=== This is the CP2K Performance-Test === Updating 6a02f6195..737e746da Fast-forward src/input_cp2k_projection_rtp.F | 3 +- .../Dockerfile.2023.2_mpich_generic_psmp | 27 +++-- .../Dockerfile.2023.2_mpich_haswell_psmp | 27 +++-- .../production/Dockerfile.2023.2_mpich_native_psmp | 27 +++-- .../Dockerfile.2023.2_mpich_skylake-avx512_psmp | 27 +++-- .../production/Dockerfile.2023.2_mpich_znver2_psmp | 27 +++-- .../production/Dockerfile.2023.2_mpich_znver3_psmp | 27 +++-- .../Dockerfile.2023.2_openmpi_generic_psmp | 25 ++-- .../Dockerfile.2023.2_openmpi_haswell_psmp | 25 ++-- .../Dockerfile.2023.2_openmpi_native_psmp | 25 ++-- .../Dockerfile.2023.2_openmpi_skylake-avx512_psmp | 25 ++-- .../Dockerfile.2023.2_openmpi_znver2_psmp | 25 ++-- .../Dockerfile.2023.2_openmpi_znver3_psmp | 25 ++-- .../Dockerfile.master_intelmpi_generic_psmp | 102 ++++++++++++++++ .../Dockerfile.master_intelmpi_haswell_psmp | 102 ++++++++++++++++ .../Dockerfile.master_intelmpi_native_psmp | 102 ++++++++++++++++ .../Dockerfile.master_intelmpi_skylake-avx512_psmp | 102 ++++++++++++++++ .../Dockerfile.master_intelmpi_znver2_psmp | 102 ++++++++++++++++ .../Dockerfile.master_intelmpi_znver3_psmp | 102 ++++++++++++++++ .../Dockerfile.master_mpich_generic_psmp | 27 +++-- .../Dockerfile.master_mpich_haswell_psmp | 27 +++-- .../production/Dockerfile.master_mpich_native_psmp | 27 +++-- .../Dockerfile.master_mpich_skylake-avx512_psmp | 27 +++-- .../production/Dockerfile.master_mpich_znver2_psmp | 27 +++-- .../production/Dockerfile.master_mpich_znver3_psmp | 27 +++-- .../Dockerfile.master_openmpi_generic_psmp | 25 ++-- .../Dockerfile.master_openmpi_haswell_psmp | 25 ++-- .../Dockerfile.master_openmpi_native_psmp | 25 ++-- .../Dockerfile.master_openmpi_skylake-avx512_psmp | 25 ++-- .../Dockerfile.master_openmpi_znver2_psmp | 25 ++-- .../Dockerfile.master_openmpi_znver3_psmp | 25 ++-- tools/docker/production/README.md | 4 +- tools/docker/production/generate_docker_files.py | 128 +++++++++++---------- 33 files changed, 1058 insertions(+), 313 deletions(-) create mode 100644 tools/docker/production/Dockerfile.master_intelmpi_generic_psmp create mode 100644 tools/docker/production/Dockerfile.master_intelmpi_haswell_psmp create mode 100644 tools/docker/production/Dockerfile.master_intelmpi_native_psmp create mode 100644 tools/docker/production/Dockerfile.master_intelmpi_skylake-avx512_psmp create mode 100644 tools/docker/production/Dockerfile.master_intelmpi_znver2_psmp create mode 100644 tools/docker/production/Dockerfile.master_intelmpi_znver3_psmp Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 737e746daa79aac850212d2d3a1eec9f85e831b5 ################# 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, SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 04.10.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/01 job id: 49387715 --- 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/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/02 job id: 49387720 --- Point --- name: 20 plot: h2o_32_ri_rpa_mp2 regex: Total MP2 Time= label: RI-MP2 (8n/6r/2t) --- Point --- name: 21 plot: h2o_32_ri_rpa_mp2_mem regex: Estimated peak process memory label: RI-MP2 (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/03 job id: 49387722 --- Point --- name: 100 plot: h2o_64_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 101 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/04 job id: 49387725 --- Point --- name: 102 plot: h2o_64_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 103 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/05 job id: 49387728 --- Point --- name: 104 plot: h2o_64_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 105 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/06 job id: 49387730 --- Point --- name: 106 plot: h2o_64_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 107 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/07 job id: 49387731 --- Point --- name: 108 plot: h2o_64_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 109 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/08 job id: 49387732 --- 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/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/09 job id: 49387733 --- 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/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/10 job id: 49387734 --- 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/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/11 job id: 49387735 --- 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/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/12 job id: 49387736 --- 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/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/13 job id: 49387737 --- 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/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/14 job id: 49387738 --- 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/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/15 job id: 49387739 --- 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/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/16 job id: 49387740 --- 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/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/17 job id: 49387741 --- 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/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/18 job id: 49387742 --- 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/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/19 job id: 49387743 --- 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/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/20 job id: 49387744 --- Point --- name: 410 plot: h2o_256_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 411 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/21 job id: 49387745 --- Point --- name: 500 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/12r/1t) --- Point --- name: 501 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/22 job id: 49387746 --- Point --- name: 502 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/6r/2t) --- Point --- name: 503 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/23 job id: 49387747 --- Point --- name: 504 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/4r/3t) --- Point --- name: 505 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/24 job id: 49387748 --- Point --- name: 506 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/3r/4t) --- Point --- name: 507 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/25 job id: 49387749 --- Point --- name: 508 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/2r/6t) --- Point --- name: 509 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/26 job id: 49387750 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/27 job id: 49387751 --- 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/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.035 133.084 133.085 farming_run 1 2.0 132.592 132.593 133.055 133.058 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.492611E+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 0.023 114.935 114.935 qs_energies 1 2.0 0.000 0.000 114.711 114.712 mp2_main 1 3.0 0.000 0.000 112.602 112.603 mp2_gpw_main 1 4.0 0.028 0.033 111.566 111.567 mp2_ri_gpw_compute_in 1 5.0 0.172 0.176 92.564 92.815 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.782 55.033 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.162 41.177 46.133 get_2c_integrals 1 6.0 0.008 0.009 36.721 37.610 integrate_v_rspace 273 8.0 0.437 0.453 24.744 29.583 pw_transfer 6555 10.6 0.375 0.391 26.850 27.198 grid_integrate_task_list 273 9.0 20.636 25.964 20.636 25.964 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.515 25.928 fft_wrap_pw1pw2_100 2178 12.4 1.133 1.268 23.083 23.471 rpa_ri_compute_en 1 5.0 0.019 0.020 18.885 19.152 compute_2c_integrals 1 7.0 0.029 0.061 19.084 19.085 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.543 18.759 mp2_eri_2c_integrate_gpw 1 9.0 2.394 2.425 18.541 18.755 cp_fm_cholesky_decompose 12 8.2 17.575 18.481 17.575 18.481 cholesky_decomp 1 7.0 0.000 0.000 16.476 17.379 fft3d_s 5443 13.4 16.150 16.469 16.172 16.491 ao_to_mo_and_store_B_mult_1 272 7.0 10.770 15.323 10.770 15.323 calculate_wavefunction 272 8.0 5.399 5.549 12.349 13.010 rpa_num_int 1 6.0 0.000 0.002 10.656 10.665 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.634 10.635 calc_mat_Q 8 8.0 0.000 0.000 9.484 9.564 contract_S_to_Q 8 9.0 0.000 0.000 8.904 8.985 calc_potential_gpw 544 9.5 0.005 0.005 8.181 8.671 parallel_gemm_fm 14 9.1 0.000 0.000 8.490 8.579 parallel_gemm_fm_cosma 14 10.1 8.490 8.579 8.490 8.579 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.038 8.277 potential_pw2rs 545 10.0 0.107 0.108 7.507 8.157 create_integ_mat 1 6.0 0.014 0.026 7.886 7.895 collocate_single_gaussian 272 10.0 0.039 0.041 7.309 7.550 array2fm 1 7.0 0.000 0.000 6.837 7.371 pw_scatter_s 2720 13.7 4.360 4.517 4.360 4.517 pw_gather_s 2722 13.2 3.477 3.789 3.477 3.789 array2fm_buffer_send 1 8.0 2.972 3.177 2.972 3.177 pw_poisson_solve 545 10.5 1.113 1.190 2.154 2.345 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.565977, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2804.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.036 420.867 420.868 farming_run 1 2.0 419.965 419.969 420.832 420.834 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.233969E+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 15104666. 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 0.011 0.030 216.875 216.876 qs_energies 1 2.0 0.000 0.000 216.660 216.668 scf_env_do_scf 1 3.0 0.000 0.000 114.752 114.752 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.815 113.823 rebuild_ks_matrix 4 6.0 0.000 0.000 113.814 113.821 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.062 113.814 113.821 hfx_ks_matrix 4 8.0 0.001 0.001 113.457 113.463 integrate_four_center 4 9.0 0.153 0.457 113.456 113.462 integrate_four_center_main 4 10.0 0.137 0.562 102.186 105.398 integrate_four_center_bin 265 11.0 102.049 105.283 102.049 105.283 mp2_main 1 3.0 0.000 0.000 101.596 101.604 mp2_gpw_main 1 4.0 0.046 0.067 100.724 100.736 init_scf_loop 1 4.0 0.000 0.000 96.725 96.725 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.009 75.068 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.788 54.846 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.161 41.505 46.522 integrate_v_rspace 95 8.0 0.397 0.564 27.983 32.858 pw_transfer 2240 10.6 0.143 0.169 29.527 29.937 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.520 29.028 mp2_ri_gpw_compute_en 1 5.0 0.054 0.062 26.555 28.661 ao_to_mo_and_store_B_mult_1 91 7.0 10.610 28.632 10.610 28.632 grid_integrate_task_list 95 9.0 23.399 28.469 23.399 28.469 fft_wrap_pw1pw2_100 730 12.4 1.312 1.498 26.258 26.773 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.844 1.896 24.847 24.857 get_2c_integrals 1 6.0 0.000 0.000 20.118 20.158 compute_2c_integrals 1 7.0 0.002 0.003 19.104 19.108 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.717 18.979 mp2_eri_2c_integrate_gpw 1 9.0 1.731 1.862 18.716 18.978 fft3d_s 1823 13.4 18.471 18.862 18.485 18.876 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.025 18.025 calculate_wavefunction 91 8.0 2.017 2.044 9.626 9.881 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.574 8.770 9.205 potential_pw2rs 186 10.0 0.033 0.035 8.422 8.975 local_gemm 172 8.0 8.214 8.647 8.214 8.647 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.204 8.516 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.519 7.853 8.248 calc_potential_gpw 182 9.5 0.002 0.002 7.951 8.138 collocate_single_gaussian 91 10.0 0.017 0.021 7.791 7.951 integrate_four_center_load 4 10.0 0.000 0.000 6.765 6.769 hfx_load_balance 1 11.0 0.000 0.000 6.764 6.769 mp2_ri_gpw_compute_en_ener 172 7.0 6.335 6.399 6.335 6.399 mp_sendrecv_dm3 2068 8.0 5.887 6.267 5.887 6.267 mp_sync 37 10.5 3.258 5.951 3.258 5.951 pw_gather_s 912 13.2 4.470 4.957 4.470 4.957 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.728477, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1495.