=== This is the CP2K Performance-Test === Updating 427e19860..d51b7226a Fast-forward tools/docker/scripts/install_python.sh | 2 +- tools/toolchain/scripts/stage2/install_openblas.sh | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: d51b7226a3a06d954c541917aa996a97b60aada9 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (12.01.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/01 job id: 44580510 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/02 job id: 44580511 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/03 job id: 44580512 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/04 job id: 44580513 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/05 job id: 44580515 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/06 job id: 44580517 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/07 job id: 44580520 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/08 job id: 44580525 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/09 job id: 44580527 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/10 job id: 44580533 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/11 job id: 44580535 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/12 job id: 44580537 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/13 job id: 44580538 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/14 job id: 44580539 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/15 job id: 44580540 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/16 job id: 44580541 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/17 job id: 44580542 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/18 job id: 44580544 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/19 job id: 44580545 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/20 job id: 44580547 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/21 job id: 44580548 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/22 job id: 44580549 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/23 job id: 44580550 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/24 job id: 44580551 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/25 job id: 44580552 --- 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/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/26 job id: 44580553 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.034 137.088 137.089 farming_run 1 2.0 136.480 136.484 137.059 137.062 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.460654E+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 229 1108280. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.027 116.829 116.829 qs_energies 1 2.0 0.000 0.000 116.604 116.605 mp2_main 1 3.0 0.000 0.000 114.537 114.538 mp2_gpw_main 1 4.0 0.020 0.026 113.653 113.654 mp2_ri_gpw_compute_in 1 5.0 0.172 0.175 93.978 94.506 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.545 56.073 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.169 41.842 47.290 get_2c_integrals 1 6.0 0.000 0.001 37.740 38.260 integrate_v_rspace 273 8.0 0.436 0.449 25.195 30.417 pw_transfer 6555 10.6 0.375 0.394 27.543 28.091 grid_integrate_task_list 273 9.0 20.991 26.715 20.991 26.715 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 26.183 26.622 fft_wrap_pw1pw2_100 2178 12.4 1.222 1.469 23.692 24.145 compute_2c_integrals 1 7.0 0.002 0.003 19.867 19.869 rpa_ri_compute_en 1 5.0 0.083 0.451 19.567 19.787 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.993 19.559 mp2_eri_2c_integrate_gpw 1 9.0 2.355 2.421 18.990 19.555 cp_fm_cholesky_decompose 12 8.2 17.867 18.389 17.867 18.389 cholesky_decomp 1 7.0 0.000 0.000 16.711 17.226 fft3d_s 5443 13.4 16.216 16.683 16.238 16.704 ao_to_mo_and_store_B_mult_1 272 7.0 10.868 15.605 10.868 15.605 calculate_wavefunction 272 8.0 5.421 5.574 12.559 13.164 rpa_num_int 1 6.0 0.002 0.038 10.885 10.894 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.633 10.839 calc_mat_Q 8 8.0 0.000 0.000 9.428 9.707 contract_S_to_Q 8 9.0 0.000 0.000 8.850 9.128 calc_potential_gpw 544 9.5 0.006 0.006 8.300 8.735 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.294 8.647 parallel_gemm_fm 14 9.1 0.000 0.000 8.441 8.636 parallel_gemm_fm_cosma 14 10.1 8.441 8.636 8.441 8.636 potential_pw2rs 545 10.0 0.107 0.110 7.753 8.400 collocate_single_gaussian 272 10.0 0.039 0.042 7.524 7.829 create_integ_mat 1 6.0 0.014 0.027 7.606 7.615 array2fm 1 7.0 0.000 0.000 6.615 7.013 pw_scatter_s 2720 13.7 4.421 4.583 4.421 4.583 pw_gather_s 2722 13.2 3.902 4.218 3.902 4.218 array2fm_buffer_send 1 8.0 2.969 3.123 2.969 3.123 pw_poisson_solve 545 10.5 1.130 1.183 2.198 2.398 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.653206, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2732.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.059 397.438 397.440 farming_run 1 2.0 396.391 396.395 397.380 397.404 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.221673E+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 704 407793. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.053 209.512 209.513 qs_energies 1 2.0 0.000 0.000 209.122 209.140 scf_env_do_scf 1 3.0 0.000 0.000 105.900 105.900 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.013 105.022 rebuild_ks_matrix 4 6.0 0.000 0.000 105.012 105.020 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.066 105.012 105.020 hfx_ks_matrix 4 8.0 0.001 0.001 104.627 104.630 integrate_four_center 4 9.0 0.144 0.458 104.626 104.630 mp2_main 1 3.0 0.000 0.000 102.919 102.938 mp2_gpw_main 1 4.0 0.032 0.047 102.038 102.059 integrate_four_center_main 4 10.0 0.089 0.537 96.518 99.087 integrate_four_center_bin 268 11.0 96.428 99.076 96.428 99.076 init_scf_loop 1 4.0 0.000 0.000 91.629 91.629 mp2_ri_gpw_compute_in 1 5.0 0.065 0.073 75.080 76.365 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.623 55.907 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.159 41.984 47.251 integrate_v_rspace 95 8.0 0.396 0.566 28.342 33.401 pw_transfer 2240 10.6 0.144 0.160 29.908 30.316 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.925 29.333 grid_integrate_task_list 95 9.0 23.634 28.893 23.634 28.893 mp2_ri_gpw_compute_en 1 5.0 0.062 0.081 26.810 28.600 ao_to_mo_and_store_B_mult_1 91 7.0 10.953 27.287 10.953 27.287 fft_wrap_pw1pw2_100 730 12.4 1.271 1.430 26.638 27.037 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.846 1.897 24.864 24.874 get_2c_integrals 1 6.0 0.001 0.007 20.350 20.394 compute_2c_integrals 1 7.0 0.003 0.009 19.332 19.337 compute_2c_integrals_loop_lm 1 8.0 0.002 0.010 18.824 19.180 mp2_eri_2c_integrate_gpw 1 9.0 1.734 1.836 18.821 19.179 fft3d_s 1823 13.4 18.420 18.730 18.433 18.744 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.269 14.269 calculate_wavefunction 91 8.0 2.033 2.065 9.756 9.951 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.579 8.762 9.345 potential_pw2rs 186 10.0 0.034 0.037 8.641 9.180 local_gemm 172 8.0 8.207 8.766 8.207 8.766 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.217 8.629 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.524 7.865 8.256 collocate_single_gaussian 91 10.0 0.016 0.019 7.873 8.159 calc_potential_gpw 182 9.5 0.002 0.002 7.883 8.151 mp2_ri_gpw_compute_en_ener 172 7.0 6.346 6.420 6.346 6.420 mp_sendrecv_dm3 2068 8.0 5.898 6.298 5.898 6.298 pw_gather_s 912 13.2 4.924 5.400 4.924 5.400 mp_sync 38 10.4 2.816 5.073 2.816 5.073 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.032735, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1509.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_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 452.104192E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 578368. 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 0.031 0.042 57.112 57.114 qs_mol_dyn_low 1 2.0 0.003 0.003 56.528 56.547 qs_forces 11 3.9 0.004 0.020 56.460 56.461 qs_energies 11 4.9 0.001 0.003 54.922 54.940 scf_env_do_scf 11 5.9 0.000 0.001 47.166 47.167 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 45.033 45.034 dbcsr_multiply_generic 2286 12.5 0.095 0.098 37.372 37.994 qs_scf_new_mos 108 7.5 0.000 0.001 34.609 34.915 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.609 34.915 ot_scf_mini 108 9.5 0.002 0.002 32.933 33.139 multiply_cannon 2286 13.5 0.184 0.192 29.173 31.171 multiply_cannon_loop 2286 14.5 1.484 1.562 28.405 30.370 velocity_verlet 10 3.0 0.001 0.001 26.234 26.234 ot_mini 108 10.5 0.001 0.001 21.377 21.642 qs_ot_get_derivative 108 11.5 0.001 0.001 18.389 18.584 mp_waitall_1 245248 16.5 9.747 17.747 9.747 17.747 multiply_cannon_metrocomm3 54864 15.5 0.068 0.073 6.974 16.221 multiply_cannon_multrec 54864 15.5 4.187 6.426 9.235 13.395 multiply_cannon_sync_h2d 54864 15.5 5.845 8.332 5.845 8.332 rebuild_ks_matrix 119 8.3 0.000 0.000 8.141 8.296 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.140 8.295 mp_sum_l 7207 12.9 6.047 7.966 6.047 7.966 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.157 7.299 dbcsr_mm_accdrv_process 76910 16.1 1.174 1.820 4.971 7.091 qs_ot_get_p 119 10.4 0.001 0.001 6.779 7.074 init_scf_run 11 5.9 0.000 0.001 6.427 6.427 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.427 6.427 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.960 6.079 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.602 6.057 jit_kernel_multiply 13 15.8 3.735 5.269 3.735 5.269 sum_up_and_integrate 119 10.3 0.012 0.015 4.680 4.687 integrate_v_rspace 119 11.3 0.002 0.003 4.668 4.676 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.295 4.399 calculate_rho_elec 119 8.7 0.011 0.017 4.295 4.399 calculate_first_density_matrix 1 7.0 0.000 0.001 4.198 4.205 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.782 3.865 multiply_cannon_metrocomm1 54864 15.5 0.051 0.057 2.033 3.691 calculate_dm_sparse 119 9.5 0.000 0.001 3.498 3.651 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.394 3.396 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.094 3.164 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.928 3.130 apply_single 119 13.6 0.000 0.000 2.927 3.130 rs_pw_transfer 974 11.9 0.011 0.012 2.893 3.022 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.938 2.939 cp_fm_diag_elpa 50 13.0 0.008 0.031 2.727 2.727 ot_diis_step 108 11.5 0.006 0.006 2.706 2.707 cp_fm_redistribute_end 50 14.0 2.454 2.673 2.459 2.675 acc_transpose_blocks 54864 15.5 0.229 0.248 2.142 2.659 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.590 2.652 cp_fm_diag_elpa_base 50 14.0 0.213 2.574 0.214 2.584 density_rs2pw 119 9.7 0.004 0.005 2.319 2.438 wfi_extrapolate 11 7.9 0.001 0.001 2.168 2.168 grid_integrate_task_list 119 12.3 2.050 2.145 2.050 2.145 init_scf_loop 11 6.9 0.000 0.000 2.115 2.115 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.990 2.011 potential_pw2rs 119 12.3 0.004 0.004 1.990 2.009 mp_sum_d 4129 12.0 1.411 2.009 1.411 2.009 pw_transfer 1439 11.6 0.053 0.060 1.935 2.009 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.858 1.936 make_m2s 4572 13.5 0.054 0.056 1.853 1.901 make_images 4572 14.5 0.132 0.137 1.770 1.818 fft3d_ps 1201 14.6 0.371 0.473 1.629 1.697 mp_waitany 12084 13.8 1.324 1.532 1.324 1.532 fft_wrap_pw1pw2_140 487 13.2 0.082 0.097 1.433 1.513 acc_transpose_blocks_kernels 54864 16.5 0.251 0.385 1.153 1.510 mp_alltoall_d11v 2130 13.8 1.320 1.463 1.320 1.463 grid_collocate_task_list 119 9.7 1.290 1.380 1.290 1.380 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.747 1.165 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=57.114000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.363636, yerr=0.881396 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.653376E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 577664. 