.341 3.347 sum_up_and_integrate 119 10.3 0.001 0.001 3.328 3.334 integrate_v_rspace 119 11.3 0.003 0.003 3.318 3.323 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.222 3.229 calculate_rho_elec 119 8.7 0.060 0.061 3.222 3.228 qs_ot_get_p 119 10.4 0.002 0.002 2.796 2.827 init_scf_loop 11 6.9 0.000 0.000 2.462 2.463 mp_sum_dm 438 4.9 2.197 2.347 2.197 2.347 md_write_output 11 3.9 0.146 2.320 0.148 2.335 md_output 10 3.0 0.000 0.000 0.148 2.333 update_particle_set 20 4.0 0.000 0.000 2.184 2.329 grid_integrate_task_list 119 12.3 1.867 1.921 1.867 1.921 make_m2s 4572 13.5 0.033 0.034 1.774 1.904 mp_waitall_1 115863 16.7 1.371 1.876 1.371 1.876 make_images 4572 14.5 0.269 0.300 1.684 1.812 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.797 1.799 density_rs2pw 119 9.7 0.003 0.003 1.609 1.738 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.729 1.738 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.709 1.730 prepare_preconditioner 11 7.9 0.000 0.000 1.708 1.713 make_preconditioner 11 8.9 0.000 0.000 1.708 1.713 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.604 1.627 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.601 1.603 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.560 1.560 acc_transpose_blocks 9144 15.5 0.042 0.043 1.425 1.454 fft_wrap_pw1pw2_140 487 12.2 0.093 0.096 1.436 1.445 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.434 1.440 grid_collocate_task_list 119 9.7 1.297 1.391 1.297 1.391 fft3d_ps 1201 13.6 0.637 0.653 1.341 1.351 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.316 1.328 qs_energies_init_hamiltonians 11 5.9 0.014 0.021 1.320 1.323 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.281 1.289 cp_fm_diag_elpa_base 50 14.0 1.254 1.272 1.280 1.288 ot_diis_step 108 11.5 0.013 0.013 1.261 1.261 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.213 1.241 apply_single 119 13.6 0.000 0.000 1.213 1.241 potential_pw2rs 119 12.3 0.010 0.010 1.141 1.143 wfi_extrapolate 11 7.9 0.001 0.001 1.066 1.066 parallel_gemm_fm 81 9.0 0.000 0.000 0.958 0.964 parallel_gemm_fm_cosma 81 10.0 0.958 0.963 0.958 0.963 transfer_rs2pw 487 10.6 0.004 0.005 0.821 0.953 hybrid_alltoall_any 4725 16.4 0.066 0.178 0.765 0.942 make_basis_sm 11 9.8 0.000 0.000 0.938 0.938 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.878 0.934 make_images_data 4572 15.5 0.042 0.046 0.749 0.897 mp_alltoall_d11v 2130 13.8 0.783 0.896 0.783 0.896 cp_fm_cholesky_invert 11 10.9 0.859 0.862 0.859 0.862 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.819 0.825 qs_env_update_s_mstruct 11 6.9 0.014 0.034 0.735 0.801 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.747 0.749 acc_transpose_blocks_sync 27432 16.5 0.720 0.746 0.720 0.746 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=33.938000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=576.636364, yerr=6.581969 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 781.094912E+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 0.073 0.076 45.736 45.736 qs_mol_dyn_low 1 2.0 0.003 0.003 45.142 45.150 qs_forces 11 3.9 0.016 0.018 45.084 45.084 qs_energies 11 4.9 0.001 0.001 43.093 43.098 scf_env_do_scf 11 5.9 0.001 0.001 34.792 34.792 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.763 26.765 velocity_verlet 10 3.0 0.002 0.002 24.274 24.279 dbcsr_multiply_generic 2286 12.5 0.102 0.103 21.412 21.638 multiply_cannon 2286 13.5 0.294 0.299 17.364 18.309 qs_scf_new_mos 108 7.5 0.001 0.001 17.467 17.560 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.466 17.559 multiply_cannon_loop 2286 14.5 0.863 0.880 16.068 17.039 ot_scf_mini 108 9.5 0.002 0.002 16.281 16.379 multiply_cannon_multrec 9144 15.5 3.406 4.956 11.053 11.148 ot_mini 108 10.5 0.001 0.001 9.880 9.992 dbcsr_mm_accdrv_process 12550 15.8 6.961 8.794 7.514 8.826 init_scf_loop 11 6.9 0.000 0.000 8.001 8.003 qs_ot_get_derivative 108 11.5 0.001 0.001 7.799 7.898 rebuild_ks_matrix 119 8.3 0.000 0.000 7.055 7.191 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.054 7.190 prepare_preconditioner 11 7.9 0.000 0.000 7.038 7.051 make_preconditioner 11 8.9 0.000 0.000 7.038 7.051 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.553 6.915 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.354 6.476 init_scf_run 11 5.9 0.000 0.001 6.177 6.177 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.177 6.177 calculate_first_density_matrix 1 7.0 0.000 0.000 4.705 4.707 cp_fm_upper_to_full 72 14.2 3.201 4.619 3.201 4.619 calculate_dm_sparse 119 9.5 0.000 0.000 4.566 4.585 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.964 3.972 calculate_rho_elec 119 8.7 0.118 0.121 3.964 3.971 sum_up_and_integrate 119 10.3 0.001 0.001 3.741 3.748 integrate_v_rspace 119 11.3 0.003 0.003 3.731 3.738 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 3.543 3.552 qs_ot_get_p 119 10.4 0.002 0.002 3.421 3.551 mp_waitall_1 94719 16.7 2.410 3.410 2.410 3.410 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.861 3.276 dbcsr_complete_redistribute 329 12.2 0.282 0.287 2.035 2.872 make_m2s 4572 13.5 0.037 0.037 2.480 2.659 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.718 2.542 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.246 2.541 apply_single 119 13.6 0.000 0.000 2.246 2.541 make_images 4572 14.5 0.351 0.381 2.360 2.539 fft_wrap_pw1pw2 1201 11.6 0.014 0.014 2.410 2.416 mp_alltoall_i22 627 13.8 1.444 2.302 1.444 2.302 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.480 2.297 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.363 2.266 grid_integrate_task_list 119 12.3 2.088 2.146 2.088 2.146 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.060 2.110 ot_diis_step 108 11.5 0.014 0.014 2.058 2.058 fft_wrap_pw1pw2_140 487 12.2 0.179 0.180 2.038 2.044 qs_ot_p2m_diag 50 11.0 0.042 0.043 2.018 2.020 density_rs2pw 119 9.7 0.003 0.003 1.981 2.016 fft3d_ps 1201 13.6 0.859 0.876 1.862 1.867 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.837 1.839 acc_transpose_blocks 9144 15.5 0.044 0.045 1.809 1.837 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.821 1.822 mp_sum_l 7287 12.8 1.056 1.747 1.056 1.747 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.666 1.666 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.573 1.614 grid_collocate_task_list 119 9.7 1.527 1.562 1.527 1.562 cp_fm_cholesky_invert 11 10.9 1.542 1.546 1.542 1.546 wfi_extrapolate 11 7.9 0.001 0.001 1.410 1.410 hybrid_alltoall_any 4725 16.4 0.091 0.152 1.129 1.402 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.375 1.375 cp_fm_diag_elpa_base 50 14.0 1.227 1.281 1.373 1.373 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.323 1.342 make_images_data 4572 15.5 0.045 0.048 1.092 1.337 mp_alltoall_d11v 2130 13.8 1.200 1.240 1.200 1.240 potential_pw2rs 119 12.3 0.013 0.014 1.221 1.223 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.163 1.177 acc_transpose_blocks_sync 27432 16.5 1.101 1.127 1.101 1.127 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.989 1.016 qs_create_task_list 11 7.9 0.000 0.000 0.932 0.942 generate_qs_task_list 11 8.9 0.369 0.389 0.931 0.942 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.910 0.923 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=45.736000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=731.454545, yerr=18.267525 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 498.491392E+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 933011. 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 0.047 0.110 85.197 85.200 qs_mol_dyn_low 1 2.0 0.006 0.031 84.358 84.367 qs_forces 11 3.9 0.022 0.063 84.285 84.289 qs_energies 11 4.9 0.003 0.017 81.402 81.434 scf_env_do_scf 11 5.9 0.000 0.001 69.456 69.459 scf_env_do_scf_inner_loop 99 6.5 0.004 0.011 63.772 63.772 dbcsr_multiply_generic 2055 12.4 0.108 0.112 52.841 53.060 qs_scf_new_mos 99 7.5 0.000 0.001 46.903 46.983 qs_scf_loop_do_ot 99 8.5 0.000 0.001 46.902 46.983 multiply_cannon 2055 13.4 0.176 0.180 44.544 45.473 ot_scf_mini 99 9.5 0.002 0.002 44.566 44.676 multiply_cannon_loop 2055 14.4 1.774 1.821 43.585 44.475 velocity_verlet 10 3.0 0.001 0.005 43.053 43.054 ot_mini 99 10.5 0.001 0.001 26.460 26.562 multiply_cannon_multrec 49320 15.4 11.366 12.071 19.606 20.216 qs_ot_get_derivative 99 11.5 0.001 0.001 19.666 19.775 rebuild_ks_matrix 110 8.3 0.000 0.000 14.537 14.614 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.537 14.614 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.759 12.829 mp_waitall_1 220248 16.4 10.614 11.315 10.614 11.315 multiply_cannon_sync_h2d 49320 15.4 9.622 10.213 9.622 10.213 qs_ot_get_p 110 10.4 0.005 0.020 9.737 9.832 init_scf_run 11 5.9 0.000 0.001 9.480 9.480 scf_env_initial_rho_setup 11 6.9 0.038 0.056 9.480 9.480 dbcsr_mm_accdrv_process 87628 16.1 3.067 3.149 8.111 8.363 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.573 8.108 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.166 7.662 apply_single 110 13.6 0.000 0.001 7.166 7.662 multiply_cannon_metrocomm3 49320 15.4 0.083 0.085 6.396 7.524 sum_up_and_integrate 110 10.3 0.002 0.004 6.943 6.957 integrate_v_rspace 110 11.3 0.003 0.003 6.917 6.938 ot_diis_step 99 11.5 0.006 0.006 6.589 6.589 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.478 6.494 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.130 6.272 calculate_rho_elec 110 8.6 0.020 0.024 6.129 6.271 calculate_dm_sparse 110 9.5 0.001 0.001 5.933 6.020 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.685 5.686 init_scf_loop 11 6.9 0.000 0.000 5.658 5.658 calculate_first_density_matrix 1 7.0 0.000 0.002 5.323 5.360 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.225 5.285 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.116 5.144 cp_fm_diag_elpa_base 48 14.0 5.103 5.130 5.114 5.142 mp_sum_l 6594 12.7 3.970 4.897 3.970 4.897 jit_kernel_multiply 13 15.9 4.766 4.795 4.766 4.795 make_m2s 4110 13.4 0.061 0.066 3.927 4.044 wfi_extrapolate 11 7.9 0.001 0.003 4.024 4.025 make_images 4110 14.4 0.177 0.190 3.831 3.950 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 3.875 3.897 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.582 