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.902912E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.060 53.976 53.987 qs_mol_dyn_low 1 2.0 0.003 0.003 52.942 53.760 qs_forces 11 3.9 0.002 0.003 52.615 52.616 qs_energies 11 4.9 0.002 0.002 51.115 51.138 scf_env_do_scf 11 5.9 0.000 0.001 45.198 45.198 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 43.142 43.142 qs_scf_new_mos 108 7.5 0.000 0.001 33.319 33.621 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.318 33.621 dbcsr_multiply_generic 2286 12.5 0.093 0.098 33.098 33.479 ot_scf_mini 108 9.5 0.002 0.002 31.685 31.881 multiply_cannon 2286 13.5 0.195 0.212 25.868 27.470 multiply_cannon_loop 2286 14.5 1.820 1.928 25.202 26.840 velocity_verlet 10 3.0 0.002 0.002 25.775 25.778 ot_mini 108 10.5 0.001 0.001 19.118 19.369 qs_ot_get_derivative 108 11.5 0.001 0.001 16.216 16.413 mp_waitall_1 245248 16.5 8.088 14.144 8.088 14.144 multiply_cannon_metrocomm3 54864 15.5 0.073 0.079 5.819 12.421 multiply_cannon_multrec 54864 15.5 3.665 5.924 7.588 11.040 qs_ot_get_p 119 10.4 0.001 0.001 7.969 8.263 rebuild_ks_matrix 119 8.3 0.000 0.000 7.752 7.889 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.752 7.889 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.818 6.943 mp_sum_l 7287 12.8 5.285 6.802 5.285 6.802 multiply_cannon_sync_h2d 54864 15.5 5.196 6.261 5.196 6.261 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.378 5.860 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.629 5.733 dbcsr_mm_accdrv_process 76910 16.1 1.832 2.889 3.836 5.496 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.185 5.244 init_scf_run 11 5.9 0.000 0.001 4.646 4.646 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.646 4.646 sum_up_and_integrate 119 10.3 0.001 0.001 4.394 4.404 integrate_v_rspace 119 11.3 0.002 0.002 4.383 4.394 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.386 4.386 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.193 4.193 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.026 4.173 calculate_rho_elec 119 8.7 0.011 0.017 4.025 4.172 cp_fm_redistribute_end 50 14.0 2.140 4.168 2.146 4.171 cp_fm_diag_elpa_base 50 14.0 2.019 4.061 2.023 4.070 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.900 3.107 apply_single 119 13.6 0.000 0.000 2.900 3.106 calculate_dm_sparse 119 9.5 0.000 0.000 2.851 3.004 acc_transpose_blocks 54864 15.5 0.233 0.261 2.226 2.749 multiply_cannon_metrocomm1 54864 15.5 0.057 0.063 1.625 2.722 ot_diis_step 108 11.5 0.006 0.006 2.666 2.666 jit_kernel_multiply 13 15.8 1.941 2.527 1.941 2.527 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.397 2.451 calculate_first_density_matrix 1 7.0 0.000 0.001 2.432 2.437 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.238 2.240 density_rs2pw 119 9.7 0.004 0.004 2.070 2.177 wfi_extrapolate 11 7.9 0.001 0.001 2.145 2.145 grid_integrate_task_list 119 12.3 2.013 2.116 2.013 2.116 init_scf_loop 11 6.9 0.000 0.000 2.039 2.039 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.940 1.997 mp_sum_d 4135 12.0 1.253 1.924 1.253 1.924 potential_pw2rs 119 12.3 0.004 0.004 1.765 1.776 make_m2s 4572 13.5 0.053 0.055 1.633 1.677 pw_transfer 1439 11.6 0.051 0.059 1.594 1.669 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.520 1.596 make_images 4572 14.5 0.134 0.139 1.551 1.594 acc_transpose_blocks_sync 164592 16.5 1.211 1.450 1.211 1.450 mp_alltoall_d11v 2130 13.8 1.294 1.440 1.294 1.440 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.378 1.418 grid_collocate_task_list 119 9.7 1.345 1.414 1.345 1.414 mp_waitany 12084 13.8 1.237 1.403 1.237 1.403 transfer_rs2pw 487 10.6 0.005 0.006 1.315 1.396 fft3d_ps 1201 14.6 0.366 0.470 1.234 1.307 transfer_pw2rs 487 13.2 0.005 0.006 1.298 1.305 fft_wrap_pw1pw2_140 487 13.2 0.141 0.155 1.182 1.261 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.987000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.272727, yerr=1.052349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_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 488.517632E+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 130672. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.026 39.087 39.088 qs_mol_dyn_low 1 2.0 0.003 0.003 38.778 38.786 qs_forces 11 3.9 0.002 0.002 38.719 38.721 qs_energies 11 4.9 0.001 0.002 37.046 37.051 scf_env_do_scf 11 5.9 0.000 0.001 31.876 31.877 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.454 29.454 dbcsr_multiply_generic 2286 12.5 0.100 0.102 22.191 22.531 qs_scf_new_mos 108 7.5 0.001 0.001 20.777 21.009 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.777 21.008 ot_scf_mini 108 9.5 0.002 0.003 19.840 20.001 velocity_verlet 10 3.0 0.001 0.006 18.447 18.448 multiply_cannon 2286 13.5 0.211 0.217 16.964 18.379 multiply_cannon_loop 2286 14.5 1.188 1.251 15.786 17.416 ot_mini 108 10.5 0.001 0.001 12.182 12.418 mp_waitall_1 200699 16.5 5.562 10.859 5.562 10.859 qs_ot_get_derivative 108 11.5 0.001 0.001 9.808 9.976 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.019 9.553 multiply_cannon_multrec 27432 15.5 1.829 4.048 6.207 9.128 rebuild_ks_matrix 119 8.3 0.000 0.000 6.804 6.947 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.804 6.947 dbcsr_mm_accdrv_process 47894 16.0 3.558 5.930 4.294 6.388 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.986 6.115 qs_ot_get_p 119 10.4 0.001 0.001 4.705 4.934 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.650 4.560 mp_sum_l 7287 12.8 2.112 4.121 2.112 4.121 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.072 4.044 apply_single 119 13.6 0.000 0.000 3.072 4.044 init_scf_run 11 5.9 0.000 0.001 3.980 3.980 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.979 3.980 sum_up_and_integrate 119 10.3 0.001 0.001 3.723 3.728 integrate_v_rspace 119 11.3 0.002 0.003 3.709 3.713 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.651 3.677 calculate_rho_elec 119 8.7 0.021 0.024 3.651 3.677 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.048 3.069 make_m2s 4572 13.5 0.052 0.053 2.664 2.989 make_images 4572 14.5 0.206 0.245 2.576 2.904 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.644 2.644 calculate_first_density_matrix 1 7.0 0.000 0.000 2.440 2.442 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.341 2.426 init_scf_loop 11 6.9 0.000 0.000 2.401 2.401 ot_diis_step 108 11.5 0.011 0.011 2.323 2.324 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.308 2.308 cp_fm_redistribute_end 50 14.0 1.170 2.281 1.174 2.284 cp_fm_diag_elpa_base 50 14.0 1.077 2.183 1.105 2.228 multiply_cannon_sync_h2d 27432 15.5 1.700 2.183 1.700 2.183 calculate_dm_sparse 119 9.5 0.000 0.001 2.075 2.150 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.051 2.053 density_rs2pw 119 9.7 0.004 0.004 1.940 2.040 jit_kernel_multiply 10 16.4 0.678 2.033 0.678 2.033 acc_transpose_blocks 27432 15.5 0.115 0.121 1.632 1.983 grid_integrate_task_list 119 12.3 1.839 1.934 1.839 1.934 pw_transfer 1439 11.6 0.064 0.068 1.866 1.898 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.801 1.839 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.775 1.810 make_images_data 4572 15.5 0.047 0.054 1.249 1.590 prepare_preconditioner 11 7.9 0.000 0.000 1.507 1.532 make_preconditioner 11 8.9 0.000 0.000 1.507 1.532 mp_alltoall_d11v 2130 13.8 1.232 1.491 1.232 1.491 wfi_extrapolate 11 7.9 0.001 0.001 1.480 1.480 potential_pw2rs 119 12.3 0.006 0.007 1.466 1.474 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.404 1.461 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.109 1.442 fft3d_ps 1201 14.6 0.517 0.571 1.410 1.440 fft_wrap_pw1pw2_140 487 13.2 0.160 0.174 1.392 1.430 grid_collocate_task_list 119 9.7 1.293 1.356 1.293 1.356 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.217 1.266 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.247 1.255 transfer_rs2pw 487 10.6 0.005 0.005 1.090 1.162 mp_sum_d 4135 12.0 0.542 1.022 0.542 1.022 mp_allgather_i34 2286 14.5 0.599 0.993 0.599 0.993 acc_transpose_blocks_sync 82296 16.5 0.813 0.940 0.813 0.940 transfer_pw2rs 487 13.2 0.004 0.005 0.923 0.930 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.926 0.928 acc_transpose_blocks_kernels 27432 16.5 0.189 0.278 0.679 0.900 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.875 0.890 make_images_sizes 4572 15.5 0.005 0.005 0.544 0.787 mp_alltoall_i44 4572 16.5 0.539 0.782 0.539 0.782 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.088000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.090909, yerr=1.564059 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 520.667136E+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 0.013 0.081 32.001 32.052 qs_mol_dyn_low 1 2.0 0.003 0.003 31.777 31.785 qs_forces 11 3.9 0.002 0.002 31.712 31.716 qs_energies 11 4.9 0.002 0.002 30.165 30.170 scf_env_do_scf 11 5.9 0.000 0.001 25.515 25.515 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 23.083 23.083 dbcsr_multiply_generic 2286 12.5 0.100 0.103 16.856 16.933 qs_scf_new_mos 108 7.5 0.001 0.001 15.412 15.435 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.411 15.434 velocity_verlet 10 3.0 0.001 0.002 15.133 15.135 ot_scf_mini 108 9.5 0.002 0.002 14.654 14.668 multiply_cannon 2286 13.5 0.193 0.197 13.643 14.305 multiply_cannon_loop 2286 14.5 0.860 0.902 12.872 13.566 ot_mini 108 10.5 0.001 0.001 9.037 9.049 qs_ot_get_derivative 108 11.5 0.001 0.001 7.556 7.570 multiply_cannon_multrec 18288 15.5 1.883 2.941 6.950 7.287 rebuild_ks_matrix 119 8.3 0.000 0.000 5.981 5.999 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.980 5.998 dbcsr_mm_accdrv_process 38222 16.0 4.914 5.704 4.973 5.768 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.265 5.282 mp_waitall_1 158411 16.6 2.625 3.724 2.625 3.724 sum_up_and_integrate 119 10.3 0.001 0.001 3.569 3.576 integrate_v_rspace 119 11.3 0.002 0.003 3.556 3.564 init_scf_run 11 5.9 0.000 0.001 3.481 3.481 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.481 3.481 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.860 3.445 qs_ot_get_p 119 10.4 0.001 0.001 3.381 3.400 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.264 3.268 calculate_rho_elec 119 8.7 0.031 0.031 3.264 3.268 init_scf_loop 11 6.9 0.000 0.000 2.416 2.416 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.438 2.402 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.994 2.375 apply_single 119 13.6 0.000 0.000 1.994 2.375 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.253 2.258 calculate_first_density_matrix 1 7.0 0.000 0.000 2.245 2.247 make_m2s 4572 13.5 0.044 0.045 1.817 1.958 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.955 1.956 density_rs2pw 119 9.7 0.004 0.004 1.830 1.929 grid_integrate_task_list 119 12.3 1.805 1.901 1.805 1.901 make_images 4572 14.5 0.193 0.204 1.732 1.872 calculate_dm_sparse 119 9.5 0.000 0.001 1.841 1.852 pw_transfer 1439 11.6 0.065 0.067 1.813 1.824 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.721 1.734 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.715 1.717 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.702 1.709 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.697 1.706 cp_fm_diag_elpa_base 50 14.0 1.673 1.686 1.695 1.704 acc_transpose_blocks 18288 15.5 0.081 0.083 1.639 1.690 prepare_preconditioner 11 7.9 0.000 0.000 1.663 1.665 make_preconditioner 11 8.9 0.000 0.000 1.663 1.665 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.527 1.606 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.516 1.524 ot_diis_step 108 11.5 0.011 0.012 1.468 1.468 mp_sum_l 7287 12.8 1.005 1.397 1.005 1.397 fft_wrap_pw1pw2_140 487 13.2 0.212 0.218 1.383 1.395 potential_pw2rs 119 12.3 0.007 0.009 1.359 1.363 grid_collocate_task_list 119 9.7 1.234 1.310 1.234 1.310 fft3d_ps 1201 14.6 0.525 0.540 1.290 1.299 multiply_cannon_sync_h2d 18288 15.5 1.036 1.203 1.036 1.203 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.181 1.185 wfi_extrapolate 11 7.9 0.001 0.001 1.180 1.180 transfer_rs2pw 487 10.6 0.005 0.005 1.000 1.095 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.940 0.956 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.944 0.947 make_images_data 4572 15.5 0.047 0.051 0.771 0.933 hybrid_alltoall_any 4725 16.4 0.058 0.115 0.664 0.845 transfer_pw2rs 487 13.2 0.004 0.004 0.837 0.840 mp_alltoall_d11v 2130 13.8 0.673 0.837 0.673 0.837 acc_transpose_blocks_kernels 18288 16.5 0.218 0.224 0.795 0.827 acc_transpose_blocks_sync 54864 16.5 0.744 0.795 0.744 0.795 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.785 0.787 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.676 0.740 cp_fm_cholesky_invert 11 10.9 0.707 0.710 0.707 0.710 mp_alltoall_z22v 1201 16.6 0.614 0.684 0.614 0.684 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.052000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.181818, yerr=1.748671 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_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 558.723072E+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 0.013 0.031 36.317 36.319 qs_mol_dyn_low 1 2.0 0.003 0.004 36.115 36.123 qs_forces 11 3.9 0.002 0.002 36.044 36.046 qs_energies 11 4.9 0.004 0.022 34.339 34.347 scf_env_do_scf 11 5.9 0.002 0.014 29.302 29.303 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.902 25.903 dbcsr_multiply_generic 2286 12.5 0.098 0.101 19.821 19.951 velocity_verlet 10 3.0 0.002 0.002 18.427 18.429 qs_scf_new_mos 108 7.5 0.001 0.001 17.872 17.923 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.871 17.922 ot_scf_mini 108 9.5 0.002 0.003 16.842 16.899 multiply_cannon 2286 13.5 0.218 0.227 16.064 16.475 multiply_cannon_loop 2286 14.5 1.556 1.640 15.115 15.450 ot_mini 108 10.5 0.001 0.001 10.329 10.401 multiply_cannon_multrec 27432 15.5 2.453 3.138 8.997 9.294 qs_ot_get_derivative 108 11.5 0.001 0.001 8.468 8.525 dbcsr_mm_accdrv_process 47916 15.9 6.012 7.662 6.440 7.764 rebuild_ks_matrix 119 8.3 0.000 0.000 6.282 6.344 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.282 6.343 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.565 5.620 qs_ot_get_p 119 10.4 0.001 0.001 3.666 3.744 init_scf_run 11 5.9 0.000 0.001 3.667 3.667 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.666 3.667 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.058 3.549 sum_up_and_integrate 119 10.3 0.001 0.001 3.461 3.468 integrate_v_rspace 119 11.3 0.002 0.003 3.449 3.456 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.350 3.391 calculate_rho_elec 119 8.7 0.040 0.046 3.349 3.391 init_scf_loop 11 6.9 0.001 0.003 3.366 3.366 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.189 2.683 apply_single 119 13.6 0.000 0.000 2.188 2.683 acc_transpose_blocks 27432 15.5 0.121 0.127 2.449 2.606 prepare_preconditioner 11 7.9 0.000 0.000 2.538 2.546 make_preconditioner 11 8.9 0.000 0.002 2.538 2.546 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.149 2.472 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.352 2.380 make_m2s 4572 13.5 0.054 0.056 2.256 2.377 calculate_first_density_matrix 1 7.0 0.001 0.008 2.268 2.269 make_images 4572 14.5 0.275 0.336 2.148 2.269 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.251 2.263 calculate_dm_sparse 119 9.5 0.000 0.000 2.143 2.196 mp_waitall_1 137007 16.6 1.646 2.177 1.646 2.177 pw_transfer 1439 11.6 0.064 0.068 1.908 1.942 density_rs2pw 119 9.7 0.004 0.004 1.796 1.891 grid_integrate_task_list 119 12.3 1.820 1.889 1.820 1.889 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.817 1.853 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.844 1.845 ot_diis_step 108 11.5 0.012 0.012 1.819 1.820 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.805 1.806 mp_sum_l 7287 12.8 1.086 1.697 1.086 1.697 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.631 1.643 acc_transpose_blocks_sync 82296 16.5 1.480 1.633 1.480 1.633 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.571 1.579 cp_fm_diag_elpa_base 50 14.0 1.537 1.553 1.569 1.578 fft_wrap_pw1pw2_140 