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 0.020 0.194 44.106 44.107 qs_mol_dyn_low 1 2.0 0.003 0.003 43.545 43.553 qs_forces 11 3.9 0.002 0.003 43.474 43.480 qs_energies 11 4.9 0.002 0.002 41.674 41.683 scf_env_do_scf 11 5.9 0.000 0.001 35.252 35.253 scf_env_do_scf_inner_loop 108 6.5 0.004 0.018 32.266 32.266 dbcsr_multiply_generic 2286 12.5 0.101 0.105 24.107 24.604 qs_scf_new_mos 108 7.5 0.001 0.001 22.002 22.253 qs_scf_loop_do_ot 108 8.5 0.001 0.001 22.002 22.253 ot_scf_mini 108 9.5 0.002 0.003 21.054 21.229 velocity_verlet 10 3.0 0.001 0.001 20.314 20.315 multiply_cannon 2286 13.5 0.209 0.217 18.235 19.881 multiply_cannon_loop 2286 14.5 0.898 0.973 16.934 18.745 ot_mini 108 10.5 0.001 0.001 13.120 13.369 mp_waitall_1 200699 16.5 7.328 13.244 7.328 13.244 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 5.024 11.093 qs_ot_get_derivative 108 11.5 0.001 0.001 10.510 10.686 multiply_cannon_multrec 27432 15.5 1.950 4.272 6.273 9.401 rebuild_ks_matrix 119 8.3 0.000 0.000 8.059 8.249 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 8.058 8.248 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.139 7.307 dbcsr_mm_accdrv_process 47894 16.0 3.343 5.663 4.254 6.562 init_scf_run 11 5.9 0.000 0.001 5.172 5.173 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.172 5.172 qs_ot_get_p 119 10.4 0.017 0.070 4.880 5.107 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.214 5.068 sum_up_and_integrate 119 10.3 0.024 0.027 4.888 4.902 integrate_v_rspace 119 11.3 0.002 0.002 4.865 4.880 mp_sum_l 7207 12.9 2.603 4.649 2.603 4.649 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.231 4.273 calculate_rho_elec 119 8.7 0.021 0.024 4.231 4.273 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.119 4.177 apply_single 119 13.6 0.000 0.000 3.119 4.176 rs_pw_transfer 974 11.9 0.010 0.011 3.403 3.906 calculate_first_density_matrix 1 7.0 0.000 0.001 3.473 3.475 make_m2s 4572 13.5 0.052 0.054 2.978 3.268 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.192 3.228 make_images 4572 14.5 0.199 0.234 2.890 3.177 density_rs2pw 119 9.7 0.004 0.004 2.517 3.026 init_scf_loop 11 6.9 0.000 0.000 2.960 2.961 multiply_cannon_sync_h2d 27432 15.5 2.188 2.834 2.188 2.834 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.788 2.792 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.691 2.707 potential_pw2rs 119 12.3 0.006 0.006 2.577 2.595 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.538 2.578 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.484 2.575 ot_diis_step 108 11.5 0.011 0.011 2.531 2.533 calculate_dm_sparse 119 9.5 0.000 0.001 2.402 2.477 acc_transpose_blocks 27432 15.5 0.108 0.114 1.522 2.280 pw_transfer 1439 11.6 0.065 0.071 2.213 2.251 jit_kernel_multiply 10 16.1 0.859 2.224 0.859 2.224 cp_fm_diag_elpa 50 13.0 0.006 0.010 2.191 2.191 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 2.121 2.161 cp_fm_redistribute_end 50 14.0 1.810 2.156 1.814 2.158 cp_fm_diag_elpa_base 50 14.0 0.329 2.017 0.342 2.072 grid_integrate_task_list 119 12.3 1.822 1.939 1.822 1.939 prepare_preconditioner 11 7.9 0.000 0.000 1.907 1.936 make_preconditioner 11 8.9 0.000 0.000 1.907 1.936 fft3d_ps 1201 14.6 0.514 0.569 1.823 1.858 make_images_data 4572 15.5 0.044 0.051 1.300 1.858 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.791 1.854 acc_transpose_blocks_kernels 27432 16.5 0.181 0.269 0.995 1.680 hybrid_alltoall_any 4725 16.4 0.050 0.111 1.144 1.676 wfi_extrapolate 11 7.9 0.001 0.001 1.638 1.639 fft_wrap_pw1pw2_140 487 13.2 0.079 0.084 1.566 1.606 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.584 1.593 mp_alltoall_d11v 2130 13.8 1.426 1.551 1.426 1.551 mp_allgather_i34 2286 14.5 0.744 1.547 0.744 1.547 jit_kernel_transpose 5 15.5 0.814 1.418 0.814 1.418 grid_collocate_task_list 119 9.7 1.228 1.358 1.228 1.358 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.309 1.348 make_images_sizes 4572 15.5 0.005 0.005 0.863 1.191 mp_alltoall_i44 4572 16.5 0.858 1.187 0.858 1.187 mp_sum_d 4129 12.0 0.699 1.160 0.699 1.160 mp_waitany 5720 13.7 0.573 1.088 0.573 1.088 rs_pw_transfer_PW2RS_50 119 14.3 0.585 0.603 1.012 1.067 rs_pw_transfer_RS2PW_140 130 11.5 0.138 0.147 0.565 1.062 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.982 1.003 mp_alltoall_z22v 1201 16.6 0.855 0.966 0.855 0.966 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.949 0.955 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=44.107000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.545455, yerr=1.499311 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_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 520.806400E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1247257. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.031 35.501 35.502 qs_mol_dyn_low 1 2.0 0.003 0.003 35.227 35.234 qs_forces 11 3.9 0.002 0.002 35.042 35.044 qs_energies 11 4.9 0.002 0.002 33.367 33.371 scf_env_do_scf 11 5.9 0.000 0.001 28.182 28.183 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.255 25.255 dbcsr_multiply_generic 2286 12.5 0.095 0.098 17.564 17.642 velocity_verlet 10 3.0 0.001 0.001 16.869 16.874 qs_scf_new_mos 108 7.5 0.001 0.001 16.241 16.255 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.240 16.254 ot_scf_mini 108 9.5 0.002 0.002 15.452 15.467 multiply_cannon 2286 13.5 0.196 0.204 13.732 14.514 multiply_cannon_loop 2286 14.5 0.635 0.663 12.839 13.675 ot_mini 108 10.5 0.001 0.001 9.431 9.444 qs_ot_get_derivative 108 11.5 0.001 0.001 7.865 7.878 multiply_cannon_multrec 18288 15.5 1.946 2.858 7.084 7.430 rebuild_ks_matrix 119 8.3 0.000 0.000 6.935 6.946 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.934 6.946 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.108 6.118 dbcsr_mm_accdrv_process 38222 16.0 4.360 5.705 5.055 5.809 mp_waitall_1 158411 16.6 3.358 4.506 3.358 4.506 sum_up_and_integrate 119 10.3 0.031 0.032 4.446 4.449 integrate_v_rspace 119 11.3 0.002 0.002 4.415 4.423 init_scf_run 11 5.9 0.000 0.001 4.000 4.001 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.999 4.001 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.883 3.893 calculate_rho_elec 119 8.7 0.030 0.031 3.882 3.893 qs_ot_get_p 119 10.4 0.001 0.001 3.745 3.766 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.947 3.564 rs_pw_transfer 974 11.9 0.009 0.010 2.824 3.094 init_scf_loop 11 6.9 0.000 0.000 2.907 2.907 calculate_first_density_matrix 1 7.0 0.000 0.000 2.646 2.647 density_rs2pw 119 9.7 0.004 0.004 2.378 2.643 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.556 2.562 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.490 2.471 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.077 2.388 apply_single 119 13.6 0.000 0.000 2.077 2.388 make_m2s 4572 13.5 0.044 0.045 2.200 2.365 make_images 4572 14.5 0.190 0.203 2.115 2.278 pw_transfer 1439 11.6 0.066 0.071 2.242 2.256 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.254 2.255 calculate_dm_sparse 119 9.5 0.000 0.000 2.198 2.208 potential_pw2rs 119 12.3 0.007 0.008 2.166 2.174 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 2.149 2.166 prepare_preconditioner 11 7.9 0.000 0.000 2.040 2.043 make_preconditioner 11 8.9 0.000 0.000 2.040 2.043 cp_fm_diag_elpa 50 13.0 0.110 0.296 1.974 1.976 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.863 1.954 grid_integrate_task_list 119 12.3 1.795 1.878 1.795 1.878 fft3d_ps 1201 14.6 0.529 0.547 1.829 1.844 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.820 1.821 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.752 1.760 cp_fm_redistribute_end 50 14.0 1.252 1.647 1.253 1.649 fft_wrap_pw1pw2_140 487 13.2 0.089 0.093 1.623 1.640 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.593 1.597 cp_fm_diag_elpa_base 50 14.0 0.377 1.524 0.392 1.575 mp_sum_l 7207 12.9 1.245 1.552 1.245 1.552 ot_diis_step 108 11.5 0.011 0.011 1.542 1.542 multiply_cannon_sync_h2d 18288 15.5 1.367 1.530 1.367 1.530 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.506 1.508 jit_kernel_multiply 8 16.5 0.644 1.495 0.644 1.495 grid_collocate_task_list 119 9.7 1.212 1.359 1.212 1.359 wfi_extrapolate 11 7.9 0.001 0.001 1.302 1.303 acc_transpose_blocks 18288 15.5 0.075 0.077 1.250 1.271 make_images_data 4572 15.5 0.044 0.048 0.983 1.191 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.858 1.088 mp_alltoall_d11v 2130 13.8 0.880 0.990 0.880 0.990 mp_alltoall_z22v 1201 16.6 0.877 0.978 0.877 0.978 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.943 0.976 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.954 0.957 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.899 0.901 multiply_cannon_metrocomm1 18288 15.5 0.028 0.029 0.611 0.871 mp_waitany 9880 13.7 0.554 0.834 0.554 0.834 acc_transpose_blocks_kernels 18288 16.5 0.209 0.217 0.812 0.832 cp_fm_cholesky_invert 11 10.9 0.813 0.817 0.813 0.817 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.121 0.526 0.797 make_images_sizes 4572 15.5 0.005 0.005 0.536 0.769 mp_alltoall_i44 4572 16.5 0.531 0.764 0.531 0.764 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.660 0.744 dbcsr_complete_redistribute 329 12.2 0.125 0.146 0.630 0.742 mp_allgather_i34 2286 14.5 0.343 0.725 0.343 0.725 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=35.502000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.545455, yerr=1.724879 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 553.140224E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.033 41.016 41.017 qs_mol_dyn_low 1 2.0 0.003 0.003 40.741 40.748 qs_forces 11 3.9 0.002 0.003 40.673 40.680 qs_energies 11 4.9 0.003 0.004 38.784 38.795 scf_env_do_scf 11 5.9 0.000 0.001 33.024 33.025 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 28.950 28.952 velocity_verlet 10 3.0 0.001 0.001 20.849 20.851 dbcsr_multiply_generic 2286 12.5 0.101 0.103 20.393 20.514 qs_scf_new_mos 108 7.5 0.001 0.001 18.745 18.797 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.744 18.796 ot_scf_mini 108 9.5 0.002 0.003 17.701 17.752 multiply_cannon 2286 13.5 0.230 0.269 15.463 15.944 multiply_cannon_loop 2286 14.5 0.934 0.967 14.266 14.676 ot_mini 108 10.5 0.001 0.001 10.643 10.717 multiply_cannon_multrec 