3.587 grid_integrate_task_list 110 12.3 3.282 3.451 3.282 3.451 multiply_cannon_metrocomm1 49320 15.4 0.064 0.067 2.322 3.438 prepare_preconditioner 11 7.9 0.000 0.000 3.384 3.399 make_preconditioner 11 8.9 0.000 0.000 3.384 3.399 density_rs2pw 110 9.6 0.004 0.004 3.187 3.351 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.162 3.208 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.161 3.204 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.153 3.197 fft_wrap_pw1pw2 1111 11.6 0.014 0.018 2.745 2.811 acc_transpose_blocks 49320 15.4 0.205 0.213 2.552 2.610 potential_pw2rs 110 12.3 0.005 0.006 2.530 2.548 mp_alltoall_d11v 2046 13.8 2.047 2.498 2.047 2.498 fft_wrap_pw1pw2_140 451 12.1 0.093 0.098 2.326 2.399 fft3d_ps 1111 13.6 0.731 0.814 2.282 2.343 grid_collocate_task_list 110 9.6 2.155 2.263 2.155 2.263 mp_waitany 14300 13.8 1.955 2.235 1.955 2.235 transfer_rs2pw 451 10.6 0.006 0.006 2.011 2.200 mp_sum_d 3893 11.9 1.546 2.144 1.546 2.144 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.920 1.932 make_images_data 4110 15.4 0.043 0.046 1.767 1.869 transfer_pw2rs 451 13.1 0.005 0.007 1.804 1.813 hybrid_alltoall_any 4261 16.3 0.085 0.485 1.539 1.813 cp_fm_cholesky_invert 11 10.9 1.797 1.800 1.797 1.800 parallel_gemm_fm 81 9.0 0.000 0.000 1.743 1.751 parallel_gemm_fm_cosma 81 10.0 1.743 1.751 1.743 1.751 qs_energies_init_hamiltonians 11 5.9 0.008 0.059 1.731 1.746 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=85.200000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=472.909091, yerr=3.342687 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 591.953920E+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 1580202. 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 0.031 0.049 71.770 71.771 qs_mol_dyn_low 1 2.0 0.007 0.038 71.255 71.266 qs_forces 11 3.9 0.016 0.022 71.137 71.142 qs_energies 11 4.9 0.004 0.021 67.796 67.816 scf_env_do_scf 11 5.9 0.000 0.001 56.367 56.371 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.845 48.845 dbcsr_multiply_generic 2055 12.4 0.134 0.168 40.352 40.537 velocity_verlet 10 3.0 0.002 0.004 35.858 35.860 multiply_cannon 2055 13.4 0.223 0.244 33.714 34.927 qs_scf_new_mos 99 7.5 0.001 0.001 33.172 33.312 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.171 33.312 multiply_cannon_loop 2055 14.4 1.155 1.188 32.459 33.274 ot_scf_mini 99 9.5 0.003 0.003 31.495 31.627 ot_mini 99 10.5 0.001 0.001 18.350 18.489 multiply_cannon_multrec 24660 15.4 6.991 8.485 16.466 18.135 rebuild_ks_matrix 110 8.3 0.000 0.000 13.535 13.634 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.020 13.535 13.634 qs_ot_get_derivative 99 11.5 0.001 0.001 12.599 12.734 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.907 11.991 mp_waitall_1 176588 16.5 7.509 9.769 7.509 9.769 dbcsr_mm_accdrv_process 52282 16.1 7.056 8.862 9.305 9.713 init_scf_run 11 5.9 0.000 0.002 8.931 8.932 scf_env_initial_rho_setup 11 6.9 0.001 0.002 8.931 8.932 multiply_cannon_metrocomm3 24660 15.4 0.072 0.074 5.080 7.830 init_scf_loop 11 6.9 0.001 0.002 7.489 7.490 multiply_cannon_sync_h2d 24660 15.4 6.371 7.230 6.371 7.230 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.422 7.108 apply_single 110 13.6 0.000 0.001 6.422 7.108 qs_ot_get_p 110 10.4 0.001 0.001 6.506 6.684 sum_up_and_integrate 110 10.3 0.002 0.003 6.271 6.283 integrate_v_rspace 110 11.3 0.002 0.003 6.244 6.257 ot_diis_step 99 11.5 0.010 0.010 5.704 5.704 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.693 5.701 calculate_rho_elec 110 8.6 0.039 0.047 5.692 5.700 prepare_preconditioner 11 7.9 0.000 0.000 5.480 5.498 make_preconditioner 11 8.9 0.000 0.000 5.480 5.498 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.714 5.432 calculate_dm_sparse 110 9.5 0.001 0.001 5.327 5.350 calculate_first_density_matrix 1 7.0 0.001 0.010 5.330 5.332 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.063 5.216 make_m2s 4110 13.4 0.056 0.060 4.253 4.712 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.603 4.624 make_images 4110 14.4 0.406 0.466 4.141 4.597 jit_kernel_multiply 11 16.4 1.878 4.483 1.878 4.483 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.139 4.140 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 3.885 3.896 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.572 3.585 cp_fm_diag_elpa_base 48 14.0 3.525 3.543 3.569 3.582 wfi_extrapolate 11 7.9 0.003 0.019 3.507 3.507 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.455 3.458 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.295 3.371 grid_integrate_task_list 110 12.3 3.160 3.342 3.160 3.342 fft_wrap_pw1pw2 1111 11.6 0.015 0.017 3.098 3.232 density_rs2pw 110 9.6 0.004 0.004 2.939 3.140 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.056 3.087 make_images_data 4110 15.4 0.048 0.052 2.298 2.779 hybrid_alltoall_any 4261 16.3 0.106 0.460 1.982 2.768 fft_wrap_pw1pw2_140 451 12.1 0.110 0.115 2.624 2.756 fft3d_ps 1111 13.6 1.069 1.268 2.475 2.600 cp_fm_cholesky_invert 11 10.9 2.546 2.554 2.546 2.554 mp_sum_l 6594 12.7 1.750 2.380 1.750 2.380 grid_collocate_task_list 110 9.6 2.168 2.283 2.168 2.283 potential_pw2rs 110 12.3 0.008 0.008 2.152 2.164 acc_transpose_blocks 24660 15.4 0.115 0.118 1.977 2.012 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.963 1.983 mp_alltoall_d11v 2046 13.8 1.723 1.960 1.723 1.960 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.940 1.958 multiply_cannon_metrocomm4 22605 15.4 0.078 0.083 0.783 1.830 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.792 1.803 transfer_rs2pw 451 10.6 0.007 0.007 1.523 1.715 mp_irecv_dv 57340 16.2 0.653 1.701 0.653 1.701 mp_allgather_i34 2055 14.4 0.564 1.576 0.564 1.576 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.389 1.492 mp_waitany 10164 13.8 1.267 1.490 1.267 1.490 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.771000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=559.181818, yerr=7.505370 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 668.676096E+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 0.177 0.203 64.513 64.516 qs_mol_dyn_low 1 2.0 0.006 0.028 63.354 63.365 qs_forces 11 3.9 0.025 0.027 63.233 63.234 qs_energies 11 4.9 0.006 0.036 59.998 60.007 scf_env_do_scf 11 5.9 0.001 0.001 49.253 49.253 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 40.458 40.458 velocity_verlet 10 3.0 0.006 0.032 32.957 32.960 dbcsr_multiply_generic 2055 12.4 0.110 0.114 31.255 31.526 multiply_cannon 2055 13.4 0.210 0.215 24.842 26.124 qs_scf_new_mos 99 7.5 0.001 0.001 25.919 26.032 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.918 26.032 multiply_cannon_loop 2055 14.4 0.813 0.836 23.621 24.814 ot_scf_mini 99 9.5 0.002 0.003 24.654 24.772 ot_mini 99 10.5 0.001 0.001 13.971 14.090 multiply_cannon_multrec 16440 15.4 3.556 4.713 12.038 13.106 rebuild_ks_matrix 110 8.3 0.000 0.000 12.246 12.429 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.246 12.428 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.794 10.954 mp_waitall_1 139946 16.5 7.075 9.945 7.075 9.945 qs_ot_get_derivative 99 11.5 0.001 0.001 9.457 9.575 init_scf_loop 11 6.9 0.001 0.004 8.758 8.759 dbcsr_mm_accdrv_process 34862 16.1 7.616 7.989 8.330 8.475 init_scf_run 11 5.9 0.000 0.002 7.971 7.971 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.971 7.971 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.453 7.255 prepare_preconditioner 11 7.9 0.000 0.000 6.928 6.943 make_preconditioner 11 8.9 0.000 0.002 6.928 6.942 make_full_inverse_cholesky 11 9.9 0.000 0.001 6.266 6.632 sum_up_and_integrate 110 10.3 0.001 0.002 6.215 6.228 integrate_v_rspace 110 11.3 0.003 0.003 6.188 6.202 qs_ot_get_p 110 10.4 0.015 0.017 5.831 5.983 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.639 5.652 calculate_rho_elec 110 8.6 0.058 0.058 5.638 5.652 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.043 5.492 apply_single 110 13.6 0.000 0.000 5.043 5.492 calculate_first_density_matrix 1 7.0 0.000 0.002 4.839 4.841 calculate_dm_sparse 110 9.5 0.001 0.001 4.780 4.813 make_m2s 4110 13.4 0.049 0.051 4.210 4.604 make_images 4110 14.4 0.397 0.518 4.093 4.487 ot_diis_step 99 11.5 0.011 0.011 4.480 4.480 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.127 4.132 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.182 3.860 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.750 3.751 multiply_cannon_sync_h2d 16440 15.4 3.227 3.739 3.227 3.739 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 3.706 3.719 grid_integrate_task_list 110 12.3 3.206 3.429 3.206 3.429 cp_fm_diag_elpa 48 13.0 0.000 0.001 3.193 3.201 cp_fm_diag_elpa_base 48 14.0 3.125 3.156 3.190 3.198 fft_wrap_pw1pw2 1111 11.6 0.013 0.017 3.180 3.196 density_rs2pw 110 9.6 0.008 0.040 2.874 3.053 wfi_extrapolate 11 7.9 0.001 0.006 2.979 2.979 make_images_data 4110 15.4 0.045 0.049 2.366 2.871 hybrid_alltoall_any 4261 16.3 0.109 0.379 2.114 2.823 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.801 2.802 fft_wrap_pw1pw2_140 451 12.1 0.130 0.152 2.720 2.741 cp_fm_cholesky_invert 11 10.9 2.694 2.701 2.694 2.701 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.498 2.564 multiply_cannon_metrocomm4 14385 15.4 0.048 0.052 0.909 2.531 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.465 2.513 fft3d_ps 1111 13.6 1.088 1.098 2.488 2.513 mp_irecv_dv 48980 15.7 0.832 2.391 0.832 2.391 grid_collocate_task_list 110 9.6 2.220 2.380 2.220 2.380 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.285 2.286 mp_sum_l 6594 12.7 1.618 2.242 1.618 2.242 potential_pw2rs 110 12.3 0.010 0.010 2.050 2.059 mp_alltoall_d11v 2046 13.8 1.813 1.989 1.813 1.989 dbcsr_complete_redistribute 325 12.2 0.327 0.374 1.501 1.965 cp_fm_upper_to_full 70 14.2 1.415 1.798 1.415 1.798 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.651 1.674 transfer_rs2pw 451 10.6 0.005 0.006 1.456 1.658 cp_fm_cholesky_decompose 22 10.9 1.632 1.653 1.632 1.653 mp_allgather_i34 2055 14.4 0.537 1.649 0.537 1.649 acc_transpose_blocks 16440 15.4 0.076 0.080 1.577 1.637 build_core_hamiltonian_matrix_ 11 4.9 0.002 0.013 1.371 1.501 mp_waitany 17072 13.8 1.222 1.447 1.222 1.447 