487 13.2 0.246 0.260 1.499 1.539 fft3d_ps 1201 14.6 0.553 0.604 1.336 1.359 wfi_extrapolate 11 7.9 0.001 0.001 1.343 1.343 grid_collocate_task_list 119 9.7 1.248 1.335 1.248 1.335 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.290 1.306 potential_pw2rs 119 12.3 0.009 0.010 1.270 1.273 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.241 1.251 cp_fm_upper_to_full 72 14.2 0.821 1.162 0.821 1.162 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.683 1.161 jit_kernel_multiply 6 15.9 0.363 1.101 0.363 1.101 qs_energies_init_hamiltonians 11 5.9 0.013 0.016 1.098 1.099 transfer_rs2pw 487 10.6 0.004 0.005 0.923 1.025 dbcsr_complete_redistribute 329 12.2 0.118 0.143 0.748 1.020 make_images_data 4572 15.5 0.047 0.052 0.828 0.952 hybrid_alltoall_any 4725 16.4 0.065 0.154 0.740 0.918 mp_alltoall_d11v 2130 13.8 0.767 0.895 0.767 0.895 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.879 acc_transpose_blocks_kernels 27432 16.5 0.271 0.279 0.819 0.827 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.550 0.821 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.807 0.813 cp_fm_cholesky_invert 11 10.9 0.741 0.744 0.741 0.744 mp_alltoall_i22 627 13.8 0.444 0.741 0.444 0.741 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.319000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.090909, yerr=3.203820 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_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 622.964736E+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 0.012 0.036 28.765 28.770 qs_mol_dyn_low 1 2.0 0.003 0.003 28.510 28.518 qs_forces 11 3.9 0.002 0.002 27.988 27.992 qs_energies 11 4.9 0.001 0.002 26.292 26.297 scf_env_do_scf 11 5.9 0.000 0.001 21.464 21.464 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.032 19.033 velocity_verlet 10 3.0 0.023 0.031 14.881 14.911 dbcsr_multiply_generic 2286 12.5 0.090 0.095 13.122 13.186 qs_scf_new_mos 108 7.5 0.001 0.001 11.551 11.575 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.550 11.574 multiply_cannon 2286 13.5 0.223 0.230 10.637 11.104 ot_scf_mini 108 9.5 0.002 0.002 10.851 10.873 multiply_cannon_loop 2286 14.5 0.637 0.653 9.726 9.863 multiply_cannon_multrec 9144 15.5 1.735 1.931 6.260 6.432 ot_mini 108 10.5 0.001 0.001 6.248 6.275 rebuild_ks_matrix 119 8.3 0.000 0.000 5.667 5.695 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.667 5.695 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.040 5.065 qs_ot_get_derivative 108 11.5 0.001 0.001 4.971 4.995 dbcsr_mm_accdrv_process 12550 15.8 3.726 4.438 4.409 4.480 init_scf_run 11 5.9 0.000 0.001 3.396 3.396 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.395 3.396 sum_up_and_integrate 119 10.3 0.001 0.001 3.362 3.367 integrate_v_rspace 119 11.3 0.003 0.003 3.352 3.357 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.260 3.267 calculate_rho_elec 119 8.7 0.060 0.061 3.259 3.267 qs_ot_get_p 119 10.4 0.001 0.001 2.741 2.773 init_scf_loop 11 6.9 0.000 0.000 2.413 2.414 calculate_first_density_matrix 1 7.0 0.000 0.000 2.285 2.285 pw_transfer 1439 11.6 0.065 0.067 1.917 1.928 grid_integrate_task_list 119 12.3 1.863 1.925 1.863 1.925 make_m2s 4572 13.5 0.034 0.034 1.746 1.903 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.825 1.837 make_images 4572 14.5 0.271 0.304 1.657 1.812 calculate_dm_sparse 119 9.5 0.000 0.000 1.782 1.798 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.783 1.785 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.744 1.763 mp_waitall_1 115863 16.7 1.205 1.724 1.205 1.724 density_rs2pw 119 9.7 0.003 0.004 1.647 1.709 prepare_preconditioner 11 7.9 0.000 0.000 1.702 1.707 make_preconditioner 11 8.9 0.000 0.000 1.702 1.707 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.698 1.699 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.593 1.620 jit_kernel_multiply 8 15.8 0.644 1.603 0.644 1.603 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.553 1.563 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.542 1.543 fft_wrap_pw1pw2_140 487 13.2 0.324 0.332 1.512 1.524 acc_transpose_blocks 9144 15.5 0.042 0.043 1.462 1.486 grid_collocate_task_list 119 9.7 1.295 1.348 1.295 1.348 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.312 1.326 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.257 1.265 ot_diis_step 108 11.5 0.013 0.013 1.264 1.265 cp_fm_diag_elpa_base 50 14.0 1.231 1.246 1.256 1.264 fft3d_ps 1201 14.6 0.555 0.570 1.238 1.246 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.202 1.241 apply_single 119 13.6 0.000 0.000 1.202 1.240 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.222 1.225 potential_pw2rs 119 12.3 0.010 0.011 1.180 1.183 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.178 1.183 wfi_extrapolate 11 7.9 0.001 0.001 1.061 1.061 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.758 0.989 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.876 0.934 make_images_data 4572 15.5 0.042 0.046 0.755 0.929 mp_alltoall_d11v 2130 13.8 0.770 0.845 0.770 0.845 cp_fm_cholesky_invert 11 10.9 0.841 0.844 0.841 0.844 transfer_rs2pw 487 10.6 0.004 0.004 0.755 0.837 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.801 0.809 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.747 0.750 acc_transpose_blocks_sync 27432 16.5 0.709 0.731 0.709 0.731 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.671 0.721 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.695 0.697 mp_allgather_i34 2286 14.5 0.229 0.650 0.229 0.650 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.310 0.633 transfer_pw2rs 487 13.2 0.003 0.004 0.629 0.631 mp_alltoall_z22v 1201 16.6 0.556 0.585 0.556 0.585 jit_kernel_transpose 5 15.6 0.577 0.579 0.577 0.579 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.770000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=592.363636, yerr=4.051324 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_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 773.201920E+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 0.028 0.046 42.237 42.238 qs_mol_dyn_low 1 2.0 0.003 0.003 42.011 42.020 qs_forces 11 3.9 0.005 0.007 41.624 41.625 qs_energies 11 4.9 0.002 0.002 39.618 39.622 scf_env_do_scf 11 5.9 0.001 0.001 34.048 34.048 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.291 26.293 velocity_verlet 10 3.0 0.002 0.002 24.020 24.066 dbcsr_multiply_generic 2286 12.5 0.098 0.100 18.713 18.944 qs_scf_new_mos 108 7.5 0.001 0.001 17.002 17.098 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.002 17.097 ot_scf_mini 108 9.5 0.002 0.002 15.847 15.948 multiply_cannon 2286 13.5 0.296 0.302 14.935 15.859 multiply_cannon_loop 2286 14.5 0.849 0.874 13.693 14.646 ot_mini 108 10.5 0.001 0.001 9.649 9.764 multiply_cannon_multrec 9144 15.5 3.360 4.726 8.796 8.891 init_scf_loop 11 6.9 0.000 0.000 7.730 7.732 qs_ot_get_derivative 108 11.5 0.001 0.001 7.622 7.723 rebuild_ks_matrix 119 8.3 0.000 0.000 7.038 7.180 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.038 7.179 prepare_preconditioner 11 7.9 0.000 0.000 6.784 6.799 make_preconditioner 11 8.9 0.000 0.000 6.784 6.799 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.338 6.681 dbcsr_mm_accdrv_process 12550 15.8 4.624 6.507 5.289 6.540 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.340 6.469 cp_fm_upper_to_full 72 14.2 3.208 4.602 3.208 4.602 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.968 3.974 calculate_rho_elec 119 8.7 0.118 0.121 3.968 3.973 sum_up_and_integrate 119 10.3 0.001 0.001 3.764 3.771 integrate_v_rspace 119 11.3 0.003 0.003 3.754 3.761 init_scf_run 11 5.9 0.000 0.001 3.529 3.529 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.529 3.529 qs_ot_get_p 119 10.4 0.001 0.001 3.304 3.439 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.810 3.199 mp_waitall_1 94719 16.7 2.163 3.138 2.163 3.138 dbcsr_complete_redistribute 329 12.2 0.288 0.292 1.970 2.781 pw_transfer 1439 11.6 0.068 0.068 2.530 2.536 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.674 2.488 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.169 2.473 apply_single 119 13.6 0.000 0.000 2.169 2.472 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.432 2.438 make_m2s 4572 13.5 0.037 0.038 2.271 2.412 make_images 4572 14.5 0.357 0.391 2.152 2.293 calculate_dm_sparse 119 9.5 0.000 0.000 2.273 2.290 mp_alltoall_i22 627 13.8 1.420 2.271 1.420 2.271 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.441 2.248 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.289 2.195 grid_integrate_task_list 119 12.3 2.087 2.151 2.087 2.151 calculate_first_density_matrix 1 7.0 0.000 0.000 2.093 2.094 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.028 2.080 fft_wrap_pw1pw2_140 487 13.2 0.574 0.577 2.063 2.069 density_rs2pw 119 9.7 0.003 0.003 2.002 2.028 ot_diis_step 108 11.5 0.014 0.014 2.001 2.002 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.954 1.957 acc_transpose_blocks 9144 15.5 0.044 0.044 1.782 1.838 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.772 1.773 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.763 1.765 mp_sum_l 7287 12.8 0.966 1.739 0.966 1.739 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.600 1.601 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.531 1.572 grid_collocate_task_list 119 9.7 1.526 1.565 1.526 1.565 fft3d_ps 1201 14.6 0.592 0.605 1.531 1.535 cp_fm_cholesky_invert 11 10.9 1.384 1.387 1.384 1.387 wfi_extrapolate 11 7.9 0.001 0.001 1.372 1.372 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.324 1.325 cp_fm_diag_elpa_base 50 14.0 1.178 1.229 1.323 1.323 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.297 1.317 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.280 1.286 potential_pw2rs 119 12.3 0.014 0.015 1.252 1.253 hybrid_alltoall_any 4725 16.4 0.089 0.150 1.014 1.207 mp_alltoall_d11v 2130 13.8 1.143 1.192 1.143 1.192 acc_transpose_blocks_sync 27432 16.5 1.077 1.128 1.077 1.128 make_images_data 4572 15.5 0.046 0.049 0.940 1.119 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.096 1.111 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.996 1.056 qs_create_task_list 11 7.9 0.000 0.000 0.927 0.937 generate_qs_task_list 11 8.9 0.367 0.386 0.927 0.937 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.908 0.921 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.238000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=724.727273, yerr=14.232427 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_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 505.028608E+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 1958955. 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 0.007 0.026 79.191 79.192 qs_mol_dyn_low 1 2.0 0.003 0.003 78.878 78.886 qs_forces 11 3.9 0.003 0.003 78.809 78.810 qs_energies 11 4.9 0.002 0.002 75.986 75.999 scf_env_do_scf 11 5.9 0.000 0.001 67.294 67.296 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 61.939 61.939 dbcsr_multiply_generic 2055 12.4 0.105 0.107 49.394 49.642 qs_scf_new_mos 99 7.5 0.000 0.001 45.179 45.324 qs_scf_loop_do_ot 99 8.5 0.001 0.001 45.178 45.323 ot_scf_mini 99 9.5 0.002 0.002 42.832 42.918 multiply_cannon 2055 13.4 0.186 0.191 41.826 42.658 multiply_cannon_loop 2055 14.4 1.761 1.808 40.909 41.766 velocity_verlet 10 3.0 0.001 0.002 41.516 41.517 ot_mini 99 10.5 0.001 0.001 25.121 25.252 qs_ot_get_derivative 99 11.5 0.001 0.001 18.376 18.494 multiply_cannon_multrec 49320 15.4 11.615 12.292 17.468 18.295 rebuild_ks_matrix 110 8.3 0.000 0.001 14.339 14.490 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.011 14.339 14.490 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.568 12.708 mp_waitall_1 220248 16.4 9.739 10.685 9.739 10.685 multiply_cannon_sync_h2d 49320 15.4 9.850 10.324 9.850 10.324 qs_ot_get_p 110 10.4 0.001 0.001 9.341 9.426 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.231 7.723 apply_single 110 13.6 0.000 0.000 7.231 7.722 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.881 7.437 multiply_cannon_metrocomm3 49320 15.4 0.084 0.086 5.953 7.119 sum_up_and_integrate 110 10.3 0.002 0.003 6.881 6.895 integrate_v_rspace 110 11.3 0.003 0.003 6.855 6.875 init_scf_run 11 5.9 0.000 0.001 6.664 6.665 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.664 6.664 ot_diis_step 99 11.5 0.006 0.006 6.589 6.589 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.167 6.332 calculate_rho_elec 110 8.6 0.020 0.024 6.166 6.332 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.228 6.243 dbcsr_mm_accdrv_process 87628 16.1 2.917 3.040 5.721 5.994 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.460 5.460 init_scf_loop 11 6.9 0.000 0.000 5.329 5.330 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.169 5.220 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.906 4.936 cp_fm_diag_elpa_base 48 14.0 4.894 4.923 4.905 4.935 mp_sum_l 6594 12.7 3.282 4.270 3.282 4.270 wfi_extrapolate 11 7.9 0.001 0.001 4.015 4.015 make_m2s 4110 13.4 0.060 0.066 3.827 3.929 make_images 4110 14.4 0.177 0.188 3.732 3.838 calculate_dm_sparse 110 9.5 0.001 0.001 3.714 3.811 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.519 3.523 grid_integrate_task_list 110 12.3 3.252 3.445 3.252 3.445 density_rs2pw 110 9.6 0.004 0.005 3.220 3.440 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.192 3.243 prepare_preconditioner 11 7.9 0.000 0.000 3.202 3.218 make_preconditioner 11 8.9 0.000 0.000 3.202 3.218 multiply_cannon_metrocomm1 49320 15.4 0.066 0.069 2.035 3.198 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.067 3.120 pw_transfer 1331 11.6 0.054 0.062 2.958 3.031 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.977 3.023 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.871 2.946 acc_transpose_blocks 49320 15.4 0.211 0.220 2.508 2.588 calculate_first_density_matrix 1 7.0 0.000 0.001 2.550 2.556 fft_wrap_pw1pw2_140 451 13.1 0.392 0.430 2.474 2.553 jit_kernel_multiply 13 15.9 2.529 2.545 2.529 2.545 potential_pw2rs 110 12.3 0.005 0.006 2.495 2.507 mp_alltoall_d11v 2046 13.8 2.016 2.437 2.016 2.437 grid_collocate_task_list 110 9.6 2.155 2.284 2.155 2.284 fft3d_ps 1111 14.6 0.795 0.884 2.165 2.219 transfer_rs2pw 451 10.6 0.005 0.006 1.824 2.015 mp_waitany 14300 13.8 1.809 1.966 1.809 1.966 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.900 1.916 make_images_data 4110 15.4 0.043 0.047 1.734 1.855 mp_sum_d 3889 11.9 1.250 1.786 1.250 1.786 hybrid_alltoall_any 4261 16.3 0.085 0.484 1.499 1.745 cp_fm_cholesky_invert 11 10.9 1.717 1.721 1.717 1.721 transfer_pw2rs 451 13.1 0.006 0.007 1.695 1.707 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.640 1.670 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=79.192000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=478.636364, yerr=3.141300 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 593.608704E+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 2553271. 