27432 15.5 2.339 3.048 8.945 9.299 qs_ot_get_derivative 108 11.5 0.001 0.001 8.682 8.737 dbcsr_mm_accdrv_process 47916 15.9 5.646 7.599 6.513 7.886 rebuild_ks_matrix 119 8.3 0.000 0.000 7.731 7.800 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.731 7.800 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.861 6.926 sum_up_and_integrate 119 10.3 0.035 0.039 4.706 4.715 integrate_v_rspace 119 11.3 0.002 0.003 4.671 4.680 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.416 4.452 calculate_rho_elec 119 8.7 0.040 0.046 4.415 4.451 init_scf_run 11 5.9 0.000 0.001 4.359 4.359 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.359 4.359 qs_ot_get_p 119 10.4 0.001 0.001 4.180 4.260 init_scf_loop 11 6.9 0.000 0.000 4.037 4.039 mp_waitall_1 137007 16.6 3.124 3.921 3.124 3.921 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.181 3.660 make_m2s 4572 13.5 0.054 0.056 2.964 3.192 rs_pw_transfer 974 11.9 0.009 0.010 2.994 3.190 make_images 4572 14.5 0.271 0.332 2.857 3.084 prepare_preconditioner 11 7.9 0.000 0.000 3.072 3.081 make_preconditioner 11 8.9 0.000 0.000 3.072 3.081 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.630 2.983 density_rs2pw 119 9.7 0.004 0.004 2.697 2.879 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.231 2.801 apply_single 119 13.6 0.000 0.000 2.230 2.800 calculate_first_density_matrix 1 7.0 0.000 0.000 2.686 2.689 pw_transfer 1439 11.6 0.066 0.071 2.637 2.678 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.544 2.590 calculate_dm_sparse 119 9.5 0.000 0.000 2.494 2.553 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.448 2.459 potential_pw2rs 119 12.3 0.008 0.009 2.333 2.346 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.303 2.340 fft3d_ps 1201 14.6 0.563 0.614 2.220 2.261 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.083 2.083 mp_sum_l 7207 12.9 1.539 2.048 1.539 2.048 acc_transpose_blocks 27432 15.5 0.111 0.113 1.501 2.047 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.020 2.022 fft_wrap_pw1pw2_140 487 13.2 0.087 0.093 1.970 2.021 jit_kernel_multiply 9 15.8 0.807 1.999 0.807 1.999 ot_diis_step 108 11.5 0.012 0.012 1.904 1.905 grid_integrate_task_list 119 12.3 1.805 1.902 1.805 1.902 cp_fm_diag_elpa 50 13.0 0.004 0.007 1.726 1.729 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.689 1.724 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 1.033 1.693 cp_fm_redistribute_end 50 14.0 1.134 1.680 1.136 1.682 cp_fm_diag_elpa_base 50 14.0 0.521 1.594 0.541 1.634 wfi_extrapolate 11 7.9 0.001 0.001 1.608 1.608 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.578 1.589 make_images_data 4572 15.5 0.045 0.049 1.205 1.473 acc_transpose_blocks_kernels 27432 16.5 0.267 0.275 0.880 1.434 dbcsr_complete_redistribute 329 12.2 0.137 0.198 1.079 1.369 grid_collocate_task_list 119 9.7 1.222 1.345 1.222 1.345 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.300 1.324 hybrid_alltoall_any 4725 16.4 0.061 0.150 1.017 1.317 mp_alltoall_d11v 2130 13.8 1.190 1.300 1.190 1.300 mp_alltoall_z22v 1201 16.6 1.239 1.282 1.239 1.282 cp_fm_upper_to_full 72 13.5 0.853 1.191 0.853 1.191 jit_kernel_transpose 5 15.6 0.613 1.171 0.613 1.171 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.087 1.095 multiply_cannon_sync_h2d 27432 15.5 0.968 1.041 0.968 1.041 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.745 1.029 cp_fm_cholesky_invert 11 10.9 0.998 1.001 0.998 1.001 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.970 0.977 make_images_sizes 4572 15.5 0.005 0.005 0.738 0.975 mp_alltoall_i44 4572 16.5 0.733 0.970 0.733 0.970 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.869 yz_to_x 606 15.1 0.191 0.225 0.850 0.866 multiply_cannon_metrocomm1 27432 15.5 0.032 0.033 0.369 0.860 mp_alltoall_i22 627 13.8 0.517 0.832 0.517 0.832 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=41.017000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=525.909091, yerr=2.874798 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_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 620.793856E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.069 0.077 33.929 33.930 qs_mol_dyn_low 1 2.0 0.003 0.003 33.323 33.330 qs_forces 11 3.9 0.002 0.002 33.252 33.256 qs_energies 11 4.9 0.001 0.002 31.383 31.390 scf_env_do_scf 11 5.9 0.000 0.001 25.400 25.400 scf_env_do_scf_inner_loop 108 6.5 0.004 0.010 22.260 22.261 velocity_verlet 10 3.0 0.001 0.001 16.862 16.865 dbcsr_multiply_generic 2286 12.5 0.114 0.116 14.728 14.910 qs_scf_new_mos 108 7.5 0.001 0.001 13.126 13.155 qs_scf_loop_do_ot 108 8.5 0.001 0.001 13.126 13.154 ot_scf_mini 108 9.5 0.002 0.002 12.372 12.397 multiply_cannon 2286 13.5 0.231 0.236 11.332 11.813 multiply_cannon_loop 2286 14.5 0.330 0.342 10.252 10.673 ot_mini 108 10.5 0.001 0.001 7.259 7.295 rebuild_ks_matrix 119 8.3 0.000 0.000 6.793 6.843 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.793 6.842 multiply_cannon_multrec 9144 15.5 1.630 1.900 6.413 6.592 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.047 6.084 qs_ot_get_derivative 108 11.5 0.001 0.001 5.816 5.844 dbcsr_mm_accdrv_process 12550 15.8 3.428 4.481 4.682 4.753 init_scf_run 11 5.9 0.000 0.001 4.477 4.477 scf_env_initial_rho_setup 11 6.9 0.003 0.003 4.477 4.477 sum_up_and_integrate 119 10.3 0.038 0.042 4.191 4.198 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.167 4.185 calculate_rho_elec 119 8.7 0.060 0.061 4.167 4.184 integrate_v_rspace 119 11.3 0.003 0.003 4.153 4.161 qs_ot_get_p 119 10.4 0.001 0.001 3.184 3.216 init_scf_loop 11 6.9 0.000 0.000 3.101 3.103 mp_waitall_1 115863 16.7 2.449 3.090 2.449 3.090 calculate_first_density_matrix 1 7.0 0.000 0.000 3.023 3.024 density_rs2pw 119 9.7 0.004 0.004 2.480 2.635 pw_transfer 1439 11.6 0.066 0.071 2.558 2.573 rs_pw_transfer 974 11.9 0.008 0.009 2.354 2.506 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.464 2.481 make_m2s 4572 13.5 0.034 0.035 2.236 2.448 jit_kernel_multiply 10 15.5 1.216 2.368 1.216 2.368 make_images 4572 14.5 0.268 0.303 2.147 2.358 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.290 2.306 prepare_preconditioner 11 7.9 0.000 0.000 2.188 2.192 make_preconditioner 11 8.9 0.000 0.000 2.188 2.192 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.155 2.156 fft3d_ps 1201 14.6 0.566 0.578 2.122 2.135 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.067 2.099 calculate_dm_sparse 119 9.5 0.000 0.000 2.056 2.081 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.041 2.044 grid_integrate_task_list 119 12.3 1.860 1.988 1.860 1.988 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.963 1.980 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.911 1.925 potential_pw2rs 119 12.3 0.010 0.010 1.831 1.836 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.806 1.806 cp_fm_diag_elpa 50 13.0 0.001 0.003 1.515 1.517 cp_fm_redistribute_end 50 14.0 0.757 1.482 0.758 1.483 cp_fm_diag_elpa_base 50 14.0 0.679 1.395 0.722 1.452 ot_diis_step 108 11.5 0.013 0.013 1.419 1.419 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.398 1.408 grid_collocate_task_list 119 9.7 1.282 1.393 1.282 1.393 acc_transpose_blocks 9144 15.5 0.038 0.039 1.356 1.388 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.370 1.381 wfi_extrapolate 11 7.9 0.001 0.001 1.381 1.381 make_images_data 4572 15.5 0.039 0.042 1.057 1.374 qs_energies_init_hamiltonians 11 5.9 0.017 0.020 1.272 1.276 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.236 1.272 apply_single 119 13.6 0.000 0.000 1.236 1.272 hybrid_alltoall_any 4725 16.4 0.062 0.173 0.924 1.174 cp_fm_cholesky_invert 11 10.9 1.147 1.150 1.147 1.150 acc_transpose_blocks_kernels 9144 16.5 0.116 0.120 1.118 1.147 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.480 1.064 mp_alltoall_z22v 1201 16.6 1.004 1.054 1.004 1.054 jit_kernel_transpose 5 15.6 1.002 1.030 1.002 1.030 mp_alltoall_d11v 2130 13.8 0.921 1.009 0.921 1.009 mp_allgather_i34 2286 14.5 0.413 0.940 0.413 0.940 mp_sum_l 7207 12.9 0.716 0.940 0.716 0.940 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.922 multiply_cannon_metrocomm1 9144 15.5 0.021 0.022 0.684 0.918 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.877 0.890 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.818 0.828 yz_to_x 606 15.1 0.263 0.274 0.798 0.816 multiply_cannon_sync_h2d 9144 15.5 0.711 0.793 0.711 0.793 make_images_sizes 4572 15.5 0.005 0.005 0.518 0.765 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.708 0.762 mp_alltoall_i44 4572 16.5 0.513 0.760 0.513 0.760 x_to_yz 595 16.2 0.276 0.288 0.746 0.753 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=33.930000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=583.181818, yerr=6.249628 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_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 749.613056E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.059 46.070 46.071 qs_mol_dyn_low 1 2.0 0.003 0.003 45.629 45.637 qs_forces 11 3.9 0.002 0.002 45.557 45.558 qs_energies 11 4.9 0.008 0.009 43.498 43.501 scf_env_do_scf 11 5.9 0.001 0.001 37.351 37.352 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 28.865 28.866 velocity_verlet 10 3.0 0.001 0.001 25.600 25.606 dbcsr_multiply_generic 2286 12.5 0.102 0.102 19.230 19.576 qs_scf_new_mos 108 7.5 0.001 0.001 17.987 18.098 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.987 18.097 ot_scf_mini 108 9.5 0.002 0.002 16.834 16.959 multiply_cannon 2286 13.5 0.306 0.315 14.324 15.770 multiply_cannon_loop 2286 14.5 0.345 0.351 12.934 14.269 ot_mini 108 10.5 0.001 0.001 10.080 10.231 multiply_cannon_multrec 9144 15.5 3.483 4.781 8.749 9.034 init_scf_loop 11 6.9 0.000 0.000 8.450 8.452 rebuild_ks_matrix 119 8.3 0.000 0.000 8.130 8.318 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 8.130 8.318 qs_ot_get_derivative 108 11.5 0.001 0.001 7.894 8.007 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.372 7.544 prepare_preconditioner 11 7.9 0.000 0.000 7.335 7.353 make_preconditioner 11 8.9 0.000 0.000 7.335 7.353 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.856 7.219 dbcsr_mm_accdrv_process 12550 15.8 4.126 5.704 5.143 6.664 mp_waitall_1 94719 16.7 3.263 4.821 3.263 4.821 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.693 4.701 calculate_rho_elec 119 8.7 0.118 0.121 4.692 4.700 cp_fm_upper_to_full 72 14.2 3.185 4.599 3.185 4.599 sum_up_and_integrate 119 10.3 0.065 0.067 4.563 4.579 integrate_v_rspace 119 11.3 0.003 0.003 4.497 4.514 qs_ot_get_p 119 10.4 0.001 0.001 3.839 3.989 init_scf_run 11 5.9 0.000 0.001 3.968 3.968 scf_env_initial_rho_setup 11 6.9 0.028 0.030 3.968 3.968 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.911 3.441 pw_transfer 1439 11.6 0.068 0.069 3.282 3.300 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.183 3.203 make_m2s 4572 13.5 0.038 0.038 2.823 3.069 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.725 3.055 dbcsr_complete_redistribute 329 12.2 0.284 0.290 2.138 