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.356 1.445 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.975 1.436 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.306 1.315 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=64.516000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.636364, yerr=8.260230 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 737.939456E+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 0.024 0.038 68.352 68.352 qs_mol_dyn_low 1 2.0 0.003 0.003 67.771 67.792 qs_forces 11 3.9 0.006 0.008 67.705 67.705 qs_energies 11 4.9 0.001 0.002 64.330 64.337 scf_env_do_scf 11 5.9 0.000 0.001 53.281 53.284 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.504 41.504 velocity_verlet 10 3.0 0.001 0.002 36.761 36.763 dbcsr_multiply_generic 2055 12.4 0.118 0.123 32.821 33.070 qs_scf_new_mos 99 7.5 0.001 0.001 27.322 27.420 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.321 27.420 multiply_cannon 2055 13.4 0.237 0.254 25.911 27.058 ot_scf_mini 99 9.5 0.002 0.003 25.635 25.728 multiply_cannon_loop 2055 14.4 1.394 1.473 24.501 25.091 multiply_cannon_multrec 24660 15.4 4.126 6.841 15.328 16.504 ot_mini 99 10.5 0.001 0.001 14.864 14.976 dbcsr_mm_accdrv_process 52304 16.0 10.157 11.481 11.047 11.968 rebuild_ks_matrix 110 8.3 0.000 0.000 11.865 11.955 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.864 11.954 init_scf_loop 11 6.9 0.000 0.000 11.738 11.739 qs_ot_get_derivative 99 11.5 0.001 0.001 10.703 10.794 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.486 10.568 prepare_preconditioner 11 7.9 0.000 0.000 9.838 9.853 make_preconditioner 11 8.9 0.000 0.000 9.838 9.853 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.048 9.521 init_scf_run 11 5.9 0.000 0.001 8.314 8.315 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.314 8.315 mp_waitall_1 121746 16.5 4.186 6.088 4.186 6.088 sum_up_and_integrate 110 10.3 0.001 0.002 6.048 6.061 integrate_v_rspace 110 11.3 0.003 0.003 6.022 6.034 qs_ot_get_p 110 10.4 0.019 0.041 5.819 5.940 make_m2s 4110 13.4 0.059 0.061 5.317 5.671 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.517 5.528 calculate_rho_elec 110 8.6 0.077 0.081 5.516 5.527 make_images 4110 14.4 0.573 0.690 5.176 5.524 calculate_dm_sparse 110 9.5 0.001 0.001 5.266 5.300 calculate_first_density_matrix 1 7.0 0.000 0.000 5.297 5.300 cp_fm_upper_to_full 70 14.2 3.388 4.904 3.388 4.904 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.046 4.142 apply_single 110 13.6 0.000 0.000 4.046 4.142 ot_diis_step 99 11.5 0.011 0.011 4.121 4.122 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.856 3.871 dbcsr_complete_redistribute 325 12.2 0.426 0.549 2.710 3.842 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 3.821 3.840 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.459 3.503 grid_integrate_task_list 110 12.3 3.272 3.441 3.272 3.441 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.311 3.360 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.206 3.327 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.302 3.303 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.139 3.161 multiply_cannon_metrocomm3 24660 15.4 0.038 0.040 1.465 3.037 make_images_data 4110 15.4 0.048 0.052 2.600 2.948 wfi_extrapolate 11 7.9 0.001 0.001 2.925 2.925 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.898 2.900 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.782 2.886 hybrid_alltoall_any 4261 16.3 0.123 0.464 2.272 2.878 mp_alltoall_i22 605 13.7 1.650 2.835 1.650 2.835 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.775 2.784 cp_fm_diag_elpa_base 48 14.0 2.629 2.688 2.773 2.783 density_rs2pw 110 9.6 0.004 0.004 2.656 2.745 fft_wrap_pw1pw2_140 451 12.1 0.145 0.147 2.688 2.712 cp_fm_cholesky_invert 11 10.9 2.616 2.624 2.616 2.624 multiply_cannon_sync_h2d 24660 15.4 2.386 2.544 2.386 2.544 acc_transpose_blocks 24660 15.4 0.109 0.112 2.392 2.512 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.470 2.498 fft3d_ps 1111 13.6 1.126 1.163 2.409 2.424 grid_collocate_task_list 110 9.6 2.265 2.394 2.265 2.394 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.221 2.224 mp_alltoall_d11v 2046 13.8 1.737 1.983 1.737 1.983 potential_pw2rs 110 12.3 0.012 0.013 1.877 1.880 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.789 1.817 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.615 1.715 cp_fm_cholesky_decompose 22 10.9 1.631 1.683 1.631 1.683 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.620 1.631 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.835 1.568 mp_sum_l 6594 12.7 0.916 1.530 0.916 1.530 mp_irecv_dv 62702 16.1 0.733 1.486 0.733 1.486 acc_transpose_blocks_sync 73980 16.4 1.383 1.485 1.383 1.485 mp_allgather_i34 2055 14.4 0.495 1.461 0.495 1.461 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=68.352000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=699.000000, yerr=8.570563 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 879.689728E+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 0.032 0.040 57.992 57.993 qs_mol_dyn_low 1 2.0 0.003 0.003 57.240 57.249 qs_forces 11 3.9 0.003 0.003 57.176 57.177 qs_energies 11 4.9 0.001 0.002 53.534 53.538 scf_env_do_scf 11 5.9 0.000 0.001 42.320 42.321 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.612 34.613 velocity_verlet 10 3.0 0.002 0.002 30.674 30.677 dbcsr_multiply_generic 2055 12.4 0.106 0.108 25.559 25.703 multiply_cannon 2055 13.4 0.237 0.247 20.137 21.251 qs_scf_new_mos 99 7.5 0.001 0.001 20.581 20.653 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.581 20.652 ot_scf_mini 99 9.5 0.002 0.002 19.331 19.393 multiply_cannon_loop 2055 14.4 0.604 0.622 18.927 19.227 rebuild_ks_matrix 110 8.3 0.000 0.000 11.423 11.494 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.423 11.493 multiply_cannon_multrec 8220 15.4 3.210 4.673 9.992 11.090 ot_mini 99 10.5 0.001 0.001 10.783 10.845 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.169 10.234 init_scf_run 11 5.9 0.000 0.001 8.031 8.031 scf_env_initial_rho_setup 11 6.9 0.002 0.002 8.031 8.031 init_scf_loop 11 6.9 0.000 0.000 7.662 7.663 dbcsr_mm_accdrv_process 17442 15.9 5.453 6.460 6.640 7.593 mp_waitall_1 103326 16.6 5.856 7.503 5.856 7.503 qs_ot_get_derivative 99 11.5 0.001 0.001 7.015 7.079 sum_up_and_integrate 110 10.3 0.001 0.002 5.999 6.014 integrate_v_rspace 110 11.3 0.003 0.003 5.972 5.986 prepare_preconditioner 11 7.9 0.000 0.000 5.971 5.976 make_preconditioner 11 8.9 0.000 0.000 5.971 5.976 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.553 5.623 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.609 5.622 calculate_rho_elec 110 8.6 0.115 0.116 5.608 5.621 calculate_first_density_matrix 1 7.0 0.000 0.000 5.285 5.288 calculate_dm_sparse 110 9.5 0.001 0.001 4.798 4.835 qs_ot_get_p 110 10.4 0.012 0.049 4.684 4.762 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 3.062 4.456 make_m2s 4110 13.4 0.037 0.038 4.082 4.269 make_images 4110 14.4 0.644 0.710 3.952 4.134 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.809 3.857 apply_single 110 13.6 0.000 0.000 3.809 3.857 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 3.750 3.762 ot_diis_step 99 11.5 0.012 0.012 3.747 3.747 grid_integrate_task_list 110 12.3 3.368 3.539 3.368 3.539 fft_wrap_pw1pw2 1111 11.6 0.014 0.016 3.302 3.313 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.277 3.280 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.943 2.944 cp_fm_cholesky_invert 11 10.9 2.866 2.870 2.866 2.870 fft_wrap_pw1pw2_140 451 12.1 0.192 0.195 2.851 2.863 density_rs2pw 110 9.6 0.003 0.004 2.651 2.795 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 2.729 2.730 wfi_extrapolate 11 7.9 0.001 0.001 2.653 2.653 grid_collocate_task_list 110 9.6 2.368 2.565 2.368 2.565 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.549 2.550 hybrid_alltoall_any 4261 16.3 0.201 0.849 2.146 2.515 fft3d_ps 1111 13.6 1.271 1.280 2.461 2.473 multiply_cannon_sync_h2d 8220 15.4 2.371 2.464 2.371 2.464 make_images_data 4110 15.4 0.041 0.047 2.159 2.426 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.408 2.415 cp_fm_diag_elpa_base 48 14.0 2.353 2.375 2.406 2.414 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.198 2.219 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.007 2.062 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.795 2.027 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.890 1.923 mp_alltoall_d11v 2046 13.8 1.618 1.856 1.618 1.856 potential_pw2rs 110 12.3 0.015 0.015 1.809 1.812 cp_fm_cholesky_decompose 22 10.9 1.712 1.777 1.712 1.777 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.539 1.653 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.585 1.590 dbcsr_complete_redistribute 325 12.2 0.557 0.574 1.412 1.502 acc_transpose_blocks 8220 15.4 0.039 0.040 1.380 1.408 mp_allgather_i34 2055 14.4 0.382 1.400 0.382 1.400 transfer_rs2pw 451 10.6 0.005 0.006 1.126 1.313 make_basis_sm 11 9.8 0.000 0.000 1.301 1.303 qs_create_task_list 11 7.9 0.000 0.001 1.204 1.301 generate_qs_task_list 11 8.9 0.374 0.441 1.203 1.301 parallel_gemm_fm 81 9.0 0.000 0.000 1.236 1.276 parallel_gemm_fm_cosma 81 10.0 1.236 1.276 1.236 1.276 jit_kernel_multiply 7 15.3 0.866 1.235 0.866 1.235 mp_waitany 9240 13.8 1.053 1.226 1.053 1.226 multiply_cannon_metrocomm4 6165 15.4 0.018 0.020 0.479 1.224 mp_irecv_dv 24056 15.7 0.452 1.179 0.452 1.179 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.828 1.175 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.153 1.174 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.993000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=818.454545, yerr=13.282686 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 1.414648E+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 0.151 0.181 91.986 91.986 qs_mol_dyn_low 1 2.0 0.003 0.003 89.839 90.644 qs_forces 11 3.9 0.016 0.017 89.654 89.654 qs_energies 11 4.9 0.015 0.016 85.534 85.536 scf_env_do_scf 11 5.9 0.000 0.001 72.445 72.445 velocity_verlet 10 3.0 0.002 0.002 55.469 55.476 scf_env_do_scf_inner_loop 99 6.5 0.013 0.034 44.062 44.065 dbcsr_multiply_generic 2055 