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 0.032 0.055 68.633 68.640 qs_mol_dyn_low 1 2.0 0.003 0.003 68.378 68.390 qs_forces 11 3.9 0.003 0.003 68.135 68.136 qs_energies 11 4.9 0.002 0.002 64.816 64.820 scf_env_do_scf 11 5.9 0.000 0.001 56.412 56.416 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 48.897 48.898 dbcsr_multiply_generic 2055 12.4 0.115 0.118 38.051 38.217 velocity_verlet 10 3.0 0.027 0.045 35.972 35.977 qs_scf_new_mos 99 7.5 0.001 0.001 33.154 33.297 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.154 33.297 multiply_cannon 2055 13.4 0.224 0.243 31.472 32.845 ot_scf_mini 99 9.5 0.003 0.003 31.490 31.613 multiply_cannon_loop 2055 14.4 1.159 1.185 30.209 31.230 ot_mini 99 10.5 0.001 0.001 18.513 18.654 multiply_cannon_multrec 24660 15.4 6.955 8.777 14.026 15.791 rebuild_ks_matrix 110 8.3 0.000 0.000 13.548 13.644 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.548 13.643 qs_ot_get_derivative 99 11.5 0.001 0.001 12.739 12.868 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.923 12.009 mp_waitall_1 176588 16.5 7.653 9.962 7.653 9.962 multiply_cannon_metrocomm3 24660 15.4 0.072 0.074 5.301 8.329 multiply_cannon_sync_h2d 24660 15.4 6.394 7.647 6.394 7.647 init_scf_loop 11 6.9 0.000 0.000 7.480 7.481 dbcsr_mm_accdrv_process 52282 16.1 5.561 6.568 6.900 7.372 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.435 7.349 apply_single 110 13.6 0.000 0.001 6.434 7.348 qs_ot_get_p 110 10.4 0.001 0.001 6.319 6.477 sum_up_and_integrate 110 10.3 0.001 0.002 6.315 6.333 integrate_v_rspace 110 11.3 0.002 0.003 6.288 6.304 init_scf_run 11 5.9 0.000 0.001 6.046 6.047 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.046 6.047 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.817 5.829 calculate_rho_elec 110 8.6 0.039 0.047 5.817 5.828 ot_diis_step 99 11.5 0.010 0.011 5.725 5.725 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.927 5.619 prepare_preconditioner 11 7.9 0.000 0.000 5.413 5.436 make_preconditioner 11 8.9 0.000 0.000 5.413 5.436 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.991 5.154 make_m2s 4110 13.4 0.057 0.060 4.196 4.677 make_images 4110 14.4 0.409 0.466 4.087 4.564 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.427 4.450 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.967 3.967 pw_transfer 1331 11.6 0.066 0.072 3.493 3.632 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.386 3.527 wfi_extrapolate 11 7.9 0.001 0.001 3.503 3.503 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.412 3.426 cp_fm_diag_elpa_base 48 14.0 3.366 3.382 3.410 3.423 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.264 3.333 grid_integrate_task_list 110 12.3 3.169 3.327 3.169 3.327 density_rs2pw 110 9.6 0.004 0.005 3.061 3.254 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.230 3.232 fft_wrap_pw1pw2_140 451 13.1 0.459 0.475 2.918 3.060 calculate_dm_sparse 110 9.5 0.001 0.001 2.962 2.990 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.891 2.936 hybrid_alltoall_any 4261 16.3 0.106 0.453 2.015 2.819 make_images_data 4110 15.4 0.048 0.052 2.303 2.785 fft3d_ps 1111 14.6 1.111 1.336 2.484 2.633 cp_fm_cholesky_invert 11 10.9 2.501 2.508 2.501 2.508 calculate_first_density_matrix 1 7.0 0.000 0.001 2.445 2.448 mp_sum_l 6594 12.7 1.755 2.405 1.755 2.405 grid_collocate_task_list 110 9.6 2.171 2.343 2.171 2.343 potential_pw2rs 110 12.3 0.009 0.009 2.198 2.207 acc_transpose_blocks 24660 15.4 0.117 0.120 1.970 2.001 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.950 1.973 mp_alltoall_d11v 2046 13.8 1.734 1.969 1.734 1.969 jit_kernel_multiply 10 16.5 0.973 1.888 0.973 1.888 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.851 1.852 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.780 1.793 mp_allgather_i34 2055 14.4 0.571 1.620 0.571 1.620 transfer_rs2pw 451 10.6 0.006 0.007 1.396 1.612 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.532 1.543 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.388 1.491 multiply_cannon_metrocomm4 22605 15.4 0.080 0.086 0.775 1.486 dbcsr_complete_redistribute 325 12.2 0.245 0.323 1.112 1.376 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.640000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=561.363636, yerr=7.113019 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_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 665.710592E+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 0.007 0.025 60.094 60.094 qs_mol_dyn_low 1 2.0 0.003 0.003 59.825 59.835 qs_forces 11 3.9 0.003 0.003 58.789 58.789 qs_energies 11 4.9 0.001 0.002 55.624 55.627 scf_env_do_scf 11 5.9 0.000 0.001 47.858 47.858 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.489 39.489 velocity_verlet 10 3.0 0.002 0.003 33.153 33.185 dbcsr_multiply_generic 2055 12.4 0.106 0.110 28.566 28.820 qs_scf_new_mos 99 7.5 0.001 0.001 24.975 25.072 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.975 25.072 ot_scf_mini 99 9.5 0.002 0.002 23.729 23.838 multiply_cannon 2055 13.4 0.212 0.220 22.389 23.604 multiply_cannon_loop 2055 14.4 0.817 0.853 21.218 22.165 ot_mini 99 10.5 0.001 0.001 13.639 13.745 rebuild_ks_matrix 110 8.3 0.000 0.000 12.032 12.144 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.032 12.143 multiply_cannon_multrec 16440 15.4 3.731 5.042 9.925 11.110 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.585 10.686 mp_waitall_1 139946 16.5 6.718 9.849 6.718 9.849 qs_ot_get_derivative 99 11.5 0.001 0.001 9.223 9.330 init_scf_loop 11 6.9 0.000 0.000 8.335 8.337 multiply_cannon_metrocomm3 16440 15.4 0.046 0.047 4.199 7.217 prepare_preconditioner 11 7.9 0.000 0.000 6.610 6.633 make_preconditioner 11 8.9 0.000 0.000 6.610 6.633 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.957 6.294 sum_up_and_integrate 110 10.3 0.001 0.002 6.215 6.229 integrate_v_rspace 110 11.3 0.003 0.003 6.189 6.203 dbcsr_mm_accdrv_process 34862 16.1 5.319 5.795 6.038 6.182 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.698 5.707 calculate_rho_elec 110 8.6 0.058 0.059 5.697 5.707 qs_ot_get_p 110 10.4 0.001 0.001 5.264 5.391 init_scf_run 11 5.9 0.000 0.001 5.390 5.390 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.389 5.389 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.894 5.358 apply_single 110 13.6 0.000 0.000 4.894 5.358 make_m2s 4110 13.4 0.049 0.051 4.163 4.546 make_images 4110 14.4 0.400 0.523 4.049 4.430 ot_diis_step 99 11.5 0.011 0.011 4.387 4.387 multiply_cannon_sync_h2d 16440 15.4 3.234 3.909 3.234 3.909 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.157 3.784 qs_ot_p2m_diag 48 11.0 0.041 0.044 3.657 3.661 pw_transfer 1331 11.6 0.065 0.074 3.533 3.542 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.426 3.438 grid_integrate_task_list 110 12.3 3.188 3.414 3.188 3.414 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.292 3.292 density_rs2pw 110 9.6 0.004 0.005 2.941 3.076 fft_wrap_pw1pw2_140 451 13.1 0.578 0.583 2.970 2.982 wfi_extrapolate 11 7.9 0.001 0.001 2.947 2.947 make_images_data 4110 15.4 0.045 0.050 2.341 2.806 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.770 2.779 cp_fm_diag_elpa_base 48 14.0 2.705 2.736 2.768 2.777 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.741 2.743 hybrid_alltoall_any 4261 16.3 0.109 0.382 2.046 2.710 cp_fm_cholesky_invert 11 10.9 2.563 2.569 2.563 2.569 calculate_dm_sparse 110 9.5 0.001 0.001 2.529 2.559 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.418 2.480 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.421 2.463 grid_collocate_task_list 110 9.6 2.213 2.401 2.213 2.401 fft3d_ps 1111 14.6 1.090 1.101 2.370 2.378 calculate_first_density_matrix 1 7.0 0.000 0.000 2.355 2.356 multiply_cannon_metrocomm4 14385 15.4 0.049 0.052 0.867 2.354 mp_irecv_dv 48980 15.7 0.791 2.219 0.791 2.219 potential_pw2rs 110 12.3 0.011 0.012 2.062 2.067 mp_sum_l 6594 12.7 1.392 1.972 1.392 1.972 mp_alltoall_d11v 2046 13.8 1.743 1.962 1.743 1.962 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.946 1.947 dbcsr_complete_redistribute 325 12.2 0.330 0.375 1.444 1.907 cp_fm_upper_to_full 70 14.2 1.370 1.732 1.370 1.732 acc_transpose_blocks 16440 15.4 0.078 0.080 1.583 1.654 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.638 1.650 cp_fm_cholesky_decompose 22 10.9 1.548 1.569 1.548 1.569 mp_allgather_i34 2055 14.4 0.484 1.492 0.484 1.492 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.467 1.479 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.369 1.473 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.949 1.394 transfer_rs2pw 451 10.6 0.005 0.006 1.260 1.388 mp_waitany 17072 13.8 1.108 1.285 1.108 1.285 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.263 1.274 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.094000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.363636, yerr=7.486627 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_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 733.540352E+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 0.008 0.028 65.861 65.862 qs_mol_dyn_low 1 2.0 0.003 0.003 65.645 65.654 qs_forces 11 3.9 0.003 0.003 65.299 65.300 qs_energies 11 4.9 0.001 0.002 61.917 61.922 scf_env_do_scf 11 5.9 0.000 0.001 53.680 53.683 scf_env_do_scf_inner_loop 99 6.5 0.010 0.057 42.064 42.064 velocity_verlet 10 3.0 0.002 0.002 37.486 37.499 dbcsr_multiply_generic 2055 12.4 0.113 0.119 30.768 30.982 qs_scf_new_mos 99 7.5 0.001 0.001 27.489 27.594 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.488 27.594 ot_scf_mini 99 9.5 0.002 0.003 25.813 25.912 multiply_cannon 2055 13.4 0.237 0.252 23.568 24.756 multiply_cannon_loop 2055 14.4 1.432 1.489 22.122 22.753 ot_mini 99 10.5 0.001 0.001 14.853 14.973 multiply_cannon_multrec 24660 15.4 4.088 6.685 12.997 14.037 rebuild_ks_matrix 110 8.3 0.000 0.000 11.938 12.043 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.937 12.043 init_scf_loop 11 6.9 0.000 0.000 11.577 11.578 qs_ot_get_derivative 99 11.5 0.001 0.001 10.669 10.773 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.543 10.636 prepare_preconditioner 11 7.9 0.000 0.000 9.835 9.851 make_preconditioner 11 8.9 0.000 0.000 9.835 9.851 dbcsr_mm_accdrv_process 52304 16.0 7.708 9.047 8.751 9.692 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.055 9.523 sum_up_and_integrate 110 10.3 0.001 0.002 6.181 6.195 mp_waitall_1 121746 16.5 4.232 6.181 4.232 6.181 integrate_v_rspace 110 11.3 0.003 0.003 6.155 6.168 make_m2s 4110 13.4 0.059 0.062 5.532 5.926 qs_ot_get_p 110 10.4 0.001 0.001 5.786 5.920 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.819 5.826 calculate_rho_elec 110 8.6 0.078 0.081 5.818 5.826 make_images 4110 14.4 0.578 0.699 5.391 5.780 init_scf_run 11 5.9 0.000 0.001 5.515 5.515 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.514 5.515 cp_fm_upper_to_full 70 14.2 3.386 4.901 3.386 4.901 ot_diis_step 99 11.5 0.011 0.012 4.146 4.146 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.048 4.133 apply_single 110 13.6 0.000 0.000 4.048 4.132 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.930 3.946 dbcsr_complete_redistribute 325 12.2 0.426 0.475 2.670 3.828 pw_transfer 1331 11.6 0.065 0.073 3.619 3.653 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.545 3.598 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.513 3.550 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.397 3.448 grid_integrate_task_list 110 12.3 3.281 3.434 3.281 3.434 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.362 3.362 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.138 3.281 hybrid_alltoall_any 4261 16.3 0.122 0.460 2.324 3.169 make_images_data 4110 15.4 0.049 0.053 2.787 3.155 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.380 3.097 density_rs2pw 110 9.6 0.004 0.004 2.920 3.088 fft_wrap_pw1pw2_140 451 13.1 0.610 0.630 3.031 3.070 calculate_dm_sparse 110 9.5 0.001 0.001 3.002 3.040 wfi_extrapolate 11 7.9 0.001 0.001 3.001 3.001 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.771 2.892 mp_alltoall_i22 605 13.7 1.657 2.879 1.657 2.879 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.830 2.838 cp_fm_diag_elpa_base 48 14.0 2.679 2.739 2.827 2.835 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.804 2.805 cp_fm_cholesky_invert 11 10.9 2.612 2.620 2.612 2.620 acc_transpose_blocks 24660 15.4 0.115 0.119 2.452 2.567 multiply_cannon_sync_h2d 24660 15.4 2.358 2.486 2.358 2.486 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.431 2.467 grid_collocate_task_list 110 9.6 2.263 2.439 2.263 2.439 calculate_first_density_matrix 1 7.0 0.000 0.000 2.416 2.420 fft3d_ps 1111 14.6 1.083 1.114 2.395 2.409 qs_energies_init_hamiltonians 11 5.9 0.002 0.006 2.230 2.231 mp_alltoall_d11v 2046 13.8 1.783 2.025 1.783 2.025 potential_pw2rs 110 12.3 0.013 0.014 1.997 2.003 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.795 1.835 mp_allgather_i34 2055 14.4 0.523 1.717 0.523 1.717 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.615 1.706 cp_fm_cholesky_decompose 22 10.9 1.645 1.689 1.645 1.689 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.635 1.646 mp_sum_l 6594 12.7 0.979 1.605 0.979 1.605 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.569 1.582 acc_transpose_blocks_sync 73980 16.4 1.440 1.549 1.440 1.549 multiply_cannon_metrocomm4 