2.968 make_images 4572 14.5 0.354 0.384 2.703 2.948 fft3d_ps 1201 14.6 0.598 0.609 2.807 2.823 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.395 2.766 apply_single 119 13.6 0.000 0.000 2.395 2.766 fft_wrap_pw1pw2_140 487 13.2 0.095 0.099 2.601 2.622 density_rs2pw 119 9.7 0.004 0.004 2.550 2.582 mp_sum_l 7207 12.9 1.547 2.570 1.547 2.570 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.745 2.567 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.352 2.354 calculate_dm_sparse 119 9.5 0.000 0.000 2.313 2.342 calculate_first_density_matrix 1 7.0 0.000 0.000 2.332 2.334 mp_alltoall_i22 627 13.8 1.477 2.323 1.477 2.323 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.474 2.291 ot_diis_step 108 11.5 0.014 0.014 2.135 2.136 grid_integrate_task_list 119 12.3 2.089 2.115 2.089 2.115 jit_kernel_multiply 10 15.5 0.990 2.075 0.990 2.075 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.043 2.044 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.930 1.985 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.970 1.972 qs_energies_init_hamiltonians 11 5.9 0.009 0.010 1.857 1.858 potential_pw2rs 119 12.3 0.014 0.015 1.828 1.847 rs_pw_transfer 974 11.9 0.009 0.009 1.795 1.823 cp_fm_cholesky_invert 11 10.9 1.775 1.779 1.775 1.779 make_images_data 4572 15.5 0.043 0.046 1.386 1.739 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.366 1.717 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.654 1.697 cp_fm_diag_elpa 50 13.0 0.001 0.002 1.680 1.680 cp_fm_diag_elpa_base 50 14.0 1.526 1.586 1.675 1.675 grid_collocate_task_list 119 9.7 1.512 1.539 1.512 1.539 wfi_extrapolate 11 7.9 0.001 0.001 1.526 1.526 mp_alltoall_d11v 2130 13.8 1.326 1.367 1.326 1.367 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.325 1.335 mp_alltoall_z22v 1201 16.6 1.256 1.293 1.256 1.293 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.232 1.262 yz_to_x 606 15.1 0.459 0.470 1.220 1.239 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.186 1.201 multiply_cannon_sync_h2d 9144 15.5 1.045 1.051 1.045 1.051 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.984 1.021 qs_create_task_list 11 7.9 0.049 0.053 1.008 1.019 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.978 0.994 x_to_yz 595 16.2 0.481 0.490 0.975 0.980 generate_qs_task_list 11 8.9 0.370 0.390 0.959 0.973 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=46.071000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=704.000000, yerr=16.365151 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.829632E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.037 91.733 91.734 qs_mol_dyn_low 1 2.0 0.003 0.003 91.334 91.344 qs_forces 11 3.9 0.002 0.003 91.152 91.155 qs_energies 11 4.9 0.001 0.003 88.095 88.113 scf_env_do_scf 11 5.9 0.000 0.001 78.045 78.048 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 71.605 71.605 dbcsr_multiply_generic 2055 12.4 0.109 0.113 55.430 55.842 qs_scf_new_mos 99 7.5 0.000 0.001 52.286 52.412 qs_scf_loop_do_ot 99 8.5 0.001 0.001 52.285 52.411 ot_scf_mini 99 9.5 0.002 0.002 49.694 49.810 velocity_verlet 10 3.0 0.001 0.001 46.839 46.841 multiply_cannon 2055 13.4 0.179 0.185 43.772 44.920 multiply_cannon_loop 2055 14.4 1.531 1.570 42.568 43.731 ot_mini 99 10.5 0.001 0.001 30.152 30.263 qs_ot_get_derivative 99 11.5 0.001 0.001 23.037 23.143 multiply_cannon_multrec 49320 15.4 12.023 12.872 17.080 17.829 rebuild_ks_matrix 110 8.3 0.000 0.001 15.858 16.072 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 15.857 16.072 mp_waitall_1 220248 16.4 13.366 14.699 13.366 14.699 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.945 14.138 qs_ot_get_p 110 10.4 0.001 0.001 10.580 10.699 multiply_cannon_sync_h2d 49320 15.4 9.774 10.554 9.774 10.554 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.998 8.587 multiply_cannon_metrocomm3 49320 15.4 0.078 0.082 7.162 8.090 sum_up_and_integrate 110 10.3 0.036 0.043 7.971 7.984 integrate_v_rspace 110 11.3 0.003 0.003 7.934 7.956 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.287 7.818 apply_single 110 13.6 0.000 0.000 7.287 7.818 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.671 7.817 calculate_rho_elec 110 8.6 0.021 0.026 7.670 7.816 init_scf_run 11 5.9 0.000 0.001 7.814 7.814 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.813 7.814 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 7.541 7.631 mp_sum_l 6514 12.8 6.149 7.285 6.149 7.285 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.893 6.918 ot_diis_step 99 11.5 0.006 0.006 6.778 6.782 init_scf_loop 11 6.9 0.000 0.000 6.401 6.401 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.723 5.724 multiply_cannon_metrocomm1 49320 15.4 0.060 0.063 3.777 5.382 dbcsr_mm_accdrv_process 87628 16.1 2.032 2.169 4.936 5.276 make_m2s 4110 13.4 0.061 0.065 5.043 5.158 make_images 4110 14.4 0.177 0.189 4.946 5.065 cp_fm_diag_elpa 48 13.0 0.001 0.002 5.051 5.059 cp_fm_redistribute_end 48 14.0 4.397 5.004 4.403 5.010 rs_pw_transfer 902 11.9 0.011 0.012 4.422 4.926 cp_fm_diag_elpa_base 48 14.0 0.594 4.768 0.598 4.799 density_rs2pw 110 9.6 0.004 0.005 4.300 4.759 wfi_extrapolate 11 7.9 0.001 0.001 4.669 4.669 pw_transfer 1331 11.6 0.054 0.065 4.215 4.297 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.127 4.213 calculate_dm_sparse 110 9.5 0.000 0.001 3.982 4.099 prepare_preconditioner 11 7.9 0.000 0.000 4.023 4.044 make_preconditioner 11 8.9 0.000 0.000 4.023 4.044 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.973 3.990 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.774 3.828 fft3d_ps 1111 14.6 0.784 0.882 3.656 3.731 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.617 3.682 fft_wrap_pw1pw2_140 451 13.1 0.170 0.191 3.365 3.452 grid_integrate_task_list 110 12.3 3.237 3.373 3.237 3.373 mp_alltoall_d11v 2046 13.8 2.758 3.353 2.758 3.353 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.291 3.353 potential_pw2rs 110 12.3 0.006 0.007 3.268 3.303 calculate_first_density_matrix 1 7.0 0.000 0.000 3.032 3.040 jit_kernel_multiply 13 15.9 2.621 2.708 2.621 2.708 mp_waitany 14300 13.8 2.068 2.637 2.068 2.637 mp_alltoall_z22v 1111 16.6 2.133 2.410 2.133 2.410 make_images_data 4110 15.4 0.042 0.045 2.135 2.390 mp_sum_d 3883 11.9 1.858 2.365 1.858 2.365 grid_collocate_task_list 110 9.6 2.083 2.308 2.083 2.308 cp_fm_cholesky_invert 11 10.9 2.278 2.282 2.278 2.282 acc_transpose_blocks 49320 15.4 0.221 0.235 2.177 2.277 hybrid_alltoall_any 4261 16.3 0.080 0.468 1.880 2.147 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.071 2.097 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=91.734000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.909091, yerr=2.193152 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_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 588.587008E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 3671289. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.074 71.096 71.114 qs_mol_dyn_low 1 2.0 0.003 0.004 70.527 70.537 qs_forces 11 3.9 0.012 0.021 70.448 70.454 qs_energies 11 4.9 0.002 0.002 67.049 67.052 scf_env_do_scf 11 5.9 0.000 0.001 57.426 57.429 scf_env_do_scf_inner_loop 99 6.5 0.008 0.013 49.610 49.610 dbcsr_multiply_generic 2055 12.4 0.115 0.120 38.929 39.140 velocity_verlet 10 3.0 0.001 0.001 36.761 36.762 multiply_cannon 2055 13.4 0.222 0.244 32.111 33.497 qs_scf_new_mos 99 7.5 0.001 0.001 32.984 33.177 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.984 33.176 ot_scf_mini 99 9.5 0.003 0.003 31.304 31.471 multiply_cannon_loop 2055 14.4 0.932 0.957 30.377 31.381 ot_mini 99 10.5 0.001 0.001 18.341 18.526 multiply_cannon_multrec 24660 15.4 7.624 8.948 14.104 15.312 rebuild_ks_matrix 110 8.3 0.000 0.001 13.963 14.119 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.963 14.119 qs_ot_get_derivative 99 11.5 0.001 0.001 12.538 12.714 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.280 12.409 mp_waitall_1 176588 16.5 7.799 10.372 7.799 10.372 multiply_cannon_sync_h2d 24660 15.4 7.017 8.304 7.017 8.304 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 5.215 7.868 init_scf_loop 11 6.9 0.000 0.000 7.780 7.780 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.513 7.258 apply_single 110 13.6 0.000 0.001 6.513 7.258 init_scf_run 11 5.9 0.000 0.001 7.225 7.225 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.225 7.225 sum_up_and_integrate 110 10.3 0.052 0.059 6.729 6.739 integrate_v_rspace 110 11.3 0.003 0.005 6.677 6.690 dbcsr_mm_accdrv_process 52282 16.1 4.920 5.939 6.321 6.552 qs_ot_get_p 110 10.4 0.001 0.001 6.249 6.446 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.350 6.361 calculate_rho_elec 110 8.6 0.040 0.047 6.349 6.361 ot_diis_step 99 11.5 0.010 0.010 5.752 5.752 prepare_preconditioner 11 7.9 0.000 0.000 5.726 5.748 make_preconditioner 11 8.9 0.000 0.000 5.726 5.748 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.314 5.478 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.687 5.419 make_m2s 4110 13.4 0.057 0.059 4.282 4.782 make_images 4110 14.4 0.397 0.443 4.174 4.670 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.296 4.316 pw_transfer 1331 11.6 0.066 0.073 3.858 4.002 density_rs2pw 110 9.6 0.004 0.004 3.486 3.994 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.751 3.898 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.850 3.851 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.684 3.686 rs_pw_transfer 902 11.9 0.012 0.014 3.110 3.660 calculate_dm_sparse 110 9.5 0.001 0.001 3.582 3.621 wfi_extrapolate 11 7.9 0.001 0.001 3.585 3.585 calculate_first_density_matrix 1 7.0 0.000 0.000 3.545 3.548 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.317 3.406 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.265 3.360 grid_integrate_task_list 110 12.3 3.149 3.331 3.149 3.331 fft_wrap_pw1pw2_140 451 13.1 0.202 0.219 3.160 3.308 cp_fm_diag_elpa 48 13.0 0.001 0.001 3.276 3.277 cp_fm_redistribute_end 48 14.0 2.448 3.246 2.450 3.247 fft3d_ps 1111 14.6 1.104 1.320 3.103 3.237 cp_fm_diag_elpa_base 48 14.0 0.763 3.097 0.794 3.183 make_images_data 4110 15.4 0.046 0.050 2.382 2.882 hybrid_alltoall_any 4261 16.3 0.101 0.447 2.058 2.745 mp_sum_l 6514 12.8 1.915 2.691 1.915 2.691 cp_fm_cholesky_invert 11 10.9 2.655 2.662 2.655 2.662 potential_pw2rs 110 12.3 0.008 0.008 2.538 2.550 grid_collocate_task_list 110 9.6 2.054 2.472 2.054 2.472 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.136 2.147 mp_alltoall_d11v 2046 13.8 1.713 2.060 1.713 2.060 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.013 2.027 jit_kernel_multiply 9 16.4 1.050 2.009 1.050 2.009 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.848 1.871 acc_transpose_blocks 24660 15.4 0.111 0.113 1.846 1.871 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.866 1.867 mp_waitany 10164 13.8 1.231 1.789 1.231 1.789 mp_allgather_i34 2055 14.4 0.713 1.714 0.713 1.714 multiply_cannon_metrocomm4 22605 15.4 0.072 0.076 0.778 1.656 mp_irecv_dv 57340 16.2 0.656 