12.4 0.121 0.125 32.043 32.139 init_scf_loop 11 6.9 0.000 0.000 28.310 28.315 qs_scf_new_mos 99 7.5 0.001 0.001 27.042 27.080 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.041 27.079 prepare_preconditioner 11 7.9 0.000 0.000 26.298 26.307 make_preconditioner 11 8.9 0.000 0.000 26.298 26.306 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.415 25.771 multiply_cannon 2055 13.4 0.341 0.382 24.816 25.537 ot_scf_mini 99 9.5 0.002 0.002 25.235 25.267 multiply_cannon_loop 2055 14.4 0.828 0.846 23.041 23.379 cp_fm_upper_to_full 70 14.2 12.741 18.346 12.741 18.346 ot_mini 99 10.5 0.001 0.001 14.224 14.257 rebuild_ks_matrix 110 8.3 0.000 0.001 13.506 13.541 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.505 13.540 multiply_cannon_multrec 8220 15.4 4.155 4.324 12.189 12.341 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.258 12.290 dbcsr_complete_redistribute 325 12.2 1.023 1.047 7.457 10.728 qs_ot_get_derivative 99 11.5 0.001 0.001 9.671 9.701 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.427 9.691 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.867 9.101 mp_alltoall_i22 605 13.7 5.482 8.767 5.482 8.767 init_scf_run 11 5.9 0.002 0.011 8.373 8.373 scf_env_initial_rho_setup 11 6.9 0.030 0.033 8.371 8.373 mp_waitall_1 84994 16.7 7.442 8.286 7.442 8.286 dbcsr_mm_accdrv_process 11614 15.7 4.289 7.191 7.884 8.116 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.891 6.931 calculate_rho_elec 110 8.6 0.227 0.227 6.891 6.931 sum_up_and_integrate 110 10.3 0.002 0.002 6.635 6.649 integrate_v_rspace 110 11.3 0.004 0.004 6.607 6.620 make_m2s 4110 13.4 0.042 0.043 5.440 5.921 qs_ot_get_p 110 10.4 0.003 0.003 5.717 5.755 make_images 4110 14.4 0.882 0.937 5.251 5.731 calculate_dm_sparse 110 9.5 0.001 0.001 5.483 5.507 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.750 5.219 apply_single 110 13.6 0.000 0.000 4.750 5.219 cp_fm_cholesky_invert 11 10.9 5.172 5.176 5.172 5.176 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.778 5.086 calculate_first_density_matrix 1 7.0 0.000 0.000 4.941 4.943 fft_wrap_pw1pw2 1111 11.6 0.016 0.016 4.537 4.545 ot_diis_step 99 11.5 0.015 0.015 4.538 4.539 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.150 4.157 jit_kernel_multiply 10 15.1 3.397 4.037 3.397 4.037 qs_energies_init_hamiltonians 11 5.9 0.108 0.113 4.007 4.012 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 3.976 3.988 fft_wrap_pw1pw2_140 451 12.1 0.363 0.366 3.950 3.960 grid_integrate_task_list 110 12.3 3.681 3.752 3.681 3.752 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.629 3.630 hybrid_alltoall_any 4261 16.3 0.268 0.591 2.814 3.565 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.034 3.543 make_images_data 4110 15.4 0.044 0.047 2.810 3.437 fft3d_ps 1111 13.6 1.838 1.859 3.395 3.405 density_rs2pw 110 9.6 0.004 0.004 3.353 3.360 wfi_extrapolate 11 7.9 0.001 0.001 3.290 3.290 multiply_cannon_sync_h2d 8220 15.4 3.124 3.139 3.124 3.139 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.039 3.041 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.036 3.036 cp_fm_diag_elpa_base 48 14.0 2.494 2.694 3.034 3.034 grid_collocate_task_list 110 9.6 2.671 2.677 2.671 2.677 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.651 2.673 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.542 2.555 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.407 2.469 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.110 2.181 potential_pw2rs 110 12.3 0.021 0.021 2.152 2.153 mp_alltoall_d11v 2046 13.8 2.027 2.094 2.027 2.094 qs_create_task_list 11 7.9 0.075 0.076 1.960 2.008 cp_fm_cholesky_decompose 22 10.9 1.984 2.006 1.984 2.006 generate_qs_task_list 11 8.9 0.733 0.786 1.885 1.932 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.848 1.851 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=91.986000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1263.636364, yerr=62.666901 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 628.989952E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57505. MP_Allreduce 11129 795. MP_Sync 87 MP_Alltoall 2226 2529087. 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 0.034 0.073 208.697 208.699 qs_mol_dyn_low 1 2.0 0.006 0.024 207.887 207.902 qs_forces 11 3.9 0.005 0.005 207.730 207.732 qs_energies 11 4.9 0.001 0.002 202.148 202.169 scf_env_do_scf 11 5.9 0.001 0.001 183.101 183.105 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 162.901 162.903 dbcsr_multiply_generic 2507 12.6 0.176 0.182 127.586 128.072 qs_scf_new_mos 117 7.6 0.001 0.001 123.885 124.069 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.884 124.068 velocity_verlet 10 3.0 0.012 0.061 122.484 122.486 ot_scf_mini 117 9.6 0.003 0.003 117.120 117.301 multiply_cannon 2507 13.6 0.237 0.247 103.859 105.140 multiply_cannon_loop 2507 14.6 2.387 2.446 101.718 103.180 ot_mini 117 10.6 0.001 0.001 65.835 66.110 multiply_cannon_multrec 60168 15.6 31.818 33.848 43.958 45.872 qs_ot_get_derivative 117 11.6 0.001 0.001 41.026 41.195 rebuild_ks_matrix 128 8.3 0.001 0.001 33.641 33.875 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.641 33.874 mp_waitall_1 267128 16.5 28.708 31.314 28.708 31.314 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.250 30.469 qs_ot_get_p 128 10.4 0.001 0.002 29.021 29.263 multiply_cannon_sync_h2d 60168 15.6 26.455 28.527 26.455 28.527 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.321 25.079 apply_single 128 13.6 0.001 0.001 24.320 25.078 ot_diis_step 117 11.6 0.008 0.008 24.560 24.562 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.449 22.535 init_scf_loop 11 6.9 0.000 0.000 20.126 20.127 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.630 19.785 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.653 19.653 multiply_cannon_metrocomm3 60168 15.6 0.118 0.123 15.970 18.239 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.696 16.729 cp_fm_diag_elpa_base 83 14.4 16.626 16.656 16.691 16.723 prepare_preconditioner 11 7.9 0.000 0.000 15.534 15.592 make_preconditioner 11 8.9 0.000 0.000 15.534 15.592 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.766 14.938 init_scf_run 11 5.9 0.000 0.001 14.828 14.829 scf_env_initial_rho_setup 11 6.9 0.001 0.001 14.828 14.829 make_m2s 5014 13.6 0.105 0.112 13.885 14.291 make_images 5014 14.6 0.398 0.416 13.699 14.115 sum_up_and_integrate 128 10.3 0.002 0.004 13.697 13.713 integrate_v_rspace 128 11.3 0.003 0.004 13.638 13.659 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.606 12.747 calculate_rho_elec 128 8.7 0.045 0.064 12.606 12.746 dbcsr_mm_accdrv_process 124484 16.2 4.746 4.867 11.701 12.146 calculate_dm_sparse 128 9.5 0.001 0.001 10.791 10.893 mp_sum_l 7950 12.9 8.631 9.952 8.631 9.952 wfi_extrapolate 11 7.9 0.001 0.001 8.958 8.958 cp_fm_cholesky_invert 11 10.9 8.785 8.793 8.785 8.793 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.117 8.219 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.002 8.211 multiply_cannon_metrocomm1 60168 15.6 0.095 0.099 6.252 7.932 make_images_data 5014 15.6 0.065 0.071 6.807 7.723 grid_integrate_task_list 128 12.3 7.063 7.447 7.063 7.447 hybrid_alltoall_any 5200 16.5 0.294 2.259 5.957 7.239 density_rs2pw 128 9.7 0.005 0.006 6.244 6.903 fft_wrap_pw1pw2 1291 11.7 0.018 0.025 6.445 6.741 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 6.668 6.678 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.726 5.826 fft_wrap_pw1pw2_140 523 12.2 0.233 0.244 5.566 5.720 calculate_first_density_matrix 1 7.0 0.000 0.001 5.695 5.700 mp_alltoall_d11v 2415 14.1 4.291 5.610 4.291 5.610 fft3d_ps 1291 13.7 2.092 2.604 5.267 5.519 grid_collocate_task_list 128 9.7 4.838 5.262 4.838 5.262 jit_kernel_multiply 13 15.9 4.854 4.876 4.854 4.876 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 4.735 4.774 cp_fm_cholesky_decompose 22 10.9 4.591 4.615 4.591 4.615 potential_pw2rs 128 12.3 0.009 0.010 4.203 4.215 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=208.699000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.636364, yerr=6.197840 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 841.752576E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 4941574. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.158 0.292 192.700 192.701 qs_mol_dyn_low 1 2.0 0.018 0.068 190.764 190.789 qs_forces 11 3.9 0.004 0.005 190.639 190.647 qs_energies 11 4.9 0.020 0.029 183.891 183.909 scf_env_do_scf 11 5.9 0.001 0.002 164.283 164.293 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 131.656 131.659 velocity_verlet 10 3.0 0.002 0.006 117.961 117.967 dbcsr_multiply_generic 2507 12.6 0.185 0.192 99.325 100.546 qs_scf_new_mos 117 7.6 0.001 0.001 93.679 94.234 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.678 94.233 ot_scf_mini 117 9.6 0.004 0.005 88.870 89.486 multiply_cannon 2507 13.6 0.503 0.562 79.479 84.172 multiply_cannon_loop 2507 14.6 1.558 1.627 76.144 78.777 ot_mini 117 10.6 0.001 0.001 50.177 50.825 multiply_cannon_multrec 30084 15.6 20.912 25.503 33.711 38.670 mp_waitall_1 214728 16.6 23.875 38.088 23.875 38.088 rebuild_ks_matrix 128 8.3 0.001 0.001 32.310 32.950 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.028 32.309 32.949 init_scf_loop 11 6.9 0.001 0.003 32.538 32.539 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.077 29.673 qs_ot_get_derivative 117 11.6 0.001 0.002 28.234 28.848 multiply_cannon_metrocomm3 30084 15.6 0.096 0.101 15.719 28.453 prepare_preconditioner 11 7.9 0.000 0.000 28.041 28.114 make_preconditioner 11 8.9 0.000 0.002 28.041 28.114 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.732 27.248 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.033 23.236 apply_single 128 13.6 0.001 0.001 22.033 23.235 ot_diis_step 117 11.6 0.014 0.015 21.760 21.762 qs_ot_get_p 128 10.4 0.004 0.024 20.943 21.644 multiply_cannon_sync_h2d 30084 15.6 17.986 20.434 17.986 20.434 cp_fm_cholesky_invert 11 10.9 16.330 16.343 16.330 16.343 qs_ot_p2m_diag 83 11.4 0.189 0.217 15.883 15.920 make_m2s 5014 13.6 0.086 0.092 13.897 15.575 make_images 5014 14.6 1.161 1.349 13.681 15.357 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.690 