20550 15.4 0.063 0.067 0.861 1.454 jit_kernel_multiply 8 15.9 0.700 1.433 0.700 1.433 mp_irecv_dv 62702 16.1 0.755 1.370 0.755 1.370 transfer_rs2pw 451 10.6 0.005 0.006 1.181 1.346 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.862000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=694.909091, yerr=8.649273 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_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 864.677888E+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 0.010 0.030 54.781 54.783 qs_mol_dyn_low 1 2.0 0.003 0.003 54.550 54.561 qs_forces 11 3.9 0.003 0.003 54.487 54.487 qs_energies 11 4.9 0.001 0.002 50.860 50.864 scf_env_do_scf 11 5.9 0.000 0.001 42.688 42.688 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.005 35.006 velocity_verlet 10 3.0 0.002 0.002 31.014 31.017 dbcsr_multiply_generic 2055 12.4 0.106 0.110 23.198 23.368 qs_scf_new_mos 99 7.5 0.001 0.001 20.614 20.655 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.613 20.655 ot_scf_mini 99 9.5 0.002 0.002 19.359 19.380 multiply_cannon 2055 13.4 0.238 0.257 17.611 19.085 multiply_cannon_loop 2055 14.4 0.603 0.627 16.335 16.713 rebuild_ks_matrix 110 8.3 0.000 0.000 11.535 11.557 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.535 11.557 ot_mini 99 10.5 0.001 0.001 10.733 10.746 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.256 10.273 multiply_cannon_multrec 8220 15.4 3.177 4.399 7.548 8.479 init_scf_loop 11 6.9 0.000 0.000 7.636 7.637 mp_waitall_1 103326 16.6 5.854 7.427 5.854 7.427 qs_ot_get_derivative 99 11.5 0.001 0.001 7.056 7.076 sum_up_and_integrate 110 10.3 0.001 0.002 6.155 6.167 integrate_v_rspace 110 11.3 0.003 0.003 6.128 6.141 prepare_preconditioner 11 7.9 0.000 0.000 5.996 6.001 make_preconditioner 11 8.9 0.000 0.000 5.996 6.001 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.887 5.901 calculate_rho_elec 110 8.6 0.114 0.114 5.886 5.900 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.572 5.651 dbcsr_mm_accdrv_process 17442 15.9 3.086 4.138 4.231 5.168 init_scf_run 11 5.9 0.000 0.001 5.055 5.055 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.054 5.055 qs_ot_get_p 110 10.4 0.001 0.001 4.662 4.685 make_m2s 4110 13.4 0.038 0.039 4.233 4.520 make_images 4110 14.4 0.650 0.711 4.104 4.390 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 3.026 4.270 pw_transfer 1331 11.6 0.065 0.069 3.802 3.812 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.720 3.766 apply_single 110 13.6 0.000 0.000 3.719 3.766 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.695 3.708 ot_diis_step 99 11.5 0.012 0.012 3.654 3.655 grid_integrate_task_list 110 12.3 3.388 3.491 3.388 3.491 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.292 3.297 fft_wrap_pw1pw2_140 451 13.1 0.778 0.789 3.199 3.221 density_rs2pw 110 9.6 0.004 0.004 2.880 2.950 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.948 2.949 cp_fm_cholesky_invert 11 10.9 2.872 2.876 2.872 2.876 hybrid_alltoall_any 4261 16.3 0.200 0.853 2.238 2.759 wfi_extrapolate 11 7.9 0.001 0.001 2.727 2.727 make_images_data 4110 15.4 0.040 0.046 2.273 2.664 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.658 2.659 calculate_dm_sparse 110 9.5 0.001 0.001 2.479 2.520 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.511 2.513 multiply_cannon_sync_h2d 8220 15.4 2.383 2.469 2.383 2.469 grid_collocate_task_list 110 9.6 2.369 2.465 2.369 2.465 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.444 2.452 cp_fm_diag_elpa_base 48 14.0 2.389 2.410 2.442 2.450 fft3d_ps 1111 14.6 1.135 1.179 2.360 2.377 calculate_first_density_matrix 1 7.0 0.000 0.000 2.232 2.234 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.132 2.146 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.995 2.021 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.783 1.990 potential_pw2rs 110 12.3 0.015 0.016 1.953 1.956 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.911 1.922 mp_alltoall_d11v 2046 13.8 1.643 1.800 1.643 1.800 cp_fm_cholesky_decompose 22 10.9 1.686 1.705 1.686 1.705 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.608 1.613 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.488 1.604 mp_allgather_i34 2055 14.4 0.446 1.549 0.446 1.549 dbcsr_complete_redistribute 325 12.2 0.560 0.577 1.434 1.528 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.430 1.442 acc_transpose_blocks 8220 15.4 0.040 0.041 1.359 1.392 qs_create_task_list 11 7.9 0.000 0.000 1.206 1.304 generate_qs_task_list 11 8.9 0.372 0.439 1.206 1.304 transfer_rs2pw 451 10.6 0.005 0.005 1.106 1.202 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.147 1.170 mp_waitany 9240 13.8 1.056 1.140 1.056 1.140 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.783000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=815.545455, yerr=14.468547 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_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 1.331442E+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 0.033 0.065 86.426 86.439 qs_mol_dyn_low 1 2.0 0.003 0.003 86.053 86.077 qs_forces 11 3.9 0.003 0.003 85.807 85.808 qs_energies 11 4.9 0.002 0.002 81.732 81.734 scf_env_do_scf 11 5.9 0.000 0.001 71.846 71.846 velocity_verlet 10 3.0 0.002 0.002 55.381 55.409 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.612 43.613 dbcsr_multiply_generic 2055 12.4 0.119 0.124 29.259 29.357 init_scf_loop 11 6.9 0.000 0.000 28.164 28.165 qs_scf_new_mos 99 7.5 0.001 0.001 26.557 26.590 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.557 26.590 prepare_preconditioner 11 7.9 0.000 0.000 26.195 26.202 make_preconditioner 11 8.9 0.000 0.000 26.195 26.202 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.287 25.677 ot_scf_mini 99 9.5 0.002 0.002 24.774 24.801 multiply_cannon 2055 13.4 0.324 0.345 22.284 23.043 multiply_cannon_loop 2055 14.4 0.821 0.834 20.516 20.785 cp_fm_upper_to_full 70 14.2 12.664 18.306 12.664 18.306 ot_mini 99 10.5 0.001 0.001 14.001 14.028 rebuild_ks_matrix 110 8.3 0.000 0.001 13.436 13.469 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.436 13.469 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.190 12.219 dbcsr_complete_redistribute 325 12.2 1.015 1.040 7.457 10.656 multiply_cannon_multrec 8220 15.4 4.109 4.265 9.756 9.873 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.455 9.652 qs_ot_get_derivative 99 11.5 0.001 0.001 9.440 9.468 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.893 9.068 mp_alltoall_i22 605 13.7 5.515 8.693 5.515 8.693 mp_waitall_1 84994 16.7 7.318 8.136 7.318 8.136 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.981 7.017 calculate_rho_elec 110 8.6 0.225 0.225 6.981 7.016 sum_up_and_integrate 110 10.3 0.002 0.002 6.654 6.668 integrate_v_rspace 110 11.3 0.004 0.004 6.626 6.640 make_m2s 4110 13.4 0.043 0.044 5.274 5.755 dbcsr_mm_accdrv_process 11614 15.7 3.787 4.112 5.500 5.715 init_scf_run 11 5.9 0.000 0.001 5.631 5.631 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.630 5.630 make_images 4110 14.4 0.890 0.948 5.085 5.566 qs_ot_get_p 110 10.4 0.001 0.001 5.505 5.536 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.711 5.178 apply_single 110 13.6 0.000 0.000 4.710 5.178 cp_fm_cholesky_invert 11 10.9 5.149 5.153 5.149 5.153 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.734 5.047 pw_transfer 1331 11.6 0.074 0.075 4.837 4.843 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.721 4.726 ot_diis_step 99 11.5 0.015 0.015 4.544 4.544 fft_wrap_pw1pw2_140 451 13.1 1.278 1.283 4.136 4.141 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.968 3.975 grid_integrate_task_list 110 12.3 3.674 3.739 3.674 3.739 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.647 3.647 hybrid_alltoall_any 4261 16.3 0.262 0.563 2.795 3.512 density_rs2pw 110 9.6 0.004 0.004 3.463 3.485 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.451 3.452 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.971 3.434 make_images_data 4110 15.4 0.045 0.049 2.754 3.349 wfi_extrapolate 11 7.9 0.001 0.001 3.299 3.299 calculate_dm_sparse 110 9.5 0.001 0.001 3.141 3.171 multiply_cannon_sync_h2d 8220 15.4 3.132 3.154 3.132 3.154 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.870 2.871 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.870 2.870 cp_fm_diag_elpa_base 48 14.0 2.327 2.527 2.868 2.868 fft3d_ps 1111 14.6 1.296 1.301 2.776 2.782 grid_collocate_task_list 110 9.6 2.663 2.674 2.663 2.674 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.613 2.628 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.363 2.375 calculate_first_density_matrix 1 7.0 0.000 0.000 2.219 2.219 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.168 2.214 potential_pw2rs 110 12.3 0.022 0.022 2.200 2.202 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.084 2.171 mp_alltoall_d11v 2046 13.8 1.997 2.071 1.997 2.071 cp_fm_cholesky_decompose 22 10.9 1.957 1.973 1.957 1.973 qs_create_task_list 11 7.9 0.000 0.000 1.879 1.925 generate_qs_task_list 11 8.9 0.729 0.781 1.879 1.925 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.843 1.846 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.747 1.782 acc_transpose_blocks 8220 15.4 0.041 0.041 1.713 1.731 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.439000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1224.181818, yerr=36.180905 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 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 635.052032E+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 3175955098256 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 1665926. 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 0.035 0.063 199.553 199.555 qs_mol_dyn_low 1 2.0 0.005 0.016 198.959 199.024 qs_forces 11 3.9 0.007 0.028 198.768 198.770 qs_energies 11 4.9 0.007 0.025 193.165 193.179 scf_env_do_scf 11 5.9 0.001 0.001 176.570 176.574 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 156.251 156.253 dbcsr_multiply_generic 2507 12.6 0.180 0.184 120.661 121.169 velocity_verlet 10 3.0 0.005 0.007 120.772 120.774 qs_scf_new_mos 117 7.6 0.001 0.001 117.240 117.496 qs_scf_loop_do_ot 117 8.6 0.001 0.001 117.239 117.496 ot_scf_mini 117 9.6 0.003 0.003 110.783 111.045 multiply_cannon 2507 13.6 0.243 0.255 100.394 102.023 multiply_cannon_loop 2507 14.6 2.264 2.339 98.268 99.938 ot_mini 117 10.6 0.001 0.001 62.528 62.808 multiply_cannon_multrec 60168 15.6 32.800 35.421 42.219 44.242 qs_ot_get_derivative 117 11.6 0.001 0.001 37.569 37.844 rebuild_ks_matrix 128 8.3 0.001 0.001 33.661 33.924 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.661 33.923 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.212 30.452 multiply_cannon_sync_h2d 60168 15.6 27.884 30.111 27.884 30.111 mp_waitall_1 267128 16.5 25.857 28.862 25.857 28.862 qs_ot_get_p 128 10.4 0.001 0.001 26.592 26.803 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.586 25.262 apply_single 128 13.6 0.001 0.001 24.586 25.262 ot_diis_step 117 11.6 0.008 0.008 24.786 24.787 qs_ot_p2m_diag 83 11.4 0.079 0.091 20.231 20.307 init_scf_loop 11 6.9 0.000 0.000 20.245 20.247 cp_dbcsr_syevd 83 12.4 0.005 0.005 18.217 18.217 multiply_cannon_metrocomm3 60168 15.6 0.120 0.125 14.351 16.397 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 16.161 16.354 prepare_preconditioner 11 7.9 0.000 0.000 15.557 15.608 make_preconditioner 11 8.9 0.000 0.000 15.557 15.608 cp_fm_diag_elpa 83 13.4 0.000 0.000 15.304 15.330 cp_fm_diag_elpa_base 83 14.4 15.237 15.282 15.300 15.328 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.818 14.986 make_m2s 5014 13.6 0.105 0.113 13.768 14.140 make_images 5014 14.6 0.397 0.417 13.585 13.965 sum_up_and_integrate 128 10.3 0.002 0.004 13.899 13.947 integrate_v_rspace 128 11.3 0.004 0.008 13.840 13.891 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.858 12.945 calculate_rho_elec 128 8.7 0.045 0.064 12.858 12.944 init_scf_run 11 5.9 0.000 0.001 12.347 12.348 scf_env_initial_rho_setup 11 6.9 0.009 0.019 12.347 12.348 dbcsr_mm_accdrv_process 124484 16.2 4.254 4.516 8.930 9.647 wfi_extrapolate 11 7.9 0.001 0.001 9.055 9.055 cp_fm_cholesky_invert 11 10.9 8.866 8.873 8.866 8.873 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.245 8.419 calculate_dm_sparse 128 9.5 0.001 0.001 8.280 8.338 pw_transfer 1547 11.6 0.073 0.085 7.581 7.780 make_images_data 5014 15.6 0.066 0.073 6.732 7.679 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.380 7.581 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.435 7.519 grid_integrate_task_list 128 12.3 7.010 7.403 7.010 7.403 density_rs2pw 128 9.7 0.006 0.007 6.643 7.203 hybrid_alltoall_any 5200 16.5 0.295 2.256 5.920 7.189 multiply_cannon_metrocomm1 60168 15.6 0.097 0.102 5.141 7.113 fft_wrap_pw1pw2_140 523 13.2 1.127 1.169 6.516 6.723 mp_sum_l 7950 12.9 5.223 6.677 5.223 6.677 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.625 6.636 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.774 5.905 fft3d_ps 1291 14.7 2.196 2.844 5.385 5.734 grid_collocate_task_list 128 9.7 4.834 5.249 4.834 5.249 mp_alltoall_d11v 2415 14.1 4.145 5.214 4.145 5.214 cp_fm_cholesky_decompose 22 10.9 4.545 4.559 4.545 4.559 potential_pw2rs 128 12.3 0.009 0.010 4.430 4.479 mp_sum_d 4465 12.1 3.386 4.021 3.386 4.021 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=199.555000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=599.727273, yerr=7.435675 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 840.273920E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57903. MP_Allreduce 11138 958. MP_Sync 87 MP_Alltoall 1983 6306357. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.103 189.315 189.326 qs_mol_dyn_low 1 2.0 0.003 0.004 188.627 188.683 qs_forces 11 3.9 0.004 0.005 188.524 188.525 qs_energies 11 4.9 0.013 0.035 181.733 181.759 scf_env_do_scf 11 5.9 0.001 0.006 165.192 165.203 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 132.777 132.780 velocity_verlet 10 3.0 0.023 0.029 119.153 119.155 dbcsr_multiply_generic 2527 12.6 0.192 0.197 98.019 99.209 qs_scf_new_mos 118 7.6 0.001 0.001 94.080 94.573 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.079 94.572 ot_scf_mini 118 9.6 0.004 0.005 89.258 89.915 multiply_cannon 2527 13.6 0.506 0.562 77.941 82.147 multiply_cannon_loop 2527 14.6 1.576 1.649 74.572 77.118 ot_mini 118 10.6 0.001 0.001 50.405 50.987 mp_waitall_1 216438 16.6 23.847 39.604 23.847 39.604 multiply_cannon_multrec 30324 15.6 21.276 25.530 31.960 36.490 rebuild_ks_matrix 129 8.3 0.001 0.001 32.488 33.168 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 32.488 33.167 init_scf_loop 11 6.9 0.001 0.004 32.321 32.324 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.191 29.818 multiply_cannon_metrocomm3 30324 15.6 0.097 0.103 15.590 29.305 qs_ot_get_derivative 118 11.6 0.001 0.001 28.329 28.987 prepare_preconditioner 11 7.9 0.000 0.001 27.986 28.056 make_preconditioner 11 8.9 0.000 0.003 27.986 28.056 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.688 27.209 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.212 23.351 apply_single 129 13.6 0.001 0.001 22.212 23.351 ot_diis_step 118 11.6 0.015 0.015 21.905 21.907 qs_ot_get_p 129 10.4 0.001 0.002 20.724 21.415 multiply_cannon_sync_h2d 30324 15.6 18.328 21.177 18.328 21.177 cp_fm_cholesky_invert 11 10.9 16.332 16.344 16.332 16.344 make_m2s 5054 13.6 0.090 0.098 14.130 15.814 qs_ot_p2m_diag 83 11.4 0.188 0.217 15.758 15.792 make_images 5054 14.6 1.184 1.392 13.915 15.597 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.563 14.564 sum_up_and_integrate 129 10.3 0.002 0.004 14.076 14.112 integrate_v_rspace 129 11.3 0.003 0.004 14.016 14.054 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.362 13.392 calculate_rho_elec 129 8.7 0.088 0.104 13.361 13.391 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.193 11.671 init_scf_run 11 5.9 0.000 0.001 11.586 11.588 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.586 11.587 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.432 11.465 cp_fm_diag_elpa_base 83 14.4 11.181 11.270 11.426 11.454 multiply_cannon_metrocomm4 27797 15.6 0.108 0.122 3.862 10.945 dbcsr_mm_accdrv_process 62734 16.2 5.502 6.136 10.120 10.740 mp_irecv_dv 70031 16.3 3.653 10.534 3.653 10.534 make_images_data 5054 15.6 0.067 0.077 8.271 10.397 hybrid_alltoall_any 5240 16.5 0.352 1.516 6.987 9.893 pw_transfer 1559 11.6 0.085 0.094 8.629 8.715 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.405 8.495 wfi_extrapolate 11 7.9 0.001 0.001 8.308 8.308 density_rs2pw 129 9.7 0.006 0.007 7.034 7.615 fft_wrap_pw1pw2_140 527 13.2 1.224 1.246 7.474 7.579 grid_integrate_task_list 129 12.3 7.228 7.516 7.228 7.516 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.363 7.201 cp_fm_cholesky_decompose 22 10.9 6.781 6.860 6.781 6.860 calculate_dm_sparse 129 9.5 0.001 0.001 6.476 6.632 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.245 6.254 fft3d_ps 1301 14.7 2.864 3.061 5.943 6.028 mp_sum_l 8010 12.9 4.145 6.025 4.145 6.025 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.333 5.526 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.397 5.464 grid_collocate_task_list 129 9.7 5.038 5.427 5.038 5.427 mp_alltoall_d11v 2423 14.1 4.111 5.092 4.111 5.092 mp_allgather_i34 2527 14.6 1.875 4.885 1.875 4.885 potential_pw2rs 129 12.3 0.016 0.019 4.497 4.517 dbcsr_complete_redistribute 395 12.7 0.783 0.863 3.128 3.944 mp_sum_d 4499 12.1 2.612 3.943 2.612 3.943 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.326000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=801.000000, yerr=2.256304 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 937.869312E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172905552 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58278. MP_Allreduce 11070 1000. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.033 172.694 172.695 qs_mol_dyn_low 1 2.0 0.003 0.003 172.205 172.218 qs_forces 11 3.9 0.004 0.005 172.091 172.093 qs_energies 11 4.9 0.012 0.019 165.588 165.596 scf_env_do_scf 11 5.9 0.001 0.001 149.618 149.620 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 115.224 115.224 velocity_verlet 10 3.0 0.006 0.033 109.920 109.922 dbcsr_multiply_generic 2485 12.5 0.179 0.183 81.195 82.206 qs_scf_new_mos 116 7.6 0.001 0.001 78.832 79.234 qs_scf_loop_do_ot 116 8.6 0.001 0.001 78.831 79.233 ot_scf_mini 116 9.6 0.003 0.004 74.654 75.085 multiply_cannon 2485 13.5 0.498 0.512 61.575 66.106 multiply_cannon_loop 2485 14.5 1.120 1.182 58.986 61.403 ot_mini 116 10.6 0.001 0.001 42.136 42.556 mp_waitall_1 169034 16.6 24.664 34.376 24.664 34.376 init_scf_loop 11 6.9 0.000 0.000 34.297 34.298 prepare_preconditioner 11 7.9 0.000 0.000 30.297 30.344 make_preconditioner 11 8.9 0.000 0.000 30.297 30.344 rebuild_ks_matrix 127 8.3 0.001 0.001 29.749 30.269 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.017 29.748 30.269 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.998 29.391 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.799 27.282 multiply_cannon_multrec 19880 15.5 12.950 16.019 22.486 25.470 multiply_cannon_metrocomm3 19880 15.5 0.061 0.064 15.295 24.716 qs_ot_get_derivative 116 11.6 0.001 0.002 22.648 23.073 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.652 20.604 apply_single 127 13.6 0.001 0.001 19.651 20.604 ot_diis_step 116 11.6 0.018 0.019 19.383 19.383 qs_ot_get_p 127 10.4 0.001 0.001 18.061 18.489 make_m2s 4970 13.5 0.079 0.083 14.727 15.432 multiply_cannon_sync_h2d 19880 15.5 13.440 15.209 13.440 15.209 make_images 4970 14.5 1.144 1.241 14.495 15.195 cp_fm_cholesky_invert 11 10.9 14.136 14.145 14.136 14.145 qs_ot_p2m_diag 82 11.4 0.261 0.269 13.916 13.926 sum_up_and_integrate 127 10.3 0.002 0.003 13.739 13.767 integrate_v_rspace 127 11.3 0.003 0.004 13.680 13.704 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.287 13.319 calculate_rho_elec 127 8.7 0.129 0.144 13.286 13.319 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.885 12.886 init_scf_run 11 5.9 0.000 0.001 10.936 10.937 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.936 10.936 make_images_data 4970 15.5 0.061 0.070 8.945 10.049 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.775 9.793 cp_fm_diag_elpa_base 82 14.4 9.376 9.529 9.772 9.790 hybrid_alltoall_any 5155 16.4 0.447 2.038 7.688 9.412 multiply_cannon_metrocomm4 17395 15.5 0.067 0.079 3.495 9.315 dbcsr_mm_accdrv_process 41158 16.2 5.824 6.220 8.997 9.187 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.811 9.116 mp_irecv_dv 49801 16.2 3.365 9.047 3.365 9.047 pw_transfer 1535 11.6 0.084 0.101 8.526 8.633 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 8.305 8.418 grid_integrate_task_list 127 12.3 7.260 7.597 7.260 7.597 fft_wrap_pw1pw2_140 519 13.2 1.296 1.330 7.396 7.523 wfi_extrapolate 11 7.9 0.001 0.001 7.353 7.353 density_rs2pw 127 9.7 0.006 0.006 6.738 7.248 cp_fm_cholesky_decompose 22 10.9 7.120 7.149 7.120 7.149 cp_fm_upper_to_full 104 14.8 5.660 7.106 5.660 7.106 dbcsr_complete_redistribute 393 12.7 1.176 1.224 4.496 6.220 calculate_dm_sparse 127 9.5 0.001 0.001 5.861 5.939 fft3d_ps 1281 14.7 2.740 2.964 5.698 5.771 grid_collocate_task_list 127 9.7 5.116 5.612 5.116 5.612 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.531 5.536 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.582 5.235 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.346 5.068 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.641 4.785 mp_alltoall_d11v 2401 14.1 4.159 4.571 4.159 4.571 mp_sum_l 7884 12.9 3.075 4.382 3.075 4.382 mp_allgather_i34 2485 14.5 1.097 4.319 1.097 4.319 potential_pw2rs 127 12.3 0.021 0.023 4.270 4.281 transfer_fm_to_dbcsr 11 9.9 0.019 0.026 2.280 3.994 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.911 3.936 qs_energies_init_hamiltonians 11 5.9 0.021 0.052 3.758 3.759 mp_alltoall_i22 712 14.1 1.854 3.716 1.854 3.716 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.442 3.492 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=172.695000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=886.636364, yerr=6.328474 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 4.320337E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.149927E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+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 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837542000 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58189. MP_Allreduce 11085 1083. MP_Sync 86 MP_Alltoall 1700 12496381. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.029 186.237 186.237 qs_mol_dyn_low 1 2.0 0.003 0.003 185.791 185.804 qs_forces 11 3.9 0.004 0.005 185.666 185.673 qs_energies 11 4.9 0.011 0.011 178.580 178.591 scf_env_do_scf 11 5.9 0.001 0.001 161.982 161.993 velocity_verlet 10 3.0 0.002 0.002 123.196 123.198 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.801 115.802 dbcsr_multiply_generic 2485 12.5 0.187 0.195 80.048 80.897 qs_scf_new_mos 116 7.6 0.001 0.001 79.922 80.210 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.922 80.209 ot_scf_mini 116 9.6 0.003 0.004 75.413 75.719 multiply_cannon 2485 13.5 0.547 0.576 55.330 57.905 multiply_cannon_loop 2485 14.5 1.841 1.921 51.724 53.684 init_scf_loop 11 6.9 0.000 0.000 46.058 46.059 ot_mini 116 10.6 0.001 0.001 42.695 43.003 prepare_preconditioner 11 7.9 0.000 0.000 41.981 42.004 make_preconditioner 11 8.9 0.000 0.000 41.981 42.004 make_full_inverse_cholesky 11 9.9 0.012 0.023 35.557 40.625 multiply_cannon_multrec 29820 15.5 13.530 18.674 26.297 31.103 rebuild_ks_matrix 127 8.3 0.001 0.001 29.040 29.309 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 29.039 29.309 mp_waitall_1 146592 16.7 17.137 26.503 17.137 26.503 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.180 26.410 qs_ot_get_derivative 116 11.6 0.001 0.002 23.280 23.584 make_m2s 4970 13.5 0.096 0.099 20.312 21.297 make_images 4970 14.5 1.938 2.216 20.010 20.998 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.835 19.417 apply_single 127 13.6 0.001 0.001 18.835 19.417 ot_diis_step 116 11.6 0.017 0.018 19.295 19.296 qs_ot_get_p 127 10.4 0.001 0.001 18.945 19.280 cp_fm_upper_to_full 104 14.8 11.602 17.081 11.602 17.081 cp_fm_cholesky_invert 11 10.9 15.867 15.876 15.867 15.876 qs_ot_p2m_diag 82 11.4 0.338 0.385 14.754 14.807 multiply_cannon_metrocomm3 29820 15.5 0.049 0.052 6.257 14.423 sum_up_and_integrate 127 10.3 0.002 0.003 13.848 13.879 integrate_v_rspace 127 11.3 0.003 0.004 13.789 13.823 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.594 13.649 calculate_rho_elec 127 8.7 0.171 0.188 13.593 13.648 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.362 13.364 dbcsr_mm_accdrv_process 61748 16.2 8.248 9.062 12.343 12.864 dbcsr_complete_redistribute 393 12.7 1.506 1.617 9.002 12.791 make_images_data 4970 15.5 0.065 0.071 10.943 12.516 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.641 11.432 hybrid_alltoall_any 5155 16.4 0.525 2.182 9.697 11.421 multiply_cannon_sync_h2d 29820 15.5 10.353 10.989 10.353 10.989 init_scf_run 11 5.9 0.000 0.001 10.700 10.702 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.700 10.701 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.325 10.338 cp_fm_diag_elpa_base 82 14.4 9.349 9.670 10.319 10.330 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.401 10.073 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.754 9.974 mp_alltoall_i22 712 14.1 5.647 9.509 5.647 9.509 pw_transfer 1535 11.6 0.084 0.099 8.817 8.894 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.594 8.677 grid_integrate_task_list 127 12.3 7.502 7.838 7.502 7.838 fft_wrap_pw1pw2_140 519 13.2 1.433 1.470 7.659 7.752 wfi_extrapolate 11 7.9 0.001 0.001 7.637 7.637 cp_fm_cholesky_decompose 22 10.9 7.414 7.502 7.414 7.502 density_rs2pw 127 9.7 0.006 0.006 6.679 7.084 multiply_cannon_metrocomm4 24850 15.5 0.085 0.094 2.786 6.682 mp_irecv_dv 75445 16.2 2.630 6.411 2.630 6.411 calculate_dm_sparse 127 9.5 0.001 0.001 6.182 6.265 fft3d_ps 1281 14.7 2.827 2.907 5.799 5.862 grid_collocate_task_list 127 9.7 5.243 5.531 5.243 5.531 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.454 5.493 mp_alltoall_d11v 2401 14.1 4.716 5.424 4.716 5.424 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.489 4.593 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.384 4.506 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 4.396 4.397 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.194 4.260 potential_pw2rs 127 12.3 0.023 0.023 4.194 4.210 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.237000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1087.545455, yerr=17.732167 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 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 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.556062E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108580288 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58130. MP_Allreduce 11095 1166. MP_Sync 86 MP_Alltoall 1700 18828153. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.079 166.969 166.978 qs_mol_dyn_low 1 2.0 0.003 0.003 166.392 166.448 qs_forces 11 3.9 0.004 0.004 166.263 166.265 qs_energies 11 4.9 0.002 0.002 158.879 158.888 scf_env_do_scf 11 5.9 0.001 0.001 141.590 141.602 velocity_verlet 10 3.0 0.028 0.032 109.344 109.348 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 106.208 106.209 dbcsr_multiply_generic 2485 12.5 0.180 0.192 70.876 71.288 qs_scf_new_mos 116 7.6 0.001 0.001 70.069 70.174 qs_scf_loop_do_ot 116 8.6 0.001 0.001 70.069 70.174 ot_scf_mini 116 9.6 0.003 0.003 65.712 65.790 multiply_cannon 2485 13.5 0.551 0.591 52.607 55.883 multiply_cannon_loop 2485 14.5 0.793 0.825 49.732 50.510 ot_mini 116 10.6 0.001 0.001 36.802 36.878 init_scf_loop 11 6.9 0.000 0.000 35.233 35.236 prepare_preconditioner 11 7.9 0.000 0.000 31.421 31.446 make_preconditioner 11 8.9 0.000 0.000 31.421 31.446 mp_waitall_1 124680 16.7 24.223 30.427 24.223 30.427 make_full_inverse_cholesky 11 9.9 0.016 0.027 29.408 29.674 rebuild_ks_matrix 127 8.3 0.001 0.001 28.600 28.691 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 28.599 28.690 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.047 26.135 multiply_cannon_multrec 9940 15.5 10.316 14.250 17.824 20.877 qs_ot_get_derivative 116 11.6 0.001 0.002 20.059 20.136 multiply_cannon_metrocomm3 9940 15.5 0.025 0.027 12.382 19.318 cp_fm_cholesky_invert 11 10.9 17.965 17.970 17.965 17.970 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 16.792 17.010 apply_single 127 13.6 0.001 0.001 16.791 17.010 ot_diis_step 116 11.6 0.020 0.020 16.675 16.675 qs_ot_get_p 127 10.4 0.001 0.001 16.367 16.446 make_m2s 4970 13.5 0.066 0.070 14.544 15.334 make_images 4970 14.5 2.151 2.581 14.239 15.021 sum_up_and_integrate 127 10.3 0.002 0.002 13.886 13.933 integrate_v_rspace 127 11.3 0.004 0.004 13.826 13.874 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.799 13.840 calculate_rho_elec 127 8.7 0.254 0.264 13.799 13.839 qs_ot_p2m_diag 82 11.4 0.489 0.494 12.711 12.729 cp_dbcsr_syevd 82 12.4 0.005 0.006 11.569 11.570 multiply_cannon_sync_h2d 9940 15.5 10.827 11.195 10.827 11.195 init_scf_run 11 5.9 0.000 0.001 10.454 10.454 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.453 10.454 make_images_data 4970 15.5 0.055 0.066 8.372 9.623 hybrid_alltoall_any 5155 16.4 0.836 3.749 8.253 9.605 pw_transfer 1535 11.6 0.084 0.092 9.085 9.110 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.865 8.896 cp_fm_diag_elpa 82 13.4 0.000 0.000 8.768 8.781 cp_fm_diag_elpa_base 82 14.4 8.532 8.609 8.763 8.776 grid_integrate_task_list 127 12.3 7.702 8.103 7.702 8.103 cp_fm_cholesky_decompose 22 10.9 7.893 8.025 7.893 8.025 fft_wrap_pw1pw2_140 519 13.2 1.788 1.832 7.876 7.915 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.825 7.876 dbcsr_mm_accdrv_process 20590 16.1 3.269 4.299 7.158 7.872 wfi_extrapolate 11 7.9 0.001 0.001 7.290 7.290 multiply_cannon_metrocomm1 9940 15.5 0.029 0.030 4.238 6.880 density_rs2pw 127 9.7 0.006 0.006 6.566 6.826 calculate_dm_sparse 127 9.5 0.001 0.001 6.010 6.073 grid_collocate_task_list 127 9.7 5.523 5.718 5.523 5.718 fft3d_ps 1281 14.7 2.747 2.817 5.624 5.684 dbcsr_complete_redistribute 393 12.7 2.123 2.224 5.145 5.548 qs_energies_init_hamiltonians 11 5.9 0.003 0.005 5.250 5.252 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.214 5.226 mp_alltoall_d11v 2401 14.1 4.461 4.764 4.461 4.764 mp_allgather_i34 2485 14.5 1.076 4.490 1.076 4.490 potential_pw2rs 127 12.3 0.026 0.028 4.138 4.144 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.064 4.115 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.576 3.884 multiply_cannon_metrocomm4 7455 15.5 0.027 0.029 1.736 3.879 mp_irecv_dv 28618 15.9 1.695 3.799 1.695 3.799 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.342 3.652 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.547 3.632 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.525 3.564 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.392 3.536 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.449 3.457 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=166.978000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1480.454545, yerr=9.099087 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 11.786061E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 3.213066E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+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 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58558. MP_Allreduce 11184 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.074 289.866 289.875 qs_mol_dyn_low 1 2.0 0.003 0.003 289.135 289.154 qs_forces 11 3.9 0.004 0.005 289.044 289.047 qs_energies 11 4.9 0.002 0.002 280.173 280.180 scf_env_do_scf 11 5.9 0.001 0.001 257.474 257.482 velocity_verlet 10 3.0 0.002 0.002 209.950 209.959 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 133.956 133.958 init_scf_loop 11 6.9 0.000 0.000 123.240 123.243 prepare_preconditioner 11 7.9 0.000 0.000 118.414 118.435 make_preconditioner 11 8.9 0.000 0.000 118.414 118.435 make_full_inverse_cholesky 11 9.9 0.038 0.039 94.905 115.505 qs_scf_new_mos 118 7.6 0.001 0.001 90.427 90.502 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.426 90.501 ot_scf_mini 118 9.6 0.004 0.004 85.536 85.575 dbcsr_multiply_generic 2529 12.6 0.214 0.227 82.705 83.221 cp_fm_upper_to_full 106 14.8 52.625 75.077 52.625 75.077 multiply_cannon 2529 13.6 0.706 0.740 59.142 60.545 multiply_cannon_loop 2529 14.6 1.039 1.076 54.983 56.424 ot_mini 118 10.6 0.001 0.001 44.889 44.942 dbcsr_complete_redistribute 397 12.7 4.023 4.070 29.553 42.347 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.021 38.766 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 23.465 36.091 rebuild_ks_matrix 129 8.3 0.001 0.001 34.054 34.108 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 34.053 34.107 mp_alltoall_i22 720 14.1 21.218 33.817 21.218 33.817 cp_fm_cholesky_invert 11 10.9 33.299 33.305 33.299 33.305 mp_waitall_1 104580 16.8 28.181 32.454 28.181 32.454 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.629 31.679 qs_ot_get_p 129 10.4 0.001 0.001 25.684 25.771 qs_ot_get_derivative 118 11.6 0.002 0.002 24.708 24.741 qs_ot_p2m_diag 84 11.4 0.890 0.895 21.622 21.652 make_m2s 5058 13.6 0.077 0.080 18.949 20.480 multiply_cannon_metrocomm3 10116 15.6 0.025 0.027 18.762 20.176 ot_diis_step 118 11.6 0.022 0.022 20.150 20.150 make_images 5058 14.6 3.100 3.293 18.464 19.995 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.813 19.816 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.463 19.586 apply_single 129 13.6 0.001 0.001 19.463 19.585 multiply_cannon_multrec 10116 15.6 10.344 12.176 18.182 18.250 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.608 16.609 cp_fm_diag_elpa_base 84 14.4 12.134 13.730 16.605 16.605 qs_rho_update_rho_low 129 7.7 0.001 0.001 16.532 16.546 calculate_rho_elec 129 8.7 0.485 0.486 16.531 16.545 sum_up_and_integrate 129 10.3 0.002 0.002 16.066 16.159 integrate_v_rspace 129 11.3 0.004 0.005 16.003 16.096 multiply_cannon_sync_h2d 10116 15.6 14.441 14.487 14.441 14.487 hybrid_alltoall_any 5245 16.5 1.322 3.089 10.929 13.095 make_images_data 5058 15.6 0.064 0.069 10.733 12.941 init_scf_run 11 5.9 0.000 0.001 12.450 12.452 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.450 12.452 pw_transfer 1559 11.6 0.095 0.097 11.361 11.367 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 11.121 11.126 fft_wrap_pw1pw2_140 527 13.2 3.072 3.144 9.932 9.942 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.818 9.838 dbcsr_mm_accdrv_process 20934 16.1 4.235 6.032 7.586 9.455 wfi_extrapolate 11 7.9 0.001 0.001 9.010 9.010 cp_fm_cholesky_decompose 22 10.9 8.862 8.884 8.862 8.884 grid_integrate_task_list 129 12.3 8.617 8.812 8.617 8.812 qs_energies_init_hamiltonians 11 5.9 0.010 0.020 8.030 8.031 density_rs2pw 129 9.7 0.005 0.005 7.500 7.545 mp_alltoall_d11v 2429 14.1 6.957 7.122 6.957 7.122 calculate_dm_sparse 129 9.5 0.001 0.001 6.846 6.893 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.500 6.552 grid_collocate_task_list 129 9.7 6.463 6.537 6.463 6.537 fft3d_ps 1301 14.7 2.872 2.887 6.330 6.370 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.199 6.286 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=289.875000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2908.454545, yerr=156.772194 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_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 1.259926E+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 18892040. 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 0.022 0.067 89.506 89.507 qs_energies 1 2.0 0.000 0.001 88.943 88.949 ls_scf 1 3.0 0.000 0.000 88.038 88.045 dbcsr_multiply_generic 111 6.7 0.014 0.014 77.115 77.331 multiply_cannon 111 7.7 0.018 0.021 60.467 61.714 multiply_cannon_loop 111 8.7 0.217 0.231 56.933 58.339 ls_scf_main 1 4.0 0.000 0.000 52.262 52.262 density_matrix_trs4 2 5.0 0.002 0.003 46.701 46.788 ls_scf_init_scf 1 4.0 0.000 0.000 32.738 32.738 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.620 31.673 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.529 29.549 mp_waitall_1 11031 10.9 22.996 28.136 22.996 28.136 multiply_cannon_multrec 2664 9.7 8.150 8.829 15.517 17.178 multiply_cannon_sync_h2d 2664 9.7 13.438 15.267 13.438 15.267 make_m2s 222 7.7 0.009 0.011 13.022 13.615 make_images 222 8.7 0.099 0.108 13.000 13.595 multiply_cannon_metrocomm1 2664 9.7 0.010 0.010 9.896 13.137 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.819 12.131 make_images_data 222 9.7 0.004 0.005 7.612 8.256 dbcsr_mm_accdrv_process 4760 10.4 0.583 0.695 6.984 7.937 hybrid_alltoall_any 227 10.6 0.215 1.843 6.509 7.823 dbcsr_mm_accdrv_process_sort 4760 11.4 6.202 7.098 6.202 7.098 calculate_norms 4752 9.8 5.489 6.099 5.489 6.099 compute_matrix_preconditioner 1 6.0 0.000 0.001 5.382 5.387 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.069 5.169 acc_transpose_blocks 2664 9.7 0.017 0.019 4.371 4.765 mp_sum_l 887 5.1 3.112 4.687 3.112 4.687 acc_transpose_blocks_kernels 2664 10.7 0.034 0.037 4.242 4.630 jit_kernel_transpose 1 13.0 4.208 4.594 4.208 4.594 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.327 3.573 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.079 3.510 mp_irecv_dv 6231 10.9 2.063 3.481 2.063 3.481 make_images_sizes 222 9.7 0.000 0.000 0.701 3.344 mp_alltoall_i44 222 10.7 0.701 3.344 0.701 3.344 arnoldi_extremal 4 6.8 0.000 0.000 3.219 3.251 arnoldi_normal_ev 4 7.8 0.001 0.002 3.218 3.251 build_subspace 16 8.4 0.009 0.012 3.126 3.128 ls_scf_post 1 4.0 0.000 0.000 3.039 3.045 ls_scf_store_result 1 5.0 0.000 0.000 2.860 2.902 dbcsr_special_finalize 555 9.7 0.005 0.006 2.354 2.818 dbcsr_merge_single_wm 555 10.7 0.479 0.616 2.345 2.809 make_images_pack 222 9.7 2.208 2.632 2.209 2.633 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.307 2.570 dbcsr_sort_data 658 11.4 2.123 2.527 2.123 2.527 dbcsr_matrix_vector_mult_local 304 10.0 2.066 2.480 2.068 2.481 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.296 2.362 buffer_matrices_ensure_size 222 8.7 1.812 2.170 1.812 2.170 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=89.507000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_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 2.164621E+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 0.061 0.132 89.901 89.902 qs_energies 1 2.0 0.005 0.038 88.951 88.959 ls_scf 1 3.0 0.000 0.000 87.487 87.506 dbcsr_multiply_generic 111 6.7 0.016 0.019 73.493 73.820 multiply_cannon 111 7.7 0.034 0.084 52.372 56.111 ls_scf_main 1 4.0 0.001 0.018 53.928 53.933 multiply_cannon_loop 111 8.7 0.135 0.146 49.610 52.700 density_matrix_trs4 2 5.0 0.003 0.028 48.225 48.410 ls_scf_init_scf 1 4.0 0.000 0.001 29.808 29.809 mp_waitall_1 9105 10.9 20.207 29.608 20.207 29.608 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.532 28.624 multiply_cannon_multrec 1332 9.7 13.413 17.978 22.697 28.411 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.158 26.169 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.136 20.395 make_m2s 222 7.7 0.007 0.008 14.853 15.424 make_images 222 8.7 1.374 1.698 14.822 15.393 dbcsr_mm_accdrv_process 4041 10.4 0.325 0.549 8.881 10.526 dbcsr_mm_accdrv_process_sort 4041 11.4 8.400 9.977 8.400 9.977 make_images_data 222 9.7 0.004 0.005 8.563 9.302 hybrid_alltoall_any 227 10.6 0.542 2.570 7.949 9.039 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.195 7.769 mp_irecv_dv 3311 11.0 3.175 7.717 3.175 7.717 mp_sum_l 887 5.1 4.804 7.627 4.804 7.627 calculate_norms 2376 9.8 6.069 6.829 6.069 6.829 multiply_cannon_sync_h2d 1332 9.7 4.872 6.273 4.872 6.273 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.543 5.949 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.940 5.167 arnoldi_extremal 4 6.8 0.000 0.000 4.623 4.640 arnoldi_normal_ev 4 7.8 0.001 0.005 4.623 4.639 build_subspace 16 8.4 0.015 0.021 4.372 4.375 ls_scf_post 1 4.0 0.002 0.014 3.751 3.771 ls_scf_store_result 1 5.0 0.000 0.000 3.421 3.565 dbcsr_matrix_vector_mult 304 9.0 0.010 0.023 3.134 3.373 dbcsr_matrix_vector_mult_local 304 10.0 2.751 3.240 2.753 3.242 mp_allgather_i34 111 8.7 0.746 2.867 0.746 2.867 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.633 2.747 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.151 2.671 dbcsr_data_new 4174 10.1 2.107 2.400 2.107 2.400 make_images_pack 222 9.7 1.826 2.109 1.829 2.111 dbcsr_sort_data 436 11.2 1.798 2.013 1.798 2.013 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.910 1.911 rebuild_ks_matrix 3 7.3 0.000 0.000 1.897 1.899 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.004 1.897 1.898 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.902000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1807.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_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 2.917908E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.053 92.660 92.661 qs_energies 1 2.0 0.000 0.000 92.044 92.049 ls_scf 1 3.0 0.000 0.000 90.615 90.619 dbcsr_multiply_generic 111 6.7 0.016 0.016 75.362 75.620 ls_scf_main 1 4.0 0.000 0.000 56.731 56.735 multiply_cannon 111 7.7 0.033 0.059 51.939 55.725 multiply_cannon_loop 111 8.7 0.117 0.132 49.176 53.692 density_matrix_trs4 2 5.0 0.002 0.003 50.869 51.063 mp_waitall_1 7281 11.0 23.483 33.846 23.483 33.846 ls_scf_init_scf 1 4.0 0.000 0.000 30.221 30.224 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.020 29.111 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.645 26.654 multiply_cannon_multrec 888 9.7 12.632 15.309 21.211 24.658 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 10.846 23.690 make_m2s 222 7.7 0.006 0.007 16.370 17.128 make_images 222 8.7 1.582 1.849 16.332 17.088 make_images_data 222 9.7 0.004 0.005 9.520 10.588 hybrid_alltoall_any 227 10.6 0.643 2.941 9.063 10.279 dbcsr_mm_accdrv_process 3754 10.4 0.308 0.508 8.086 9.335 mp_sum_l 887 5.1 5.364 9.076 5.364 9.076 dbcsr_mm_accdrv_process_sort 3754 11.4 7.649 8.827 7.649 8.827 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.134 7.279 multiply_cannon_sync_h2d 888 9.7 6.064 7.245 6.064 7.245 multiply_cannon_metrocomm4 777 9.7 0.004 0.006 2.496 7.197 mp_irecv_dv 2335 11.1 2.480 7.130 2.480 7.130 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.763 6.978 arnoldi_extremal 4 6.8 0.000 0.000 5.066 5.080 arnoldi_normal_ev 4 7.8 0.001 0.005 5.066 5.080 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.711 4.948 build_subspace 16 8.4 0.014 0.020 4.768 4.774 calculate_norms 1584 9.8 4.331 4.747 4.331 4.747 mp_allgather_i34 111 8.7 0.884 3.821 0.884 3.821 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.441 3.776 ls_scf_post 1 4.0 0.000 0.000 3.663 3.667 dbcsr_matrix_vector_mult_local 304 10.0 3.037 3.615 3.039 3.618 ls_scf_store_result 1 5.0 0.000 0.000 3.417 3.487 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.827 2.915 dbcsr_data_new 4116 9.9 2.108 2.461 2.108 2.461 make_images_sizes 222 9.7 0.000 0.000 1.002 2.209 mp_alltoall_i44 222 10.7 1.002 2.208 1.002 2.208 dbcsr_sort_data 325 11.1 1.882 2.185 1.882 2.185 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.886 1.888 make_images_pack 222 9.7 1.624 1.882 1.627 1.885 dbcsr_finalize 304 7.8 0.026 0.032 1.597 1.877 rebuild_ks_matrix 3 7.3 0.000 0.000 1.868 1.870 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.868 1.870 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.661000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2285.