1.512 0.656 1.512 rs_pw_transfer_RS2PW_140 121 11.5 0.206 0.214 0.962 1.508 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.368 1.482 dbcsr_complete_redistribute 325 12.2 0.247 0.302 1.185 1.452 mp_alltoall_z22v 1111 16.6 1.315 1.433 1.315 1.433 cp_fm_cholesky_decompose 22 10.9 1.424 1.430 1.424 1.430 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.114000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.909091, yerr=7.427890 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_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 659.623936E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.418 0.480 63.646 63.648 qs_mol_dyn_low 1 2.0 0.003 0.005 62.613 62.622 qs_forces 11 3.9 0.030 0.044 62.545 62.547 qs_energies 11 4.9 0.003 0.011 59.277 59.309 scf_env_do_scf 11 5.9 0.001 0.001 50.280 50.280 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.345 41.346 velocity_verlet 10 3.0 0.001 0.001 33.395 33.397 dbcsr_multiply_generic 2055 12.4 0.138 0.144 29.660 29.928 qs_scf_new_mos 99 7.5 0.001 0.001 25.960 26.046 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.960 26.046 ot_scf_mini 99 9.5 0.003 0.003 24.725 24.819 multiply_cannon 2055 13.4 0.213 0.227 23.014 24.211 multiply_cannon_loop 2055 14.4 0.616 0.629 21.776 23.113 ot_mini 99 10.5 0.001 0.001 14.245 14.343 rebuild_ks_matrix 110 8.3 0.000 0.000 12.608 12.740 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.607 12.739 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.120 11.237 multiply_cannon_multrec 16440 15.4 3.960 4.921 10.121 11.015 mp_waitall_1 139946 16.5 7.131 10.397 7.131 10.397 qs_ot_get_derivative 99 11.5 0.001 0.001 9.771 9.863 init_scf_loop 11 6.9 0.002 0.009 8.899 8.900 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.296 7.422 prepare_preconditioner 11 7.9 0.000 0.001 7.081 7.097 make_preconditioner 11 8.9 0.001 0.004 7.081 7.097 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.405 6.759 sum_up_and_integrate 110 10.3 0.060 0.061 6.693 6.708 integrate_v_rspace 110 11.3 0.003 0.003 6.633 6.649 init_scf_run 11 5.9 0.001 0.003 6.393 6.394 scf_env_initial_rho_setup 11 6.9 0.008 0.028 6.392 6.394 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.157 6.167 calculate_rho_elec 110 8.6 0.058 0.059 6.156 6.167 dbcsr_mm_accdrv_process 34862 16.1 4.706 5.600 6.016 6.142 qs_ot_get_p 110 10.4 0.001 0.001 5.623 5.769 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.988 5.408 apply_single 110 13.6 0.000 0.000 4.988 5.408 make_m2s 4110 13.4 0.050 0.051 4.331 4.721 make_images 4110 14.4 0.394 0.514 4.217 4.606 ot_diis_step 99 11.5 0.010 0.011 4.437 4.438 density_rs2pw 110 9.6 0.004 0.004 3.267 4.436 multiply_cannon_sync_h2d 16440 15.4 3.722 4.296 3.722 4.296 rs_pw_transfer 902 11.9 0.010 0.011 2.943 4.038 qs_ot_p2m_diag 48 11.0 0.041 0.044 3.942 3.948 pw_transfer 1331 11.6 0.065 0.072 3.798 3.811 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.169 3.780 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.691 3.706 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.588 3.589 grid_integrate_task_list 110 12.3 3.184 3.439 3.184 3.439 calculate_first_density_matrix 1 7.0 0.008 0.034 3.311 3.318 fft_wrap_pw1pw2_140 451 13.1 0.213 0.217 3.168 3.187 cp_fm_diag_elpa 48 13.0 0.001 0.001 3.045 3.047 cp_fm_redistribute_end 48 14.0 1.903 3.018 1.906 3.018 fft3d_ps 1111 14.6 1.093 1.105 2.991 3.005 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.998 3.000 wfi_extrapolate 11 7.9 0.001 0.001 2.984 2.985 make_images_data 4110 15.4 0.043 0.047 2.496 2.980 cp_fm_diag_elpa_base 48 14.0 1.049 2.855 1.108 2.961 calculate_dm_sparse 110 9.5 0.001 0.001 2.895 2.926 hybrid_alltoall_any 4261 16.3 0.105 0.371 2.212 2.887 cp_fm_cholesky_invert 11 10.9 2.868 2.874 2.868 2.874 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.584 2.632 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.540 2.600 grid_collocate_task_list 110 9.6 2.080 2.478 2.080 2.478 mp_waitany 17072 13.8 1.307 2.457 1.307 2.457 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.862 2.441 potential_pw2rs 110 12.3 0.010 0.010 2.422 2.433 mp_irecv_dv 48980 15.7 0.793 2.316 0.793 2.316 mp_sum_l 6514 12.8 1.683 2.273 1.683 2.273 qs_energies_init_hamiltonians 11 5.9 0.007 0.016 2.109 2.138 mp_alltoall_d11v 2046 13.8 1.773 2.099 1.773 2.099 rs_pw_transfer_RS2PW_140 121 11.5 0.182 0.199 0.974 2.055 dbcsr_complete_redistribute 325 12.2 0.370 0.453 1.533 1.996 jit_kernel_multiply 8 16.5 0.920 1.920 0.920 1.920 cp_fm_upper_to_full 70 13.6 1.393 1.862 1.393 1.862 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.845 1.855 cp_fm_cholesky_decompose 22 10.9 1.711 1.736 1.711 1.736 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.688 1.702 mp_allgather_i34 2055 14.4 0.563 1.647 0.563 1.647 acc_transpose_blocks 16440 15.4 0.073 0.075 1.602 1.630 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.074 1.526 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.367 1.511 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.312 1.322 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.648000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.000000, yerr=9.390517 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_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 720.617472E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.061 66.290 66.298 qs_mol_dyn_low 1 2.0 0.003 0.003 65.887 65.898 qs_forces 11 3.9 0.002 0.002 65.811 65.813 qs_energies 11 4.9 0.011 0.019 62.386 62.389 scf_env_do_scf 11 5.9 0.000 0.001 53.759 53.762 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.844 41.845 velocity_verlet 10 3.0 0.001 0.001 37.265 37.267 dbcsr_multiply_generic 2055 12.4 0.115 0.117 29.867 30.051 qs_scf_new_mos 99 7.5 0.001 0.001 26.576 26.676 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.575 26.675 ot_scf_mini 99 9.5 0.003 0.003 24.958 25.063 multiply_cannon 2055 13.4 0.242 0.257 22.871 23.833 multiply_cannon_loop 2055 14.4 0.881 0.907 21.426 21.927 ot_mini 99 10.5 0.001 0.001 14.415 14.541 multiply_cannon_multrec 24660 15.4 4.319 7.073 12.797 13.933 rebuild_ks_matrix 110 8.3 0.000 0.001 12.332 12.424 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.332 12.423 init_scf_loop 11 6.9 0.000 0.000 11.870 11.870 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.906 10.986 qs_ot_get_derivative 99 11.5 0.001 0.001 10.277 10.386 prepare_preconditioner 11 7.9 0.000 0.000 10.129 10.145 make_preconditioner 11 8.9 0.000 0.000 10.129 10.145 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.354 9.817 dbcsr_mm_accdrv_process 52304 16.0 6.953 8.685 8.332 9.192 mp_waitall_1 121746 16.5 4.642 6.679 4.642 6.679 sum_up_and_integrate 110 10.3 0.068 0.071 6.601 6.613 integrate_v_rspace 110 11.3 0.003 0.003 6.533 6.545 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.213 6.222 calculate_rho_elec 110 8.6 0.077 0.081 6.213 6.222 init_scf_run 11 5.9 0.000 0.001 5.881 5.882 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.881 5.882 make_m2s 4110 13.4 0.060 0.063 5.400 5.753 qs_ot_get_p 110 10.4 0.001 0.001 5.521 5.672 make_images 4110 14.4 0.578 0.696 5.260 5.610 cp_fm_upper_to_full 70 13.8 3.269 4.657 3.269 4.657 ot_diis_step 99 11.5 0.011 0.012 4.103 4.103 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.012 4.085 apply_single 110 13.6 0.000 0.000 4.012 4.084 pw_transfer 1331 11.6 0.065 0.075 3.901 3.934 dbcsr_complete_redistribute 325 12.2 0.516 0.586 2.780 3.920 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.796 3.831 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.755 3.770 density_rs2pw 110 9.6 0.004 0.004 3.134 3.647 grid_integrate_task_list 110 12.3 3.237 3.428 3.237 3.428 multiply_cannon_sync_h2d 24660 15.4 3.182 3.414 3.182 3.414 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.254 3.375 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.281 3.334 fft_wrap_pw1pw2_140 451 13.1 0.202 0.212 3.250 3.287 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.270 3.270 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.573 3.249 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.064 3.122 hybrid_alltoall_any 4261 16.3 0.119 0.458 2.317 3.119 fft3d_ps 1111 14.6 1.093 1.132 3.089 3.110 make_images_data 4110 15.4 0.046 0.050 2.738 3.096 calculate_dm_sparse 110 9.5 0.001 0.001 3.043 3.078 wfi_extrapolate 11 7.9 0.001 0.001 2.992 2.992 rs_pw_transfer 902 11.9 0.010 0.011 2.452 2.976 cp_fm_cholesky_invert 11 10.9 2.917 2.925 2.917 2.925 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.899 2.900 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.767 2.875 mp_alltoall_i22 605 13.7 1.644 2.807 1.644 2.807 calculate_first_density_matrix 1 7.0 0.000 0.000 2.795 2.798 cp_fm_diag_elpa 48 13.0 0.001 0.002 2.708 2.711 cp_fm_redistribute_end 48 14.0 1.356 2.683 1.358 2.684 cp_fm_diag_elpa_base 48 14.0 1.245 2.543 1.323 2.655 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.426 2.455 grid_collocate_task_list 110 9.6 2.181 2.393 2.181 2.393 potential_pw2rs 110 12.3 0.012 0.013 2.264 2.272 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.244 2.246 mp_alltoall_d11v 2046 13.8 1.787 2.026 1.787 2.026 jit_kernel_multiply 11 15.8 1.048 2.018 1.048 2.018 cp_fm_cholesky_decompose 22 10.9 1.704 1.766 1.704 1.766 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.712 1.744 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.597 1.697 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.672 1.684 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.659 1.678 mp_sum_l 6514 12.8 0.923 1.548 0.923 1.548 acc_transpose_blocks 24660 15.4 0.105 0.108 1.513 1.545 mp_waitany 13376 13.8 1.063 1.527 1.063 1.527 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.823 1.496 mp_allgather_i34 2055 14.4 0.544 1.421 0.544 1.421 mp_irecv_dv 62702 16.1 0.725 1.418 0.725 1.418 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.298000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=684.454545, yerr=5.867687 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_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 838.381568E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.439 0.475 57.696 57.704 qs_mol_dyn_low 1 2.0 0.003 0.003 56.917 56.929 qs_forces 11 3.9 0.002 0.003 56.837 56.847 qs_energies 11 4.9 0.001 0.002 53.091 53.103 scf_env_do_scf 11 5.9 0.000 0.001 44.495 44.495 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 36.451 36.451 velocity_verlet 10 3.0 0.001 0.001 31.950 31.953 dbcsr_multiply_generic 2055 12.4 0.106 0.108 23.118 23.233 qs_scf_new_mos 99 7.5 0.001 0.001 20.992 21.041 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.991 21.040 ot_scf_mini 99 9.5 0.002 0.002 19.740 19.777 multiply_cannon 2055 13.4 0.248 0.269 17.530 18.807 multiply_cannon_loop 2055 14.4 0.321 0.332 16.155 16.444 rebuild_ks_matrix 110 8.3 0.000 0.000 12.206 12.242 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.206 12.241 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.825 