14.691 init_scf_run 11 5.9 0.002 0.016 14.575 14.577 scf_env_initial_rho_setup 11 6.9 0.001 0.002 14.573 14.576 sum_up_and_integrate 128 10.3 0.003 0.006 13.705 13.731 integrate_v_rspace 128 11.3 0.003 0.004 13.644 13.675 dbcsr_mm_accdrv_process 62242 16.2 7.313 8.289 12.245 12.864 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.759 12.798 calculate_rho_elec 128 8.7 0.088 0.106 12.759 12.797 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.161 11.607 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.547 11.582 cp_fm_diag_elpa_base 83 14.4 11.295 11.399 11.541 11.572 multiply_cannon_metrocomm4 27577 15.6 0.106 0.122 3.861 10.821 mp_irecv_dv 69486 16.3 3.652 10.420 3.652 10.420 make_images_data 5014 15.6 0.065 0.074 8.178 10.144 hybrid_alltoall_any 5200 16.5 0.349 1.497 6.955 9.973 calculate_dm_sparse 128 9.5 0.001 0.001 8.572 8.747 wfi_extrapolate 11 7.9 0.004 0.022 8.281 8.282 grid_integrate_task_list 128 12.3 7.168 7.531 7.168 7.531 fft_wrap_pw1pw2 1291 11.7 0.019 0.025 7.350 7.434 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.313 7.097 cp_fm_cholesky_decompose 22 10.9 6.857 6.959 6.857 6.959 density_rs2pw 128 9.7 0.007 0.015 6.483 6.867 fft_wrap_pw1pw2_140 523 12.2 0.248 0.257 6.458 6.524 mp_sum_l 7950 12.9 4.205 6.374 4.205 6.374 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.275 6.286 calculate_first_density_matrix 1 7.0 0.004 0.030 6.099 6.104 fft3d_ps 1291 13.7 2.763 2.938 5.796 5.862 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.274 5.464 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.367 5.447 grid_collocate_task_list 128 9.7 5.006 5.375 5.006 5.375 mp_allgather_i34 2507 14.6 1.857 4.884 1.857 4.884 mp_alltoall_d11v 2415 14.1 4.088 4.736 4.088 4.736 jit_kernel_multiply 12 15.9 1.828 4.421 1.828 4.421 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 4.341 4.372 potential_pw2rs 128 12.3 0.016 0.029 4.192 4.208 mp_sum_d 4474 12.1 2.834 4.142 2.834 4.142 dbcsr_complete_redistribute 395 12.7 0.788 0.950 3.201 4.030 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=192.701000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=801.727273, yerr=2.339033 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 956.239872E+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 0.056 0.085 179.383 179.384 qs_mol_dyn_low 1 2.0 0.003 0.003 178.638 178.653 qs_forces 11 3.9 0.004 0.004 178.549 178.550 qs_energies 11 4.9 0.001 0.002 171.981 171.991 scf_env_do_scf 11 5.9 0.001 0.001 153.752 153.753 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 118.514 118.515 velocity_verlet 10 3.0 0.002 0.002 112.986 112.988 dbcsr_multiply_generic 2507 12.6 0.184 0.189 84.602 85.711 qs_scf_new_mos 117 7.6 0.001 0.001 81.529 81.828 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.528 81.827 ot_scf_mini 117 9.6 0.003 0.004 77.272 77.652 multiply_cannon 2507 13.6 0.501 0.521 64.515 69.445 multiply_cannon_loop 2507 14.6 1.124 1.195 61.506 63.941 ot_mini 117 10.6 0.001 0.001 42.876 43.269 init_scf_loop 11 6.9 0.000 0.000 35.133 35.134 mp_waitall_1 170520 16.6 24.548 33.932 24.548 33.932 prepare_preconditioner 11 7.9 0.000 0.000 31.104 31.150 make_preconditioner 11 8.9 0.000 0.000 31.104 31.150 rebuild_ks_matrix 128 8.3 0.001 0.001 30.447 30.895 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 30.446 30.895 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.742 30.165 multiply_cannon_multrec 20056 15.6 13.106 16.209 24.895 28.037 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.451 27.859 multiply_cannon_metrocomm3 20056 15.6 0.060 0.063 15.045 23.813 qs_ot_get_derivative 117 11.6 0.001 0.002 23.376 23.773 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.646 20.580 apply_single 128 13.6 0.001 0.001 19.646 20.580 qs_ot_get_p 128 10.4 0.001 0.002 19.762 20.180 ot_diis_step 117 11.6 0.018 0.018 19.389 19.389 make_m2s 5014 13.6 0.077 0.084 14.843 16.076 make_images 5014 14.6 1.145 1.235 14.607 15.836 qs_ot_p2m_diag 83 11.4 0.265 0.272 15.360 15.368 multiply_cannon_sync_h2d 20056 15.6 13.629 15.228 13.629 15.228 cp_fm_cholesky_invert 11 10.9 14.470 14.479 14.470 14.479 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.284 14.285 sum_up_and_integrate 128 10.3 0.002 0.003 13.803 13.828 integrate_v_rspace 128 11.3 0.003 0.004 13.744 13.769 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.215 13.246 calculate_rho_elec 128 8.7 0.132 0.146 13.214 13.245 init_scf_run 11 5.9 0.000 0.001 13.133 13.133 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.133 13.133 dbcsr_mm_accdrv_process 41502 16.2 8.073 8.345 11.244 11.363 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.034 11.053 cp_fm_diag_elpa_base 83 14.4 10.613 10.763 11.029 11.048 make_images_data 5014 15.6 0.061 0.071 8.821 10.540 hybrid_alltoall_any 5200 16.5 0.448 2.025 7.671 9.904 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.253 9.531 multiply_cannon_metrocomm4 17549 15.6 0.066 0.075 3.493 9.127 mp_irecv_dv 50230 16.2 3.363 8.868 3.363 8.868 calculate_dm_sparse 128 9.5 0.001 0.001 8.124 8.220 fft_wrap_pw1pw2 1291 11.7 0.019 0.025 7.694 7.794 grid_integrate_task_list 128 12.3 7.306 7.760 7.306 7.760 wfi_extrapolate 11 7.9 0.001 0.001 7.463 7.463 cp_fm_cholesky_decompose 22 10.9 7.381 7.420 7.381 7.420 cp_fm_upper_to_full 105 14.8 5.828 7.373 5.828 7.373 density_rs2pw 128 9.7 0.005 0.006 6.507 6.860 fft_wrap_pw1pw2_140 523 12.2 0.284 0.290 6.717 6.839 dbcsr_complete_redistribute 395 12.7 1.157 1.186 4.616 6.411 fft3d_ps 1291 13.7 2.761 2.988 6.002 6.071 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.589 5.598 grid_collocate_task_list 128 9.7 5.162 5.506 5.162 5.506 calculate_first_density_matrix 1 7.0 0.000 0.000 5.484 5.487 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.617 5.313 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.422 5.212 mp_alltoall_d11v 2415 14.1 4.469 4.933 4.469 4.933 mp_sum_l 7950 12.9 3.408 4.901 3.408 4.901 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.682 4.824 mp_allgather_i34 2507 14.6 1.516 4.583 1.516 4.583 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 4.423 4.448 potential_pw2rs 128 12.3 0.019 0.022 4.178 4.187 transfer_fm_to_dbcsr 11 9.9 0.019 0.024 2.343 4.084 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.946 3.980 mp_alltoall_i22 716 14.1 1.995 3.952 1.995 3.952 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.745 3.746 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=179.384000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=903.636364, yerr=8.402086 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.160618E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57949. MP_Allreduce 11127 1081. 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 0.040 0.134 191.465 191.471 qs_mol_dyn_low 1 2.0 0.014 0.090 190.299 190.313 qs_forces 11 3.9 0.007 0.015 190.084 190.098 qs_energies 11 4.9 0.044 0.142 183.000 183.018 scf_env_do_scf 11 5.9 0.001 0.001 162.436 162.447 velocity_verlet 10 3.0 0.020 0.078 123.931 123.934 scf_env_do_scf_inner_loop 117 6.6 0.004 0.008 116.472 116.473 dbcsr_multiply_generic 2507 12.6 0.207 0.226 83.247 83.959 qs_scf_new_mos 117 7.6 0.001 0.001 80.712 81.016 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.711 81.015 ot_scf_mini 117 9.6 0.003 0.004 75.957 76.268 multiply_cannon 2507 13.6 0.550 0.585 58.059 61.165 multiply_cannon_loop 2507 14.6 1.833 1.919 54.343 56.059 init_scf_loop 11 6.9 0.001 0.003 45.843 45.843 ot_mini 117 10.6 0.001 0.001 43.289 43.602 prepare_preconditioner 11 7.9 0.000 0.000 41.710 41.736 make_preconditioner 11 8.9 0.000 0.002 41.710 41.736 make_full_inverse_cholesky 11 9.9 0.013 0.023 35.335 40.329 multiply_cannon_multrec 30084 15.6 13.658 19.024 28.843 33.764 rebuild_ks_matrix 128 8.3 0.001 0.001 29.248 29.549 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 29.247 29.549 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.412 26.679 mp_waitall_1 147882 16.7 17.124 26.610 17.124 26.610 qs_ot_get_derivative 117 11.6 0.001 0.002 23.642 23.965 make_m2s 5014 13.6 0.092 0.096 20.746 21.902 make_images 5014 14.6 1.968 2.424 20.436 21.592 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.025 19.604 apply_single 128 13.6 0.001 0.001 19.024 19.604 ot_diis_step 117 11.6 0.018 0.019 19.517 19.518 qs_ot_get_p 128 10.4 0.010 0.012 18.833 19.209 cp_fm_upper_to_full 105 14.8 11.563 16.950 11.563 16.950 cp_fm_cholesky_invert 11 10.9 15.680 15.688 15.680 15.688 dbcsr_mm_accdrv_process 62264 16.2 10.221 11.888 14.750 15.215 multiply_cannon_metrocomm3 30084 15.6 0.050 0.053 6.204 14.766 qs_ot_p2m_diag 83 11.4 0.343 0.390 14.632 14.688 init_scf_run 11 5.9 0.001 0.014 13.998 14.001 scf_env_initial_rho_setup 11 6.9 0.011 0.013 13.997 14.001 sum_up_and_integrate 128 10.3 0.002 0.003 13.634 13.662 integrate_v_rspace 128 11.3 0.003 0.004 13.574 13.604 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.220 13.268 calculate_rho_elec 128 8.7 0.174 0.190 13.220 13.267 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.242 13.242 dbcsr_complete_redistribute 395 12.7 1.549 1.775 9.247 13.047 make_images_data 5014 15.6 0.064 0.070 11.049 12.811 copy_fm_to_dbcsr 209 11.7 0.002 0.006 7.922 11.704 hybrid_alltoall_any 5200 16.5 0.533 2.211 9.846 11.559 multiply_cannon_sync_h2d 30084 15.6 10.415 11.205 10.415 11.205 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 10.019 10.257 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.052 10.066 cp_fm_diag_elpa_base 83 14.4 9.076 9.386 10.045 10.059 transfer_fm_to_dbcsr 11 9.9 0.001 0.008 6.354 10.044 mp_alltoall_i22 716 14.1 5.553 9.365 5.553 9.365 calculate_dm_sparse 128 9.5 0.001 0.001 8.684 8.777 fft_wrap_pw1pw2 1291 11.7 0.019 0.023 7.743 7.858 grid_integrate_task_list 128 12.3 7.497 7.842 7.497 7.842 wfi_extrapolate 11 7.9 0.001 0.001 7.647 7.647 cp_fm_cholesky_decompose 22 10.9 7.396 7.503 7.396 7.503 multiply_cannon_metrocomm4 25070 15.6 0.079 0.090 2.813 7.357 mp_irecv_dv 76098 16.2 2.660 7.079 2.660 7.079 fft_wrap_pw1pw2_140 523 12.2 0.326 0.333 6.860 6.999 density_rs2pw 128 9.7 0.006 0.016 6.218 6.574 calculate_first_density_matrix 1 7.0 0.001 0.006 6.130 6.137 fft3d_ps 1291 13.7 2.950 3.009 5.924 5.973 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.807 5.842 grid_collocate_task_list 128 9.7 5.276 5.750 5.276 5.750 mp_alltoall_d11v 2415 14.1 5.108 5.715 5.108 5.715 qs_energies_init_hamiltonians 11 5.9 0.016 0.021 4.731 4.742 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.584 4.680 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 4.604 4.620 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.509 4.598 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.226 4.322 jit_kernel_multiply 11 15.8 1.694 4.267 1.694 4.267 potential_pw2rs 128 12.3 0.022 0.022 3.924 3.936 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=191.471000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1096.000000, yerr=19.423510 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 5.909858E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.561752E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591807936 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57807. MP_Allreduce 11152 1164. MP_Sync 87 MP_Alltoall 1724 18848449. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.048 171.056 171.057 qs_mol_dyn_low 1 2.0 0.003 0.003 170.515 170.550 qs_forces 11 3.9 0.004 0.004 170.419 170.423 qs_energies 11 4.9 0.004 0.007 163.039 163.046 scf_env_do_scf 11 5.9 0.001 0.001 142.984 142.999 velocity_verlet 10 3.0 0.002 0.002 110.733 110.737 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 107.821 107.823 dbcsr_multiply_generic 2527 12.6 0.185 0.192 74.405 74.814 qs_scf_new_mos 118 7.6 0.001 0.001 71.280 71.382 qs_scf_loop_do_ot 118 8.6 0.001 0.001 71.279 71.382 ot_scf_mini 118 9.6 0.003 0.004 66.838 66.913 multiply_cannon 2527 13.6 0.567 0.611 55.854 59.195 multiply_cannon_loop 2527 14.6 0.816 0.850 52.933 53.757 ot_mini 118 10.6 0.001 0.001 37.521 37.594 init_scf_loop 11 6.9 0.000 0.000 35.013 35.015 prepare_preconditioner 11 7.9 0.000 0.000 31.204 31.231 make_preconditioner 11 8.9 0.000 0.000 31.204 31.231 mp_waitall_1 126780 16.7 24.610 30.947 24.610 30.947 make_full_inverse_cholesky 11 9.9 0.016 0.026 29.116 29.389 rebuild_ks_matrix 129 8.3 0.001 0.001 29.154 29.236 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.153 29.236 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.614 26.682 multiply_cannon_multrec 10108 15.6 10.489 14.936 20.461 23.489 qs_ot_get_derivative 118 11.6 0.002 0.002 20.438 20.512 multiply_cannon_metrocomm3 10108 15.6 0.025 0.027 12.427 19.417 cp_fm_cholesky_invert 11 10.9 17.739 17.746 17.739 17.746 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.084 17.207 apply_single 129 13.6 0.001 0.001 17.084 17.207 ot_diis_step 118 11.6 0.020 0.020 17.015 17.015 qs_ot_get_p 129 10.4 0.002 0.002 16.544 16.632 make_m2s 5054 13.6 0.065 0.069 14.725 15.573 make_images 5054 14.6 2.178 2.613 14.413 15.252 sum_up_and_integrate 129 10.3 0.002 0.002 13.838 13.880 integrate_v_rspace 129 11.3 0.004 0.004 13.778 13.821 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.457 13.500 calculate_rho_elec 129 8.7 0.259 0.270 13.457 13.500 init_scf_run 11 5.9 0.000 0.001 12.963 12.964 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.963 12.964 qs_ot_p2m_diag 83 11.4 0.496 0.501 12.799 12.817 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.643 11.644 multiply_cannon_sync_h2d 10108 15.6 10.964 11.272 10.964 11.272 dbcsr_mm_accdrv_process 20926 16.1 3.559 6.447 9.619 10.331 hybrid_alltoall_any 5240 16.5 0.849 3.794 8.348 9.812 make_images_data 5054 15.6 0.055 0.066 8.480 9.733 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.793 8.803 cp_fm_diag_elpa_base 83 14.4 8.556 8.635 8.790 8.800 calculate_dm_sparse 129 9.5 0.001 0.001 8.391 8.454 grid_integrate_task_list 129 12.3 7.841 8.104 7.841 8.104 fft_wrap_pw1pw2 1301 11.7 0.018 0.019 8.073 8.093 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.947 7.999 cp_fm_cholesky_decompose 22 10.9 7.779 7.887 7.779 7.887 multiply_cannon_metrocomm1 10108 15.6 0.031 0.031 4.502 7.486 wfi_extrapolate 11 7.9 0.001 0.001 7.236 7.236 fft_wrap_pw1pw2_140 527 12.2 0.438 0.444 7.121 7.155 density_rs2pw 129 9.7 0.005 0.005 6.138 6.551 fft3d_ps 1301 13.7 3.136 3.246 6.036 6.053 grid_collocate_task_list 129 9.7 5.602 5.912 5.602 5.912 dbcsr_complete_redistribute 395 12.7 2.100 2.133 5.140 5.543 calculate_first_density_matrix 1 7.0 0.000 0.000 5.508 5.509 qs_energies_init_hamiltonians 11 5.9 0.025 0.049 5.488 5.492 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.147 5.162 mp_alltoall_d11v 2423 14.1 4.515 4.921 4.515 4.921 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 4.614 4.634 mp_allgather_i34 2527 14.6 1.098 4.569 1.098 4.569 jit_kernel_multiply 12 15.5 3.405 4.386 3.405 4.386 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.089 4.138 multiply_cannon_metrocomm4 7581 15.6 0.025 0.028 1.749 3.996 potential_pw2rs 129 12.3 0.026 0.027 3.923 3.929 mp_irecv_dv 29086 15.9 1.708 3.918 1.708 3.918 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.600 3.882 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.672 3.739 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.311 3.626 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.577 3.605 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.406 3.520 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.451 3.464 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=171.057000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1457.818182, yerr=49.239338 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 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 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 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 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.134423E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3943 59990. MP_Allreduce 10935 1507. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.068 0.074 290.000 290.000 qs_mol_dyn_low 1 2.0 0.003 0.003 288.960 288.988 qs_forces 11 3.9 0.005 0.005 288.825 288.826 qs_energies 11 4.9 0.002 0.002 280.067 280.071 scf_env_do_scf 11 5.9 0.001 0.001 254.924 254.932 velocity_verlet 10 3.0 0.042 0.044 207.417 207.425 scf_env_do_scf_inner_loop 116 6.6 0.004 0.008 130.839 130.841 init_scf_loop 11 6.9 0.000 0.000 123.798 123.800 prepare_preconditioner 11 7.9 0.000 0.000 119.079 119.098 make_preconditioner 11 8.9 0.000 0.000 119.079 119.098 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.047 116.210 qs_scf_new_mos 116 7.6 0.001 0.001 88.559 88.595 qs_scf_loop_do_ot 116 8.6 0.001 0.001 88.558 88.594 dbcsr_multiply_generic 2485 12.5 0.214 0.224 83.551 83.998 ot_scf_mini 116 9.6 0.003 0.004 83.771 83.778 cp_fm_upper_to_full 104 14.8 52.821 75.833 52.821 75.833 multiply_cannon 2485 13.5 0.684 0.728 60.764 62.289 multiply_cannon_loop 2485 14.5 1.028 1.046 56.798 57.955 ot_mini 116 10.6 0.001 0.001 43.723 43.733 dbcsr_complete_redistribute 393 12.7 3.983 4.071 30.007 43.288 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.495 39.716 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.989 36.982 mp_alltoall_i22 712 14.1 21.772 35.083 21.772 35.083 rebuild_ks_matrix 127 8.3 0.001 0.001 33.386 33.398 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 33.386 33.397 cp_fm_cholesky_invert 11 10.9 33.143 33.149 33.143 33.149 mp_waitall_1 102768 16.8 27.535 31.991 27.535 31.991 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.012 31.030 qs_ot_get_p 127 10.4 0.002 0.002 25.081 25.129 qs_ot_get_derivative 116 11.6 0.002 0.002 24.597 24.603 qs_ot_p2m_diag 82 11.4 0.868 0.873 21.094 21.119 multiply_cannon_multrec 9940 15.5 10.359 12.269 20.440 20.547 make_m2s 4970 13.5 0.073 0.076 18.321 19.798 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 18.541 19.687 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.334 19.337 make_images 4970 14.5 3.042 3.234 17.843 19.321 ot_diis_step 116 11.6 0.021 0.022 19.101 19.101 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.455 18.533 apply_single 127 13.6 0.001 0.001 18.455 18.532 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.213 16.213 cp_fm_diag_elpa_base 82 14.4 11.787 13.462 16.209 16.209 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.693 15.701 calculate_rho_elec 127 8.7 0.478 0.479 15.692 15.700 sum_up_and_integrate 127 10.3 0.002 0.002 15.593 15.686 integrate_v_rspace 127 11.3 0.004 0.005 15.530 15.624 init_scf_run 11 5.9 0.000 0.001 14.773 14.773 scf_env_initial_rho_setup 11 6.9 0.001 0.001 14.772 14.773 multiply_cannon_sync_h2d 9940 15.5 14.213 14.239 14.213 14.239 hybrid_alltoall_any 5155 16.4 1.305 3.050 10.593 12.849 make_images_data 4970 15.5 0.062 0.068 10.284 12.394 dbcsr_mm_accdrv_process 20590 16.0 4.416 6.332 9.832 11.838 fft_wrap_pw1pw2 1281 11.7 0.021 0.021 9.860 9.874 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.531 9.535 wfi_extrapolate 11 7.9 0.001 0.001 8.957 8.957 cp_fm_cholesky_decompose 22 10.9 8.930 8.950 8.930 8.950 calculate_dm_sparse 127 9.5 0.001 0.001 8.865 8.910 fft_wrap_pw1pw2_140 519 12.2 0.818 0.820 8.737 8.766 grid_integrate_task_list 127 12.3 8.500 8.691 8.500 8.691 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.171 8.172 fft3d_ps 1281 13.7 3.945 3.962 7.189 7.209 mp_alltoall_d11v 2401 14.1 6.919 7.140 6.919 7.140 density_rs2pw 127 9.7 0.005 0.005 6.768 6.814 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.585 6.616 grid_collocate_task_list 127 9.7 6.366 6.442 6.366 6.442 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.169 6.267 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=290.000000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2773.727273, yerr=185.886618 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 1.265652E+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 10245265. 