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_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 3.382530E+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 0.033 0.053 96.583 96.584 qs_energies 1 2.0 0.000 0.000 96.008 96.014 ls_scf 1 3.0 0.000 0.000 94.353 94.358 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.156 78.333 ls_scf_main 1 4.0 0.000 0.000 58.577 58.578 multiply_cannon 111 7.7 0.052 0.100 51.662 55.991 density_matrix_trs4 2 5.0 0.002 0.003 52.475 52.589 multiply_cannon_loop 111 8.7 0.154 0.170 46.630 49.748 ls_scf_init_scf 1 4.0 0.000 0.000 32.563 32.566 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.364 31.425 mp_waitall_1 6369 11.0 22.632 30.279 22.632 30.279 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.879 28.896 multiply_cannon_multrec 1332 9.7 14.171 17.419 22.062 25.436 make_m2s 222 7.7 0.007 0.008 21.080 22.455 make_images 222 8.7 3.155 3.626 21.030 22.408 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.243 16.780 make_images_data 222 9.7 0.004 0.004 11.729 13.349 hybrid_alltoall_any 227 10.6 0.798 3.789 11.122 12.911 dbcsr_mm_accdrv_process 3641 10.4 0.293 0.477 7.524 9.027 dbcsr_mm_accdrv_process_sort 3641 11.4 7.086 8.534 7.086 8.534 mp_sum_l 887 5.1 3.843 6.922 3.843 6.922 multiply_cannon_sync_h2d 1332 9.7 5.440 6.080 5.440 6.080 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.091 6.060 mp_irecv_dv 3229 10.9 2.066 5.986 2.066 5.986 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.032 5.483 arnoldi_extremal 4 6.8 0.000 0.000 5.233 5.248 arnoldi_normal_ev 4 7.8 0.001 0.004 5.233 5.247 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.490 4.926 build_subspace 16 8.4 0.014 0.021 4.899 4.907 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.533 4.692 calculate_norms 2376 9.8 4.206 4.645 4.206 4.645 mp_allgather_i34 111 8.7 2.146 4.563 2.146 4.563 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.596 3.899 dbcsr_matrix_vector_mult_local 304 10.0 3.215 3.718 3.217 3.719 dbcsr_sort_data 658 11.4 3.074 3.330 3.074 3.330 ls_scf_post 1 4.0 0.000 0.000 3.213 3.219 dbcsr_special_finalize 555 9.7 0.006 0.007 2.832 3.173 dbcsr_merge_single_wm 555 10.7 0.539 0.663 2.824 3.165 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.019 3.077 ls_scf_store_result 1 5.0 0.000 0.000 2.961 3.021 dbcsr_data_release 10477 10.7 1.582 2.445 1.582 2.445 dbcsr_finalize 304 7.8 0.049 0.061 1.792 1.986 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.584000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2724.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_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 4.747866E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.061 91.927 91.928 qs_energies 1 2.0 0.000 0.000 91.034 91.039 ls_scf 1 3.0 0.000 0.000 89.093 89.098 dbcsr_multiply_generic 111 6.7 0.018 0.019 70.569 70.796 ls_scf_main 1 4.0 0.000 0.000 56.494 56.494 multiply_cannon 111 7.7 0.070 0.130 52.294 55.905 multiply_cannon_loop 111 8.7 0.089 0.096 49.712 51.097 density_matrix_trs4 2 5.0 0.002 0.003 49.514 49.595 ls_scf_init_scf 1 4.0 0.000 0.000 29.195 29.196 mp_waitall_1 5436 11.0 24.192 28.702 24.192 28.702 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.926 27.968 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.820 25.835 multiply_cannon_multrec 444 9.7 13.760 16.153 20.821 23.094 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.696 15.321 make_m2s 222 7.7 0.005 0.005 13.612 14.540 make_images 222 8.7 2.041 2.479 13.545 14.470 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 5.974 13.678 hybrid_alltoall_any 227 10.6 0.801 3.839 8.116 9.769 make_images_data 222 9.7 0.003 0.004 8.338 9.621 multiply_cannon_sync_h2d 444 9.7 6.745 8.250 6.745 8.250 dbcsr_mm_accdrv_process 3003 10.4 0.351 0.404 6.757 7.888 dbcsr_mm_accdrv_process_sort 3003 11.4 6.395 7.491 6.395 7.491 arnoldi_extremal 4 6.8 0.000 0.000 5.926 5.938 arnoldi_normal_ev 4 7.8 0.001 0.005 5.926 5.938 build_subspace 16 8.4 0.015 0.019 5.534 5.542 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.453 4.659 mp_sum_l 887 5.1 2.663 4.616 2.663 4.616 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.240 4.468 dbcsr_matrix_vector_mult_local 304 10.0 3.738 4.221 3.740 4.223 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.525 3.893 mp_irecv_dv 1241 11.2 1.512 3.867 1.512 3.867 calculate_norms 792 9.8 3.616 3.792 3.616 3.792 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.878 3.688 mp_allgather_i34 111 8.7 1.149 3.640 1.149 3.640 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.520 3.606 ls_scf_post 1 4.0 0.000 0.000 3.404 3.409 make_images_sizes 222 9.7 0.000 0.000 0.892 3.362 mp_alltoall_i44 222 10.7 0.892 3.362 0.892 3.362 ls_scf_store_result 1 5.0 0.000 0.000 3.193 3.226 dbcsr_finalize 304 7.8 0.062 0.076 2.197 2.317 dbcsr_data_new 4608 9.7 1.804 2.229 1.804 2.229 dbcsr_merge_all 275 8.9 0.478 0.523 2.055 2.147 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.026 2.027 rebuild_ks_matrix 3 7.3 0.000 0.000 1.994 1.995 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.994 1.995 qs_energies_init_hamiltonians 1 3.0 0.003 0.010 1.925 1.925 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.928000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3783.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_performance_tests/26/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 30.358840E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 339931 0.0% 0.0% 100.0% average stack size 0.0 0.0 28341.7 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 8.810639E+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 0.074 0.094 105.874 105.875 qs_energies 1 2.0 0.000 0.000 104.419 104.424 ls_scf 1 3.0 0.000 0.000 101.497 101.501 dbcsr_multiply_generic 111 6.7 0.024 0.026 75.119 75.253 ls_scf_main 1 4.0 0.000 0.000 64.128 64.128 density_matrix_trs4 2 5.0 0.002 0.003 55.179 55.246 multiply_cannon 111 7.7 0.142 0.220 48.934 51.188 multiply_cannon_loop 111 8.7 0.097 0.099 45.974 46.327 ls_scf_init_scf 1 4.0 0.000 0.000 33.567 33.567 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.089 32.095 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.379 29.387 mp_waitall_1 4527 11.1 22.468 26.112 22.468 26.112 make_m2s 222 7.7 0.005 0.005 22.656 23.720 make_images 222 8.7 3.575 3.873 22.548 23.611 multiply_cannon_multrec 444 9.7 17.796 18.446 22.430 23.143 hybrid_alltoall_any 227 10.6 1.651 3.615 12.826 15.744 make_images_data 222 9.7 0.003 0.004 13.044 15.180 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.819 11.136 multiply_cannon_sync_h2d 444 9.7 8.770 8.824 8.770 8.824 arnoldi_extremal 4 6.8 0.000 0.000 7.581 7.591 arnoldi_normal_ev 4 7.8 0.002 0.008 7.580 7.590 build_subspace 16 8.4 0.026 0.036 7.025 7.037 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.618 5.773 dbcsr_matrix_vector_mult_local 304 10.0 5.137 5.451 5.139 5.454 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.064 5.154 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.891 5.147 dbcsr_mm_accdrv_process 1814 10.4 0.269 0.354 4.448 4.573 dbcsr_mm_accdrv_process_sort 1814 11.4 4.111 4.243 4.111 4.243 ls_scf_post 1 4.0 0.000 0.000 3.803 3.808 make_images_sizes 222 9.7 0.000 0.000 1.489 3.647 mp_alltoall_i44 222 10.7 1.489 3.646 1.489 3.646 ls_scf_store_result 1 5.0 0.000 0.000 3.524 3.558 mp_allgather_i34 111 8.7 1.078 3.529 1.078 3.529 calculate_norms 792 9.8 3.251 3.324 3.251 3.324 dbcsr_finalize 304 7.8 0.082 0.089 3.068 3.148 dbcsr_merge_all 275 8.9 0.883 0.913 2.855 2.928 qs_energies_init_hamiltonians 1 3.0 0.000 0.000 2.892 2.892 dbcsr_complete_redistribute 5 7.6 1.440 1.486 2.750 2.856 dbcsr_sort_data 325 11.1 2.435 2.491 2.435 2.491 matrix_ls_to_qs 2 6.0 0.000 0.000 2.393 2.491 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.382 2.384 rebuild_ks_matrix 3 7.3 0.000 0.000 2.316 2.318 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.316 2.318 dbcsr_data_new 6591 9.6 1.858 2.308 1.858 2.308 dbcsr_new_transposed 4 7.5 0.240 0.251 2.279 2.292 dbcsr_frobenius_norm 74 6.6 2.055 2.132 2.190 2.224 dbcsr_add_d 103 6.2 0.000 0.000 2.123 2.200 dbcsr_add_anytype 103 7.2 0.858 0.890 2.123 2.200 dbcsr_data_release 12724 10.6 1.969 2.159 1.969 2.159 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.875000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6976.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/737e746daa79aac850212d2d3a1eec9f85e831b5_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 593.309696E+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 4803995. 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 0.038 0.116 225.621 225.674 qs_mol_dyn_low 1 2.0 0.005 0.066 224.358 224.412 qs_forces 5 3.8 0.004 0.005 224.218 224.224 qs_energies 5 4.8 0.004 0.034 221.178 221.209 scf_env_do_scf 5 5.8 0.000 0.009 205.673 205.676 scf_env_do_scf_inner_loop 105 6.6 0.003 0.029 179.558 179.580 qs_scf_new_mos 105 7.6 0.000 0.001 140.454 140.624 qs_scf_loop_do_ot 105 8.6 0.001 0.001 140.454 140.623 dbcsr_multiply_generic 1445 12.2 0.127 0.137 134.293 134.784 ot_scf_mini 105 9.6 0.003 0.004 130.438 130.589 multiply_cannon 1445 13.2 0.278 0.350 115.358 117.501 multiply_cannon_loop 1445 14.2 2.712 2.916 113.631 115.286 velocity_verlet 4 3.0 0.002 0.008 106.264 106.265 ot_mini 105 10.6 0.001 0.002 59.995 60.143 mp_waitall_1 488190 16.1 34.953 45.202 34.953 45.202 multiply_cannon_multrec 69360 15.2 30.199 35.251 39.479 44.888 qs_ot_get_p 112 10.4 0.001 0.002 40.363 40.760 qs_ot_get_derivative 55 11.6 0.001 0.001 38.149 38.293 multiply_cannon_metrocomm3 69360 15.2 0.204 0.216 25.866 36.277 multiply_cannon_sync_h2d 69360 15.2 29.670 34.229 29.670 34.229 qs_ot_p2m_diag 40 11.0 0.020 0.031 29.248 29.365 rebuild_ks_matrix 110 8.4 0.000 0.000 28.608 28.772 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.017 28.608 28.772 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.174 26.322 init_scf_loop 7 6.6 0.000 0.005 26.083 26.085 cp_dbcsr_syevd 40 12.0 0.002 0.002 26.075 26.075 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.073 23.326 apply_single 62 13.6 0.000 0.000 23.073 23.326 prepare_preconditioner 7 7.6 0.000 0.000 21.251 21.283 make_preconditioner 7 8.6 0.000 0.002 21.251 21.282 cp_fm_syevd 40 13.0 0.000 0.001 21.074 21.208 ot_new_cg_direction 55 11.6 0.001 0.002 21.109 21.110 qs_rho_update_rho_low 110 7.6 0.001 0.001 16.255 16.695 calculate_rho_elec 110 8.6 0.030 0.032 16.254 16.695 cp_fm_redistribute_end 40 14.0 8.206 16.375 8.211 16.377 cp_fm_syevd_base 40 14.0 8.158 16.328 8.158 16.328 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.817 14.982 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.245 14.312 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.918 14.023 mp_sum_l 4764 12.2 12.249 13.147 12.249 13.147 init_scf_run 5 5.8 0.000 0.001 12.808 12.809 scf_env_initial_rho_setup 5 6.8 0.001 0.003 12.808 12.809 calculate_dm_sparse 110 9.5 0.000 0.001 11.573 11.742 pw_transfer 1645 12.4 0.079 0.101 11.448 11.692 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 11.310 11.559 density_rs2pw 110 9.6 0.006 0.007 10.500 10.858 qs_vxc_create 110 10.4 0.002 0.005 10.302 10.350 dbcsr_mm_accdrv_process 154766 15.8 5.700 6.286 9.148 10.261 fft_wrap_pw1pw2_240 915 15.0 1.147 1.259 9.958 10.184 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.021 10.083 cp_fm_cholesky_invert 7 10.6 10.018 10.026 10.018 10.026 check_diag 80 13.5 8.687 8.938 9.379 9.508 acc_transpose_blocks 69360 15.2 0.329 0.369 8.040 8.888 fft3d_pb 915 16.0 2.401 2.720 8.150 8.429 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 8.379 8.395 sum_up_and_integrate 60 10.3 0.001 0.003 8.287 8.301 integrate_v_rspace 60 11.3 0.002 0.002 8.270 8.284 multiply_cannon_metrocomm1 69360 15.2 0.097 0.104 4.716 8.243 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.001 7.987 8.036 calculate_first_density_matrix 1 7.0 0.000 0.003 7.991 8.011 transfer_rs2pw 445 10.6 0.007 0.008 7.323 7.675 xc_rho_set_and_dset_create 110 12.4 0.077 0.098 7.241 7.500 xc_vxc_pw_create 60 11.3 0.038 0.050 6.891 6.938 make_m2s 2890 13.2 0.078 0.087 6.241 6.880 make_images 2890 14.2 0.240 0.259 6.134 6.774 make_full_single_inverse 7 9.6 0.001 0.004 6.731 6.762 xc_pw_derive 510 13.4 0.006 0.007 6.066 6.154 acc_transpose_blocks_kernels 69360 16.2 0.798 0.883 5.523 6.058 mp_alltoall_z22v 2340 17.7 5.163 5.415 5.163 5.415 jit_kernel_transpose 5 15.0 4.725 5.214 4.725 5.214 mp_waitany 7680 13.5 4.479 5.108 4.479 5.108 potential_pw2rs 60 12.3 0.003 0.003 4.570 4.598 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=225.674000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.200000, yerr=3.487119 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 737e746daa79aac850212d2d3a1eec9f85e831b5 Summary: empty Status: OK