10.853 ot_mini 99 10.5 0.001 0.001 10.613 10.647 multiply_cannon_multrec 8220 15.4 3.230 4.572 7.530 8.501 init_scf_loop 11 6.9 0.000 0.000 7.994 7.996 mp_waitall_1 103326 16.6 6.199 7.953 6.199 7.953 qs_ot_get_derivative 99 11.5 0.001 0.001 6.756 6.796 sum_up_and_integrate 110 10.3 0.079 0.082 6.699 6.710 integrate_v_rspace 110 11.3 0.003 0.003 6.619 6.631 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.496 6.511 calculate_rho_elec 110 8.6 0.116 0.116 6.495 6.510 prepare_preconditioner 11 7.9 0.000 0.000 6.298 6.303 make_preconditioner 11 8.9 0.000 0.000 6.298 6.303 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.884 5.961 init_scf_run 11 5.9 0.000 0.001 5.455 5.455 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.455 5.455 qs_ot_get_p 110 10.4 0.001 0.001 5.154 5.190 dbcsr_mm_accdrv_process 17442 15.9 2.956 4.028 4.169 5.066 make_m2s 4110 13.4 0.038 0.040 4.265 4.516 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.116 4.410 make_images 4110 14.4 0.640 0.698 4.135 4.386 pw_transfer 1331 11.6 0.066 0.071 4.256 4.266 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.148 4.163 ot_diis_step 99 11.5 0.012 0.012 3.832 3.832 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.744 3.795 apply_single 110 13.6 0.000 0.000 3.744 3.795 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.769 3.772 density_rs2pw 110 9.6 0.004 0.004 3.236 3.611 fft_wrap_pw1pw2_140 451 13.1 0.216 0.220 3.558 3.579 grid_integrate_task_list 110 12.3 3.358 3.534 3.358 3.534 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.457 3.458 fft3d_ps 1111 14.6 1.147 1.157 3.391 3.402 cp_fm_cholesky_invert 11 10.9 3.126 3.131 3.126 3.131 multiply_cannon_sync_h2d 8220 15.4 2.916 3.064 2.916 3.064 cp_fm_diag_elpa 48 13.0 0.008 0.013 2.908 2.909 cp_fm_redistribute_end 48 14.0 0.738 2.858 0.743 2.859 wfi_extrapolate 11 7.9 0.001 0.001 2.841 2.841 make_images_data 4110 15.4 0.038 0.044 2.404 2.828 cp_fm_diag_elpa_base 48 14.0 1.930 2.652 2.107 2.815 hybrid_alltoall_any 4261 16.3 0.200 0.864 2.339 2.761 rs_pw_transfer 902 11.9 0.010 0.010 2.302 2.697 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.686 2.696 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.561 2.563 calculate_dm_sparse 110 9.5 0.001 0.001 2.513 2.540 grid_collocate_task_list 110 9.6 2.277 2.540 2.277 2.540 calculate_first_density_matrix 1 7.0 0.000 0.000 2.514 2.515 potential_pw2rs 110 12.3 0.015 0.015 2.229 2.232 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.180 2.197 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.008 2.024 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.774 1.983 mp_alltoall_d11v 2046 13.8 1.683 1.864 1.683 1.864 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.835 1.859 jit_kernel_multiply 10 15.5 0.903 1.796 0.903 1.796 cp_fm_cholesky_decompose 22 10.9 1.729 1.755 1.729 1.755 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.719 1.727 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.521 1.644 mp_allgather_i34 2055 14.4 0.559 1.632 0.559 1.632 dbcsr_complete_redistribute 325 12.2 0.555 0.574 1.483 1.592 mp_waitany 9240 13.8 1.105 1.520 1.105 1.520 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.463 1.474 qs_create_task_list 11 7.9 0.001 0.002 1.224 1.331 generate_qs_task_list 11 8.9 0.379 0.446 1.224 1.330 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.197 1.234 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.169 0.805 1.210 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.474 1.183 mp_alltoall_z22v 1111 16.6 1.135 1.165 1.135 1.165 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.704000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=793.090909, yerr=12.522377 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_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 1.469788E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.249 0.295 91.290 91.300 qs_mol_dyn_low 1 2.0 0.003 0.003 90.681 90.692 qs_forces 11 3.9 0.007 0.012 90.606 90.613 qs_energies 11 4.9 0.002 0.002 86.422 86.435 scf_env_do_scf 11 5.9 0.001 0.001 75.739 75.739 velocity_verlet 10 3.0 0.001 0.001 56.679 56.685 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 47.082 47.084 dbcsr_multiply_generic 2055 12.4 0.131 0.139 30.753 30.821 init_scf_loop 11 6.9 0.000 0.000 28.583 28.587 qs_scf_new_mos 99 7.5 0.001 0.001 28.215 28.250 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.215 28.250 prepare_preconditioner 11 7.9 0.000 0.000 26.514 26.521 make_preconditioner 11 8.9 0.000 0.000 26.514 26.520 ot_scf_mini 99 9.5 0.002 0.002 26.426 26.468 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.857 25.987 multiply_cannon 2055 13.4 0.346 0.369 22.449 23.696 multiply_cannon_loop 2055 14.4 0.341 0.343 20.616 21.987 cp_fm_upper_to_full 70 14.2 12.646 18.021 12.646 18.021 ot_mini 99 10.5 0.001 0.001 15.234 15.274 rebuild_ks_matrix 110 8.3 0.001 0.001 14.519 14.550 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.518 14.550 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.173 13.204 multiply_cannon_multrec 8220 15.4 4.365 4.531 10.075 11.023 qs_ot_get_derivative 99 11.5 0.001 0.001 10.639 10.680 dbcsr_complete_redistribute 325 12.2 1.025 1.071 7.284 10.394 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.232 9.343 mp_waitall_1 84994 16.7 7.977 8.790 7.977 8.790 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.641 8.731 mp_alltoall_i22 605 13.7 5.288 8.462 5.288 8.462 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.988 8.022 calculate_rho_elec 110 8.6 0.227 0.227 7.988 8.022 sum_up_and_integrate 110 10.3 0.150 0.151 7.628 7.640 integrate_v_rspace 110 11.3 0.004 0.004 7.478 7.489 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.724 6.547 apply_single 110 13.6 0.000 0.000 5.724 6.546 dbcsr_mm_accdrv_process 11614 15.7 3.338 4.367 5.566 6.373 init_scf_run 11 5.9 0.000 0.001 6.357 6.357 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.357 6.357 qs_ot_get_p 110 10.4 0.001 0.001 5.965 6.002 make_m2s 4110 13.4 0.043 0.044 5.438 5.979 make_images 4110 14.4 0.884 0.934 5.250 5.791 cp_fm_cholesky_invert 11 10.9 5.565 5.571 5.565 5.571 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 5.078 5.530 pw_transfer 1331 11.6 0.075 0.075 5.429 5.436 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.312 5.320 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.296 4.776 fft_wrap_pw1pw2_140 451 13.1 0.222 0.226 4.588 4.597 ot_diis_step 99 11.5 0.015 0.016 4.570 4.570 fft3d_ps 1111 14.6 1.294 1.302 4.505 4.511 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.356 4.363 density_rs2pw 110 9.6 0.004 0.004 3.956 4.001 multiply_cannon_sync_h2d 8220 15.4 3.956 3.961 3.956 3.961 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.902 3.903 grid_integrate_task_list 110 12.3 3.712 3.840 3.712 3.840 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.718 3.729 hybrid_alltoall_any 4261 16.3 0.258 0.559 2.994 3.693 make_images_data 4110 15.4 0.042 0.045 2.952 3.648 calculate_dm_sparse 110 9.5 0.001 0.001 3.539 3.569 wfi_extrapolate 11 7.9 0.001 0.001 3.390 3.390 cp_fm_diag_elpa 48 13.0 0.001 0.002 3.276 3.276 cp_fm_diag_elpa_base 48 14.0 2.742 2.936 3.271 3.272 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.089 3.090 calculate_first_density_matrix 1 7.0 0.000 0.000 2.842 2.843 grid_collocate_task_list 110 9.6 2.673 2.732 2.673 2.732 mp_sum_l 6514 12.8 1.982 2.651 1.982 2.651 potential_pw2rs 110 12.3 0.021 0.022 2.620 2.626 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.540 2.554 rs_pw_transfer 902 11.9 0.011 0.011 2.438 2.523 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.475 2.493 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.254 2.316 mp_alltoall_d11v 2046 13.8 2.114 2.226 2.114 2.226 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.082 2.153 jit_kernel_multiply 10 15.4 2.027 2.140 2.027 2.140 cp_fm_cholesky_decompose 22 10.9 2.109 2.125 2.109 2.125 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.052 2.069 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.956 1.959 qs_create_task_list 11 7.9 0.001 0.001 1.893 1.941 generate_qs_task_list 11 8.9 0.735 0.787 1.893 1.940 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.842 1.892 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=91.300000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1274.272727, yerr=89.546424 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 629.395456E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56513. MP_Allreduce 11298 784. MP_Sync 170 MP_Alltoall 2226 2520936. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.058 207.567 207.568 qs_mol_dyn_low 1 2.0 0.003 0.004 207.057 207.070 qs_forces 11 3.9 0.003 0.005 206.958 206.962 qs_energies 11 4.9 0.001 0.002 201.308 201.335 scf_env_do_scf 11 5.9 0.001 0.001 184.119 184.123 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 163.131 163.133 dbcsr_multiply_generic 2507 12.6 0.179 0.183 125.506 126.483 velocity_verlet 10 3.0 0.001 0.001 124.897 124.898 qs_scf_new_mos 117 7.6 0.001 0.001 122.453 122.787 qs_scf_loop_do_ot 117 8.6 0.001 0.001 122.452 122.786 ot_scf_mini 117 9.6 0.003 0.003 115.852 116.150 multiply_cannon 2507 13.6 0.235 0.243 101.913 104.641 multiply_cannon_loop 2507 14.6 2.095 2.175 99.717 102.126 ot_mini 117 10.6 0.001 0.001 66.867 67.240 multiply_cannon_multrec 60168 15.6 33.184 35.380 42.068 43.657 qs_ot_get_derivative 117 11.6 0.001 0.001 41.901 42.222 rebuild_ks_matrix 128 8.3 0.001 0.001 34.151 34.453 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.027 34.150 34.453 mp_waitall_1 267128 16.5 29.087 31.747 29.087 31.747 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.690 30.929 multiply_cannon_sync_h2d 60168 15.6 27.190 29.256 27.190 29.256 qs_ot_get_p 128 10.4 0.001 0.002 26.956 27.280 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.315 25.102 apply_single 128 13.6 0.001 0.001 24.315 25.102 ot_diis_step 117 11.6 0.008 0.008 24.575 24.576 init_scf_loop 11 6.9 0.000 0.000 20.912 20.913 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.247 20.339 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.134 19.355 multiply_cannon_metrocomm3 60168 15.6 0.114 0.119 15.948 18.578 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.504 17.506 prepare_preconditioner 11 7.9 0.000 0.000 16.247 16.300 make_preconditioner 11 8.9 0.000 0.000 16.247 16.300 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.425 15.593 cp_fm_diag_elpa 83 13.4 0.000 0.002 14.658 14.662 cp_fm_redistribute_end 83 14.4 11.588 14.585 11.599 14.587 sum_up_and_integrate 128 10.3 0.089 0.107 14.524 14.545 integrate_v_rspace 128 11.3 0.003 0.004 14.435 14.458 cp_fm_diag_elpa_base 83 14.4 2.946 14.216 2.977 14.333 make_m2s 5014 13.6 0.104 0.111 13.835 14.137 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.983 14.122 calculate_rho_elec 128 8.7 0.045 0.064 13.982 14.122 make_images 5014 14.6 0.399 0.419 