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 0.133 0.446 87.148 87.153 qs_energies 1 2.0 0.000 0.001 84.626 84.643 ls_scf 1 3.0 0.000 0.000 83.684 83.704 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.631 72.834 multiply_cannon 111 7.7 0.017 0.020 55.912 57.469 multiply_cannon_loop 111 8.7 0.226 0.244 52.460 54.167 ls_scf_main 1 4.0 0.000 0.000 52.351 52.351 density_matrix_trs4 2 5.0 0.002 0.003 46.809 46.890 ls_scf_init_scf 1 4.0 0.000 0.000 28.282 28.290 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.127 27.197 mp_waitall_1 11031 10.9 22.593 25.169 22.593 25.169 matrix_sqrt_Newton_Schulz 2 6.5 0.027 0.051 24.998 25.020 multiply_cannon_multrec 2664 9.7 8.126 8.874 15.627 17.381 multiply_cannon_sync_h2d 2664 9.7 13.389 15.384 13.389 15.384 make_m2s 222 7.7 0.008 0.011 13.048 13.596 make_images 222 8.7 0.099 0.109 13.026 13.577 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.692 11.623 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.556 9.693 make_images_data 222 9.7 0.004 0.005 7.688 8.249 dbcsr_mm_accdrv_process 4760 10.4 0.595 0.702 7.120 8.106 hybrid_alltoall_any 227 10.6 0.215 1.828 6.598 7.868 dbcsr_mm_accdrv_process_sort 4760 11.4 6.324 7.215 6.324 7.215 calculate_norms 4752 9.8 5.488 6.105 5.488 6.105 mp_sum_l 887 5.1 3.215 5.251 3.215 5.251 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.052 5.190 multiply_cannon_metrocomm4 2442 9.7 0.011 0.015 2.050 4.216 mp_irecv_dv 6231 10.9 2.034 4.187 2.034 4.187 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.362 3.842 arnoldi_extremal 4 6.8 0.000 0.000 3.232 3.276 arnoldi_normal_ev 4 7.8 0.020 0.058 3.232 3.276 make_images_sizes 222 9.7 0.000 0.000 0.699 3.235 mp_alltoall_i44 222 10.7 0.698 3.235 0.698 3.235 build_subspace 16 8.4 0.009 0.012 3.092 3.095 ls_scf_post 1 4.0 0.000 0.000 3.051 3.066 ls_scf_store_result 1 5.0 0.000 0.000 2.854 2.904 dbcsr_special_finalize 555 9.7 0.005 0.006 2.302 2.719 dbcsr_merge_single_wm 555 10.7 0.460 0.605 2.294 2.710 make_images_pack 222 9.7 2.212 2.630 2.214 2.632 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.317 2.557 dbcsr_matrix_vector_mult_local 304 10.0 2.062 2.453 2.064 2.455 dbcsr_sort_data 658 11.4 2.092 2.445 2.092 2.445 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.281 2.379 buffer_matrices_ensure_size 222 8.7 1.757 2.072 1.757 2.072 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.805 1.806 rebuild_ks_matrix 3 7.3 0.000 0.000 1.795 1.797 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.795 1.797 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=87.153000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1136.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 2.177724E+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 0.115 0.184 90.388 90.411 qs_energies 1 2.0 0.000 0.000 89.387 89.391 ls_scf 1 3.0 0.000 0.000 88.027 88.032 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.181 74.498 multiply_cannon 111 7.7 0.028 0.041 52.665 57.457 ls_scf_main 1 4.0 0.000 0.000 54.304 54.307 multiply_cannon_loop 111 8.7 0.135 0.146 49.891 53.548 density_matrix_trs4 2 5.0 0.002 0.003 48.628 48.862 ls_scf_init_scf 1 4.0 0.000 0.000 30.048 30.050 mp_waitall_1 9105 10.9 20.395 29.545 20.395 29.545 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.878 28.958 multiply_cannon_multrec 1332 9.7 13.359 16.947 22.769 27.512 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.469 26.481 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.235 20.602 make_m2s 222 7.7 0.006 0.008 14.981 15.496 make_images 222 8.7 1.375 1.691 14.950 15.466 dbcsr_mm_accdrv_process 4041 10.4 0.330 0.539 9.005 10.615 dbcsr_mm_accdrv_process_sort 4041 11.4 8.523 10.076 8.523 10.076 hybrid_alltoall_any 227 10.6 0.542 2.556 7.996 9.538 make_images_data 222 9.7 0.004 0.004 8.647 9.493 mp_sum_l 887 5.1 5.195 8.759 5.195 8.759 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.194 7.998 mp_irecv_dv 3311 11.0 3.173 7.935 3.173 7.935 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.864 6.996 calculate_norms 2376 9.8 6.070 6.754 6.070 6.754 multiply_cannon_sync_h2d 1332 9.7 4.927 6.174 4.927 6.174 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.897 5.102 arnoldi_extremal 4 6.8 0.000 0.000 4.630 4.648 arnoldi_normal_ev 4 7.8 0.001 0.005 4.630 4.648 build_subspace 16 8.4 0.014 0.021 4.379 4.382 ls_scf_post 1 4.0 0.000 0.000 3.676 3.681 ls_scf_store_result 1 5.0 0.000 0.000 3.381 3.500 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.154 3.365 dbcsr_matrix_vector_mult_local 304 10.0 2.747 3.223 2.749 3.225 mp_allgather_i34 111 8.7 0.760 2.945 0.760 2.945 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.605 2.688 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.161 2.562 dbcsr_data_new 4174 10.1 2.114 2.404 2.114 2.404 make_images_pack 222 9.7 1.820 2.120 1.822 2.122 dbcsr_sort_data 436 11.2 1.831 2.042 1.831 2.042 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.880 1.883 rebuild_ks_matrix 3 7.3 0.000 0.000 1.868 1.871 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.868 1.871 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.411000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1783.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 2.853306E+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 0.036 0.103 92.227 92.231 qs_energies 1 2.0 0.000 0.000 91.436 91.441 ls_scf 1 3.0 0.000 0.000 89.982 89.986 dbcsr_multiply_generic 111 6.7 0.016 0.022 74.669 75.055 multiply_cannon 111 7.7 0.042 0.132 51.813 56.947 ls_scf_main 1 4.0 0.000 0.000 56.197 56.203 multiply_cannon_loop 111 8.7 0.116 0.129 49.103 52.312 density_matrix_trs4 2 5.0 0.002 0.003 50.273 50.431 mp_waitall_1 7281 11.0 23.182 32.301 23.182 32.301 ls_scf_init_scf 1 4.0 0.000 0.000 30.114 30.128 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.940 28.996 matrix_sqrt_Newton_Schulz 2 6.5 0.012 0.024 26.616 26.631 multiply_cannon_multrec 888 9.7 12.647 15.006 21.350 24.487 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.890 21.914 make_m2s 222 7.7 0.006 0.008 16.337 17.038 make_images 222 8.7 1.586 1.866 16.299 17.001 make_images_data 222 9.7 0.004 0.004 9.491 10.482 hybrid_alltoall_any 227 10.6 0.640 2.945 8.962 10.122 dbcsr_mm_accdrv_process 3754 10.4 0.323 0.495 8.218 9.477 dbcsr_mm_accdrv_process_sort 3754 11.4 7.783 8.982 7.783 8.982 mp_sum_l 887 5.1 4.865 8.193 4.865 8.193 multiply_cannon_sync_h2d 888 9.7 6.075 7.330 6.075 7.330 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.461 7.182 mp_irecv_dv 2335 11.1 2.445 7.125 2.445 7.125 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.441 6.538 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.682 6.388 arnoldi_extremal 4 6.8 0.000 0.000 5.170 5.193 arnoldi_normal_ev 4 7.8 0.063 0.088 5.170 5.192 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.778 5.006 calculate_norms 1584 9.8 4.407 4.840 4.407 4.840 build_subspace 16 8.4 0.014 0.020 4.787 4.793 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.453 3.798 mp_allgather_i34 111 8.7 0.833 3.718 0.833 3.718 ls_scf_post 1 4.0 0.000 0.000 3.671 3.678 dbcsr_matrix_vector_mult_local 304 10.0 3.032 3.614 3.034 3.616 ls_scf_store_result 1 5.0 0.000 0.000 3.405 3.490 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.828 2.926 dbcsr_data_new 4116 9.9 2.103 2.468 2.103 2.468 make_images_sizes 222 9.7 0.000 0.000 1.010 2.315 mp_alltoall_i44 222 10.7 1.010 2.315 1.010 2.315 dbcsr_sort_data 325 11.1 1.865 2.136 1.865 2.136 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.917 1.928 rebuild_ks_matrix 3 7.3 0.000 0.000 1.898 1.910 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.898 1.910 make_images_pack 222 9.7 1.624 1.883 1.627 1.886 dbcsr_finalize 304 7.8 0.026 0.032 1.600 1.865 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.231000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2274.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 3.387720E+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 0.089 0.110 98.584 98.590 qs_energies 1 2.0 0.001 0.008 97.360 97.373 ls_scf 1 3.0 0.004 0.016 95.265 95.282 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.862 79.095 ls_scf_main 1 4.0 0.001 0.012 59.089 59.092 multiply_cannon 111 7.7 0.040 0.089 51.816 56.217 density_matrix_trs4 2 5.0 0.002 0.004 52.993 53.120 multiply_cannon_loop 111 8.7 0.152 0.166 46.807 50.112 ls_scf_init_scf 1 4.0 0.000 0.001 32.873 32.877 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.501 31.581 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 29.002 29.013 mp_waitall_1 6369 11.0 22.787 28.438 22.787 28.438 multiply_cannon_multrec 1332 9.7 14.057 17.384 22.093 24.981 make_m2s 222 7.7 0.006 0.008 21.137 22.447 make_images 222 8.7 3.141 3.601 21.087 22.399 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.265 17.488 make_images_data 222 9.7 0.004 0.004 11.736 13.326 hybrid_alltoall_any 227 10.6 0.796 3.787 11.144 12.979 dbcsr_mm_accdrv_process 3641 10.4 0.288 0.485 7.668 9.216 dbcsr_mm_accdrv_process_sort 3641 11.4 7.226 8.720 7.226 8.720 mp_sum_l 887 5.1 4.352 8.466 4.352 8.466 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.333 6.753 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.115 6.150 mp_irecv_dv 3229 10.9 2.090 6.064 2.090 6.064 multiply_cannon_sync_h2d 1332 9.7 5.397 5.917 5.397 5.917 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.616 5.678 arnoldi_extremal 4 6.8 0.000 0.000 5.212 5.224 arnoldi_normal_ev 4 7.8 0.001 0.004 5.212 5.224 build_subspace 16 8.4 0.014 0.021 4.877 4.886 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.542 4.750 calculate_norms 2376 9.8 4.213 4.598 4.213 4.598 mp_allgather_i34 111 8.7 2.120 4.359 2.120 4.359 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 3.597 3.888 dbcsr_matrix_vector_mult_local 304 10.0 3.205 3.703 3.207 3.705 dbcsr_sort_data 658 11.4 3.069 3.432 3.069 3.432 ls_scf_post 1 4.0 0.000 0.000 3.300 3.314 dbcsr_special_finalize 555 9.7 0.006 0.007 2.830 3.275 dbcsr_merge_single_wm 555 10.7 0.540 0.666 2.821 3.267 ls_scf_store_result 1 5.0 0.000 0.000 3.000 3.077 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.977 3.035 dbcsr_data_release 10477 10.7 1.577 2.380 1.577 2.380 qs_energies_init_hamiltonians 1 3.0 0.014 0.022 2.083 2.084 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.043 2.048 rebuild_ks_matrix 3 7.3 0.000 0.000 2.020 2.025 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.006 2.020 2.025 dbcsr_finalize 304 7.8 0.049 0.061 1.797 1.988 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.590000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2801.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 4.803584E+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 0.103 0.107 93.350 93.351 qs_energies 1 2.0 0.000 0.000 92.057 92.069 ls_scf 1 3.0 0.000 0.000 90.082 90.094 dbcsr_multiply_generic 111 6.7 0.017 0.019 71.601 71.848 ls_scf_main 1 4.0 0.000 0.000 57.200 57.200 multiply_cannon 111 7.7 0.078 0.161 53.053 56.103 multiply_cannon_loop 111 8.7 0.088 0.095 50.442 52.340 density_matrix_trs4 2 5.0 0.002 0.003 50.251 50.329 ls_scf_init_scf 1 4.0 0.000 0.000 29.468 29.470 mp_waitall_1 5436 11.0 24.938 29.127 24.938 29.127 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.242 28.285 matrix_sqrt_Newton_Schulz 2 6.5 0.004 0.012 26.157 26.163 multiply_cannon_multrec 444 9.7 13.734 16.235 20.909 21.974 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.713 15.777 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.592 15.257 make_m2s 222 7.7 0.005 0.005 13.671 14.643 make_images 222 8.7 2.039 2.480 13.604 14.574 hybrid_alltoall_any 227 10.6 0.807 3.873 8.262 9.946 make_images_data 222 9.7 0.003 0.004 8.454 9.818 dbcsr_mm_accdrv_process 3003 10.4 0.349 0.397 6.873 8.016 multiply_cannon_sync_h2d 444 9.7 6.690 7.992 6.690 7.992 dbcsr_mm_accdrv_process_sort 3003 11.4 6.512 7.619 6.512 7.619 arnoldi_extremal 4 6.8 0.000 0.000 5.908 5.926 arnoldi_normal_ev 4 7.8 0.009 0.019 5.908 5.926 build_subspace 16 8.4 0.015 0.019 5.500 5.510 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.403 4.623 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.192 4.397 mp_sum_l 887 5.1 2.875 4.373 2.875 4.373 dbcsr_matrix_vector_mult_local 304 10.0 3.731 4.201 3.733 4.203 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.565 3.809 mp_allgather_i34 111 8.7 1.189 3.795 1.189 3.795 mp_irecv_dv 1241 11.2 1.549 3.766 1.549 3.766 calculate_norms 792 9.8 3.613 3.735 3.613 3.735 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.486 3.559 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.034 3.530 ls_scf_post 1 4.0 0.000 0.000 3.414 3.426 make_images_sizes 222 9.7 0.000 0.000 0.823 3.260 mp_alltoall_i44 222 10.7 0.823 3.259 0.823 3.259 ls_scf_store_result 1 5.0 0.000 0.000 3.185 3.226 dbcsr_data_new 4608 9.7 1.807 2.300 1.807 2.300 dbcsr_finalize 304 7.8 0.062 0.077 2.207 2.264 dbcsr_merge_all 275 8.9 0.482 0.532 2.062 2.108 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.028 2.029 rebuild_ks_matrix 3 7.3 0.000 0.000 1.995 1.996 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.995 1.996 qs_energies_init_hamiltonians 1 3.0 0.016 0.036 1.959 1.960 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=93.351000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3735.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 9.056772E+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 0.163 0.183 106.168 106.168 qs_energies 1 2.0 0.000 0.000 104.055 104.059 ls_scf 1 3.0 0.000 0.000 101.045 101.053 dbcsr_multiply_generic 111 6.7 0.024 0.028 74.869 75.040 ls_scf_main 1 4.0 0.000 0.000 63.621 63.622 density_matrix_trs4 2 5.0 0.002 0.003 54.847 54.921 multiply_cannon 111 7.7 0.128 0.227 48.467 50.388 multiply_cannon_loop 111 8.7 0.098 0.099 45.462 46.115 ls_scf_init_scf 1 4.0 0.000 0.000 33.608 33.614 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.207 32.226 matrix_sqrt_Newton_Schulz 2 6.5 0.036 0.044 29.477 29.488 mp_waitall_1 4527 11.1 21.713 25.843 21.713 25.843 make_m2s 222 7.7 0.005 0.005 22.640 23.747 make_images 222 8.7 3.579 3.878 22.533 23.639 multiply_cannon_multrec 444 9.7 17.841 18.562 22.559 23.287 hybrid_alltoall_any 227 10.6 1.658 3.615 12.700 15.748 make_images_data 222 9.7 0.003 0.004 12.989 15.179 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.158 10.908 multiply_cannon_sync_h2d 444 9.7 8.775 8.817 8.775 8.817 arnoldi_extremal 4 6.8 0.000 0.000 7.514 7.523 arnoldi_normal_ev 4 7.8 0.078 0.088 7.514 7.523 build_subspace 16 8.4 0.026 0.036 6.887 6.901 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.524 5.691 dbcsr_matrix_vector_mult_local 304 10.0 5.113 5.435 5.116 5.437 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.885 5.144 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.907 4.996 dbcsr_mm_accdrv_process 1814 10.4 0.243 0.340 4.532 4.660 dbcsr_mm_accdrv_process_sort 1814 11.4 4.198 4.331 4.198 4.331 ls_scf_post 1 4.0 0.000 0.000 3.815 3.818 make_images_sizes 222 9.7 0.000 0.000 1.490 3.767 mp_alltoall_i44 222 10.7 1.490 3.766 1.490 3.766 mp_allgather_i34 111 8.7 1.124 3.664 1.124 3.664 ls_scf_store_result 1 5.0 0.000 0.000 3.505 3.544 calculate_norms 792 9.8 3.247 3.288 3.247 3.288 dbcsr_finalize 304 7.8 0.082 0.089 3.093 3.178 qs_energies_init_hamiltonians 1 3.0 0.019 0.047 2.979 2.988 dbcsr_merge_all 275 8.9 0.886 0.924 2.873 2.951 dbcsr_complete_redistribute 5 7.6 1.433 1.499 2.749 2.888 matrix_ls_to_qs 2 6.0 0.000 0.000 2.403 2.540 dbcsr_sort_data 325 11.1 2.446 2.498 2.446 2.498 dbcsr_data_new 6591 9.6 1.873 2.337 1.873 2.337 dbcsr_new_transposed 4 7.5 0.242 0.252 2.297 2.311 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.280 2.282 mp_sum_l 887 5.1 1.364 2.241 1.364 2.241 dbcsr_frobenius_norm 74 6.6 2.055 2.136 2.190 2.222 rebuild_ks_matrix 3 7.3 0.000 0.000 2.214 2.216 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.214 2.216 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.211 dbcsr_add_anytype 103 7.2 0.858 0.890 2.132 2.211 dbcsr_data_release 12724 10.6 1.988 2.201 1.988 2.201 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.168000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7124.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/0f470484747789dc0da415bf6a25bd77b7b21d8e_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 585.654272E+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 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 0.046 0.119 228.831 228.836 qs_mol_dyn_low 1 2.0 0.046 0.172 227.516 227.540 qs_forces 5 3.8 0.004 0.020 227.208 227.229 qs_energies 5 4.8 0.053 0.131 224.302 224.354 scf_env_do_scf 5 5.8 0.001 0.009 207.536 207.539 scf_env_do_scf_inner_loop 105 6.6 0.008 0.053 180.848 180.866 qs_scf_new_mos 105 7.6 0.000 0.001 141.695 141.857 qs_scf_loop_do_ot 105 8.6 0.001 0.001 141.694 141.856 dbcsr_multiply_generic 1445 12.2 0.142 0.154 134.121 134.515 ot_scf_mini 105 9.6 0.004 0.022 131.875 131.988 multiply_cannon 1445 13.2 0.273 0.285 115.440 117.729 multiply_cannon_loop 1445 14.2 2.838 2.987 113.750 114.989 velocity_verlet 4 3.0 0.024 0.102 105.117 105.138 ot_mini 105 10.6 0.001 0.001 59.729 59.852 multiply_cannon_multrec 69360 15.2 29.704 34.528 41.685 46.756 qs_ot_get_p 112 10.4 0.001 0.001 42.279 42.588 mp_waitall_1 488190 16.1 34.515 41.714 34.515 41.714 qs_ot_get_derivative 55 11.6 0.001 0.001 38.047 38.207 multiply_cannon_sync_h2d 69360 15.2 29.370 33.403 29.370 33.403 multiply_cannon_metrocomm3 69360 15.2 0.201 0.212 25.325 32.846 qs_ot_p2m_diag 40 11.0 0.020 0.030 31.431 31.542 rebuild_ks_matrix 110 8.4 0.000 0.000 28.426 28.609 qs_ks_build_kohn_sham_matrix 110 9.4 0.013 0.047 28.425 28.608 cp_dbcsr_syevd 40 12.0 0.002 0.003 28.199 28.199 init_scf_loop 7 6.6 0.001 0.041 26.657 26.659 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.135 26.305 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.991 23.221 apply_single 62 13.6 0.000 0.000 22.991 23.221 cp_fm_syevd 40 13.0 0.000 0.000 23.088 23.208 prepare_preconditioner 7 7.6 0.000 0.000 21.771 21.799 make_preconditioner 7 8.6 0.002 0.057 21.771 21.799 ot_new_cg_direction 55 11.6 0.001 0.001 20.983 20.983 cp_fm_redistribute_end 40 14.0 9.166 18.285 9.171 18.286 cp_fm_syevd_base 40 14.0 9.108 18.228 9.108 18.228 qs_rho_update_rho_low 110 7.6 0.001 0.001 16.348 16.709 calculate_rho_elec 110 8.6 0.030 0.045 16.348 16.709 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.669 14.893 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.513 14.583 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.792 13.878 init_scf_run 5 5.8 0.000 0.003 13.837 13.838 scf_env_initial_rho_setup 5 6.8 0.002 0.030 13.837 13.838 calculate_dm_sparse 110 9.5 0.000 0.001 13.379 13.563 mp_sum_l 4764 12.2 11.858 12.732 11.858 12.732 dbcsr_mm_accdrv_process 154766 15.8 6.266 6.450 11.847 12.610 density_rs2pw 110 9.6 0.004 0.005 10.665 11.214 fft_wrap_pw1pw2 1425 12.5 0.019 0.026 10.606 10.786 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.092 10.161 cp_fm_cholesky_invert 7 10.6 10.083 10.090 10.083 10.090 qs_vxc_create 110 10.4 0.002 0.003 9.824 9.843 check_diag 80 13.5 8.704 8.935 9.577 9.699 fft_wrap_pw1pw2_240 915 14.0 0.199 0.231 9.285 9.415 calculate_first_density_matrix 1 7.0 0.001 0.034 9.031 9.041 multiply_cannon_metrocomm1 69360 15.2 0.099 0.105 4.657 8.488 fft3d_pb 915 15.0 2.258 2.446 8.240 8.409 sum_up_and_integrate 60 10.3 0.001 0.010 8.349 8.367 integrate_v_rspace 60 11.3 0.002 0.010 8.331 8.351 transfer_rs2pw 445 10.6 0.008 0.009 7.712 8.314 acc_transpose_blocks 69360 15.2 0.349 0.368 6.799 7.565 xc_rho_set_and_dset_create 110 12.4 0.072 0.078 6.980 7.207 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.010 7.102 7.117 make_full_single_inverse 7 9.6 0.001 0.032 6.972 7.002 make_m2s 2890 13.2 0.079 0.088 6.348 6.929 make_images 2890 14.2 0.237 0.258 6.239 6.819 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.524 6.593 xc_vxc_pw_create 60 11.3 0.036 0.038 6.544 6.567 xc_pw_derive 510 13.4 0.005 0.006 5.706 5.761 mp_alltoall_z22v 2340 16.7 5.420 5.606 5.420 5.606 mp_waitany 7680 13.5 4.475 5.258 4.475 5.258 jit_kernel_multiply 13 15.2 4.711 4.757 4.711 4.757 acc_transpose_blocks_kernels 69360 16.2 0.854 0.890 4.053 4.712 multiply_cannon_metrocomm4 67915 15.2 0.181 0.194 2.004 4.666 potential_pw2rs 60 12.3 0.002 0.003 4.605 4.632 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=228.836000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=557.600000, yerr=1.959592 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 0f470484747789dc0da415bf6a25bd77b7b21d8e Summary: empty Status: OK