13.654 13.964 init_scf_run 11 5.9 0.000 0.001 12.931 12.931 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.931 12.931 density_rs2pw 128 9.7 0.006 0.007 7.324 10.561 mp_sum_l 7870 13.0 8.493 9.977 8.493 9.977 cp_fm_cholesky_invert 11 10.9 9.257 9.265 9.257 9.265 wfi_extrapolate 11 7.9 0.001 0.001 9.230 9.231 dbcsr_mm_accdrv_process 124484 16.2 3.227 3.414 8.450 9.172 rs_pw_transfer 1046 11.9 0.016 0.019 5.873 9.126 calculate_dm_sparse 128 9.5 0.001 0.001 8.721 8.902 pw_transfer 1547 11.6 0.076 0.093 8.242 8.494 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 8.037 8.299 multiply_cannon_metrocomm1 60168 15.6 0.086 0.092 6.563 8.247 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.005 8.174 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.893 8.044 make_images_data 5014 15.6 0.066 0.071 6.786 7.732 grid_integrate_task_list 128 12.3 7.021 7.532 7.021 7.532 fft_wrap_pw1pw2_140 523 13.2 0.444 0.511 6.919 7.224 hybrid_alltoall_any 5200 16.5 0.288 2.240 5.964 7.181 fft3d_ps 1291 14.7 2.106 2.878 6.824 6.991 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.778 6.787 mp_waitany 16020 13.9 2.730 5.985 2.730 5.985 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.868 5.944 grid_collocate_task_list 128 9.7 4.563 5.812 4.563 5.812 mp_alltoall_d11v 2415 14.1 4.378 5.594 4.378 5.594 rs_pw_transfer_RS2PW_140 139 11.5 0.280 0.295 2.139 5.389 potential_pw2rs 128 12.3 0.009 0.011 4.880 4.907 cp_fm_cholesky_decompose 22 10.9 4.731 4.746 4.731 4.746 mp_sum_d 4473 12.2 3.732 4.537 3.732 4.537 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.568000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.818182, yerr=7.132744 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 832.155648E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57341. MP_Allreduce 11227 947. MP_Sync 170 MP_Alltoall 1969 4718641. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.407 0.491 196.186 196.190 qs_mol_dyn_low 1 2.0 0.004 0.007 195.112 195.125 qs_forces 11 3.9 0.009 0.011 193.473 193.475 qs_energies 11 4.9 0.002 0.005 186.391 186.401 scf_env_do_scf 11 5.9 0.005 0.055 168.622 168.632 scf_env_do_scf_inner_loop 117 6.6 0.005 0.025 135.060 135.064 velocity_verlet 10 3.0 0.001 0.001 122.597 122.631 dbcsr_multiply_generic 2507 12.6 0.187 0.193 98.508 99.748 qs_scf_new_mos 117 7.6 0.001 0.001 94.473 95.103 qs_scf_loop_do_ot 117 8.6 0.001 0.002 94.472 95.102 ot_scf_mini 117 9.6 0.004 0.005 89.676 90.364 multiply_cannon 2507 13.6 0.479 0.530 78.429 82.504 multiply_cannon_loop 2507 14.6 1.265 1.309 75.378 77.989 ot_mini 117 10.6 0.001 0.001 49.886 50.533 mp_waitall_1 214728 16.6 24.718 38.630 24.718 38.630 multiply_cannon_multrec 30084 15.6 22.036 26.642 32.081 37.229 rebuild_ks_matrix 128 8.3 0.001 0.001 33.606 34.158 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 33.605 34.158 init_scf_loop 11 6.9 0.020 0.160 33.429 33.430 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.243 30.762 prepare_preconditioner 11 7.9 0.000 0.001 28.837 28.898 make_preconditioner 11 8.9 0.000 0.002 28.837 28.898 qs_ot_get_derivative 117 11.6 0.001 0.002 28.002 28.683 multiply_cannon_metrocomm3 30084 15.6 0.095 0.101 15.807 28.638 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.518 28.072 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.953 23.058 apply_single 128 13.6 0.001 0.001 21.952 23.058 qs_ot_get_p 128 10.4 0.001 0.001 21.544 22.374 multiply_cannon_sync_h2d 30084 15.6 19.400 22.172 19.400 22.172 ot_diis_step 117 11.6 0.014 0.015 21.708 21.710 cp_fm_cholesky_invert 11 10.9 16.903 16.916 16.903 16.916 qs_ot_p2m_diag 83 11.4 0.187 0.215 16.712 16.751 make_m2s 5014 13.6 0.089 0.094 14.271 15.582 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.560 15.561 make_images 5014 14.6 1.154 1.337 14.063 15.373 sum_up_and_integrate 128 10.3 0.116 0.130 14.927 14.958 integrate_v_rspace 128 11.3 0.003 0.003 14.810 14.843 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.356 14.385 calculate_rho_elec 128 8.7 0.088 0.105 14.355 14.385 init_scf_run 11 5.9 0.000 0.001 12.865 12.867 scf_env_initial_rho_setup 11 6.9 0.002 0.004 12.865 12.867 cp_fm_diag_elpa 83 13.4 0.004 0.021 12.335 12.343 cp_fm_redistribute_end 83 14.4 7.195 12.253 7.207 12.254 cp_fm_diag_elpa_base 83 14.4 4.808 11.754 5.030 12.114 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.031 11.558 multiply_cannon_metrocomm4 27577 15.6 0.094 0.108 3.708 10.945 mp_irecv_dv 69486 16.3 3.517 10.551 3.517 10.551 make_images_data 5014 15.6 0.063 0.072 8.670 10.249 density_rs2pw 128 9.7 0.006 0.007 7.658 10.109 dbcsr_mm_accdrv_process 62242 16.2 4.790 5.834 9.508 10.083 hybrid_alltoall_any 5200 16.5 0.341 1.510 7.536 9.355 pw_transfer 1547 11.6 0.087 0.109 9.125 9.203 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 8.899 8.982 wfi_extrapolate 11 7.9 0.001 0.001 8.482 8.482 rs_pw_transfer 1046 11.9 0.014 0.016 5.919 8.389 fft_wrap_pw1pw2_140 523 13.2 0.472 0.529 7.810 7.917 grid_integrate_task_list 128 12.3 7.152 7.650 7.152 7.650 fft3d_ps 1291 14.7 2.748 2.922 7.268 7.311 cp_fm_cholesky_decompose 22 10.9 7.074 7.156 7.074 7.156 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.329 7.067 calculate_dm_sparse 128 9.5 0.001 0.001 6.827 6.958 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.922 6.928 mp_sum_l 7870 13.0 4.052 6.308 4.052 6.308 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.005 6.210 grid_collocate_task_list 128 9.7 4.711 5.808 4.711 5.808 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.371 5.431 potential_pw2rs 128 12.3 0.014 0.016 5.235 5.255 mp_alltoall_d11v 2415 14.1 4.346 5.144 4.346 5.144 mp_waitany 11748 13.9 2.506 5.031 2.506 5.031 mp_allgather_i34 2507 14.6 1.621 4.646 1.621 4.646 rs_pw_transfer_RS2PW_140 139 11.5 0.354 0.380 2.094 4.545 mp_sum_d 4459 12.1 2.908 4.277 2.908 4.277 calculate_first_density_matrix 1 7.0 0.000 0.001 4.202 4.206 dbcsr_complete_redistribute 395 12.7 0.773 0.883 3.282 4.111 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=196.190000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=791.727273, yerr=2.525899 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 945.483776E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.079 180.060 180.066 qs_mol_dyn_low 1 2.0 0.004 0.007 179.387 179.400 qs_forces 11 3.9 0.009 0.041 178.529 178.535 qs_energies 11 4.9 0.002 0.008 171.854 171.873 scf_env_do_scf 11 5.9 0.001 0.002 155.791 155.793 scf_env_do_scf_inner_loop 117 6.6 0.007 0.024 120.204 120.206 velocity_verlet 10 3.0 0.001 0.001 115.153 115.178 dbcsr_multiply_generic 2507 12.6 0.181 0.185 82.135 83.299 qs_scf_new_mos 117 7.6 0.001 0.001 81.601 81.896 qs_scf_loop_do_ot 117 8.6 0.001 0.002 81.600 81.895 ot_scf_mini 117 9.6 0.004 0.005 77.461 77.770 multiply_cannon 2507 13.6 0.500 0.530 62.167 66.172 multiply_cannon_loop 2507 14.6 0.855 0.882 59.064 61.677 ot_mini 117 10.6 0.001 0.001 43.098 43.414 init_scf_loop 11 6.9 0.001 0.004 35.485 35.496 mp_waitall_1 170520 16.6 25.062 34.276 25.062 34.276 rebuild_ks_matrix 128 8.3 0.001 0.001 31.005 31.415 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.020 31.005 31.414 prepare_preconditioner 11 7.9 0.000 0.001 31.315 31.365 make_preconditioner 11 8.9 0.000 0.002 31.315 31.365 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.936 30.341 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.905 28.282 multiply_cannon_multrec 20056 15.6 13.583 16.632 22.462 25.422 multiply_cannon_metrocomm3 20056 15.6 0.060 0.064 15.141 24.482 qs_ot_get_derivative 117 11.6 0.001 0.002 23.309 23.620 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.919 20.888 apply_single 128 13.6 0.001 0.001 19.919 20.888 qs_ot_get_p 128 10.4 0.004 0.014 19.952 20.372 ot_diis_step 117 11.6 0.018 0.018 19.686 19.687 multiply_cannon_sync_h2d 20056 15.6 14.240 15.742 14.240 15.742 make_m2s 5014 13.6 0.081 0.085 14.779 15.662 make_images 5014 14.6 1.181 1.256 14.546 15.430 qs_ot_p2m_diag 83 11.4 0.265 0.272 15.412 15.426 sum_up_and_integrate 128 10.3 0.135 0.146 14.900 14.930 integrate_v_rspace 128 11.3 0.010 0.030 14.764 14.796 cp_fm_cholesky_invert 11 10.9 14.656 14.665 14.656 14.665 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.607 14.659 calculate_rho_elec 128 8.7 0.132 0.147 14.606 14.659 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.357 14.361 cp_fm_diag_elpa 83 13.4 0.001 0.003 11.293 11.294 cp_fm_redistribute_end 83 14.4 4.273 11.224 4.292 11.227 cp_fm_diag_elpa_base 83 14.4 6.497 10.610 6.910 11.076 init_scf_run 11 5.9 0.000 0.002 10.867 10.867 scf_env_initial_rho_setup 11 6.9 0.002 0.014 10.866 10.867 make_images_data 5014 15.6 0.060 0.068 9.147 10.393 density_rs2pw 128 9.7 0.006 0.006 7.424 9.607 hybrid_alltoall_any 5200 16.5 0.433 1.984 7.891 9.394 pw_transfer 1547 11.6 0.086 0.108 9.226 9.345 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.059 9.284 multiply_cannon_metrocomm4 17549 15.6 0.061 0.070 3.400 9.182 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.002 9.126 mp_irecv_dv 50230 16.2 3.280 8.936 3.280 8.936 dbcsr_mm_accdrv_process 41502 16.2 4.534 5.434 8.332 8.504 fft_wrap_pw1pw2_140 523 13.2 0.477 0.518 7.953 8.082 grid_integrate_task_list 128 12.3 7.298 7.886 7.298 7.886 rs_pw_transfer 1046 11.9 0.013 0.014 5.415 7.670 cp_fm_cholesky_decompose 22 10.9 7.487 7.525 7.487 7.525 wfi_extrapolate 11 7.9 0.001 0.001 7.479 7.479 cp_fm_upper_to_full 105 14.5 5.756 7.476 5.756 7.476 fft3d_ps 1291 14.7 2.674 2.906 7.261 7.347 dbcsr_complete_redistribute 395 12.7 1.164 1.193 4.506 6.268 calculate_dm_sparse 128 9.5 0.001 0.002 6.015 6.118 grid_collocate_task_list 128 9.7 4.902 5.727 4.902 5.727 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.770 5.465 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.423 5.429 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.310 5.066 potential_pw2rs 128 12.3 0.020 0.023 4.955 4.971 mp_alltoall_d11v 2415 14.1 4.352 4.967 4.352 4.967 mp_sum_l 7870 13.0 3.398 4.805 3.398 4.805 mp_waitany 11748 13.9 2.475 4.750 2.475 4.750 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.538 4.660 mp_allgather_i34 2507 14.6 1.625 4.574 1.625 4.574 rs_pw_transfer_RS2PW_140 139 11.5 0.330 0.356 2.056 4.287 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.358 4.085 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.039 4.077 qs_energies_init_hamiltonians 11 5.9 0.004 0.008 3.892 3.903 mp_alltoall_i22 716 14.1 1.939 3.825 1.939 3.825 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.618 3.660 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=180.066000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=898.454545, yerr=5.789703 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.145504E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4077 57304. MP_Allreduce 11231 1068. MP_Sync 170 MP_Alltoall 1712 12503084. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.463 0.696 192.515 192.612 qs_mol_dyn_low 1 2.0 0.005 0.020 190.854 190.867 qs_forces 11 3.9 0.004 0.016 190.716 190.723 qs_energies 11 4.9 0.008 0.053 183.401 183.410 scf_env_do_scf 11 5.9 0.002 0.017 165.978 165.987 velocity_verlet 10 3.0 0.001 0.001 126.433 126.436 scf_env_do_scf_inner_loop 117 6.6 0.008 0.027 118.667 118.669 dbcsr_multiply_generic 2507 12.6 0.190 0.197 80.327 81.373 qs_scf_new_mos 117 7.6 0.001 0.001 80.713 81.046 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.712 81.045 ot_scf_mini 117 9.6 0.004 0.006 76.225 76.530 multiply_cannon 2507 13.6 0.555 0.593 55.470 59.293 multiply_cannon_loop 2507 14.6 1.179 1.208 51.690 54.165 init_scf_loop 11 6.9 0.001 0.004 47.159 47.161 ot_mini 117 10.6 0.001 0.001 43.035 43.316 prepare_preconditioner 11 7.9 0.000 0.001 42.956 42.989 make_preconditioner 11 8.9 0.000 0.002 42.956 42.989 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.597 41.613 multiply_cannon_multrec 30084 15.6 14.105 19.263 26.590 31.298 rebuild_ks_matrix 128 8.3 0.001 0.001 30.286 30.648 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.021 30.286 30.647 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.227 27.547 mp_waitall_1 147882 16.7 17.531 26.771 17.531 26.771 qs_ot_get_derivative 117 11.6 0.001 0.002 23.343 23.650 make_m2s 5014 13.6 0.096 0.098 20.224 21.179 make_images 5014 14.6 1.955 2.273 19.918 20.874 qs_ot_get_p 128 10.4 0.001 0.001 19.415 19.753 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.013 19.673 apply_single 128 13.6 0.001 0.001 19.013 19.673 ot_diis_step 117 11.6 0.018 0.019 19.553 19.555 cp_fm_cholesky_invert 11 10.9 16.649 16.658 16.649 16.658 cp_fm_upper_to_full 105 14.7 11.060 16.362 11.060 16.362 multiply_cannon_metrocomm3 30084 15.6 0.047 0.050 6.568 15.360 qs_ot_p2m_diag 83 11.4 0.342 0.389 15.097 15.150 sum_up_and_integrate 128 10.3 0.141 0.154 14.944 14.983 integrate_v_rspace 128 11.3 0.004 0.005 14.803 14.846 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.785 14.812 calculate_rho_elec 128 8.7 0.175 0.190 14.785 14.812 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.795 13.796 multiply_cannon_sync_h2d 30084 15.6 11.699 13.197 11.699 13.197 dbcsr_complete_redistribute 395 12.7 1.502 1.626 8.949 12.673 dbcsr_mm_accdrv_process 62264 16.2 7.702 8.833 12.061 12.647 make_images_data 5014 15.6 0.064 0.069 10.901 12.180 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.565 11.317 init_scf_run 11 5.9 0.001 0.003 11.264 11.266 scf_env_initial_rho_setup 11 6.9 0.002 0.006 11.264 11.266 hybrid_alltoall_any 5200 16.5 0.525 2.206 9.947 11.264 cp_fm_diag_elpa 83 13.4 0.009 0.013 10.644 10.647 cp_fm_redistribute_end 83 14.4 1.847 10.558 1.863 10.563 cp_fm_diag_elpa_base 83 14.4 8.081 9.941 8.668 10.427 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.336 10.027 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.559 9.788 pw_transfer 1547 11.6 0.088 0.109 9.581 9.654 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 9.352 9.433 mp_alltoall_i22 716 14.1 5.593 9.343 5.593 9.343 density_rs2pw 128 9.7 0.006 0.006 7.240 8.506 fft_wrap_pw1pw2_140 523 13.2 0.482 0.494 8.285 8.376 grid_integrate_task_list 128 12.3 7.491 7.959 7.491 7.959 cp_fm_cholesky_decompose 22 10.9 7.788 7.894 7.788 7.894 wfi_extrapolate 11 7.9 0.001 0.001 7.776 7.777 fft3d_ps 1291 14.7 2.774 2.859 7.576 7.640 multiply_cannon_metrocomm4 25070 15.6 0.076 0.086 2.795 7.412 mp_irecv_dv 76098 16.2 2.650 7.145 2.650 7.145 calculate_dm_sparse 128 9.5 0.001 0.002 6.513 6.596 rs_pw_transfer 1046 11.9 0.013 0.014 4.844 6.165 grid_collocate_task_list 128 9.7 5.085 5.759 5.085 5.759 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.419 5.463 mp_alltoall_d11v 2415 14.1 4.812 5.329 4.812 5.329 potential_pw2rs 128 12.3 0.023 0.023 4.785 4.808 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.618 4.697 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.551 4.555 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.345 4.437 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.228 4.280 mp_sum_l 7870 13.0 3.091 4.145 3.091 4.145 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=192.612000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1074.454545, yerr=23.507692 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/19/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 5.865088E+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 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.522471E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+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 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350154512 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58352. MP_Allreduce 10977 1175. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.076 0.165 179.022 179.023 qs_mol_dyn_low 1 2.0 0.003 0.003 178.064 178.078 qs_forces 11 3.9 0.003 0.003 177.950 177.957 qs_energies 11 4.9 0.001 0.002 170.381 170.392 scf_env_do_scf 11 5.9 0.001 0.001 152.038 152.043 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 115.617 115.619 velocity_verlet 10 3.0 0.001 0.001 115.548 115.552 qs_scf_new_mos 117 7.6 0.001 0.001 76.356 76.415 qs_scf_loop_do_ot 117 8.6 0.001 0.001 76.355 76.414 dbcsr_multiply_generic 2507 12.6 0.183 0.188 75.077 75.622 ot_scf_mini 117 9.6 0.003 0.004 71.923 71.982 multiply_cannon 2507 13.6 0.583 0.623 55.206 59.813 multiply_cannon_loop 2507 14.6 0.445 0.455 50.604 51.690 ot_mini 117 10.6 0.001 0.001 39.977 40.029 init_scf_loop 11 6.9 0.000 0.000 36.276 36.278 mp_waitall_1 125778 16.7 25.681 33.236 25.681 33.236 prepare_preconditioner 11 7.9 0.000 0.000 32.309 32.331 make_preconditioner 11 8.9 0.000 0.000 32.309 32.331 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.133 30.418 rebuild_ks_matrix 128 8.3 0.001 0.001 30.238 30.328 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 30.237 30.328 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.505 27.589 multiply_cannon_multrec 10028 15.6 10.473 14.656 17.791 21.003 qs_ot_get_derivative 117 11.6 0.001 0.002 20.014 20.071 ot_diis_step 117 11.6 0.020 0.021 19.887 19.887 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.456 19.760 apply_single 128 13.6 0.001 0.001 19.456 19.760 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 12.084 19.178 qs_ot_get_p 128 10.4 0.001 0.001 19.063 19.157 cp_fm_cholesky_invert 11 10.9 18.496 18.502 18.496 18.502 make_m2s 5014 13.6 0.066 0.072 16.196 18.457 make_images 5014 14.6 2.293 2.659 15.890 18.143 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.700 15.764 calculate_rho_elec 128 8.7 0.259 0.269 15.699 15.763 sum_up_and_integrate 128 10.3 0.180 0.190 15.311 15.359 qs_ot_p2m_diag 83 11.4 0.495 0.501 15.268 15.283 integrate_v_rspace 128 11.3 0.004 0.004 15.131 15.187 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.117 14.120 make_images_data 5014 15.6 0.052 0.060 9.884 12.402 multiply_cannon_sync_h2d 10028 15.6 11.608 12.140 11.608 12.140 hybrid_alltoall_any 5200 16.5 0.837 3.816 9.657 12.136 init_scf_run 11 5.9 0.000 0.001 11.134 11.134 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.134 11.134 cp_fm_diag_elpa 83 13.4 0.001 0.001 10.972 10.983 cp_fm_diag_elpa_base 83 14.4 10.709 10.805 10.960 10.970 pw_transfer 1547 11.6 0.086 0.092 10.014 10.056 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.789 9.836 fft_wrap_pw1pw2_140 523 13.2 0.495 0.518 8.616 8.669 grid_integrate_task_list 128 12.3 7.705 8.261 7.705 8.261 cp_fm_cholesky_decompose 22 10.9 8.057 8.167 8.057 8.167 density_rs2pw 128 9.7 0.005 0.006 7.228 8.120 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 7.944 7.981 fft3d_ps 1291 14.7 2.708 2.787 7.946 7.964 dbcsr_mm_accdrv_process 20762 16.1 2.734 3.430 6.949 7.598 mp_allgather_i34 2507 14.6 2.803 7.554 2.803 7.554 wfi_extrapolate 11 7.9 0.001 0.001 7.394 7.394 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 4.432 7.330 calculate_dm_sparse 128 9.5 0.001 0.001 6.085 6.186 mp_alltoall_d11v 2415 14.1 5.172 6.150 5.172 6.150 grid_collocate_task_list 128 9.7 5.391 6.070 5.391 6.070 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.835 5.844 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 5.712 5.714 dbcsr_complete_redistribute 395 12.7 2.123 2.206 5.238 5.702 rs_pw_transfer 1046 11.9 0.012 0.013 4.383 5.520 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.773 4.811 potential_pw2rs 128 12.3 0.027 0.028 4.689 4.703 multiply_cannon_metrocomm4 7521 15.6 0.024 0.027 1.865 4.408 mp_irecv_dv 28860 15.9 1.828 4.329 1.828 4.329 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.558 3.870 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.785 3.813 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.434 3.792 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.707 3.756 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.551 3.664 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.641 3.648 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=179.023000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1417.090909, yerr=47.054040 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_performance_tests/20/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 11.606413E+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 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.020026E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108580288 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 59070. MP_Allreduce 11015 1514. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.091 305.762 305.767 qs_mol_dyn_low 1 2.0 0.003 0.003 305.066 305.077 qs_forces 11 3.9 0.003 0.003 304.966 304.974 qs_energies 11 4.9 0.002 0.002 295.588 295.603 scf_env_do_scf 11 5.9 0.001 0.001 272.344 272.355 velocity_verlet 10 3.0 0.001 0.001 217.861 217.869 scf_env_do_scf_inner_loop 116 6.6 0.004 0.008 144.776 144.777 init_scf_loop 11 6.9 0.000 0.000 127.288 127.291 prepare_preconditioner 11 7.9 0.000 0.000 122.072 122.104 make_preconditioner 11 8.9 0.000 0.000 122.072 122.104 make_full_inverse_cholesky 11 9.9 0.000 0.000 98.014 119.203 qs_scf_new_mos 116 7.6 0.001 0.001 94.025 94.248 qs_scf_loop_do_ot 116 8.6 0.001 0.001 94.024 94.248 ot_scf_mini 116 9.6 0.003 0.004 89.218 89.382 dbcsr_multiply_generic 2485 12.5 0.225 0.231 84.697 85.506 cp_fm_upper_to_full 104 14.8 53.430 76.458 53.430 76.458 multiply_cannon 2485 13.5 0.703 0.747 59.291 59.820 multiply_cannon_loop 2485 14.5 0.467 0.477 55.316 57.124 ot_mini 116 10.6 0.001 0.001 45.896 46.090 dbcsr_complete_redistribute 393 12.7 3.972 4.034 29.810 42.807 copy_fm_to_dbcsr 208 11.6 0.002 0.002 26.357 39.370 rebuild_ks_matrix 127 8.3 0.001 0.001 38.394 38.530 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 38.393 38.530 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.013 36.880 qs_ks_update_qs_env 127 7.6 0.001 0.001 35.321 35.452 mp_alltoall_i22 712 14.1 21.905 35.270 21.905 35.270 cp_fm_cholesky_invert 11 10.9 35.037 35.043 35.037 35.043 mp_waitall_1 102768 16.8 28.853 33.039 28.853 33.039 qs_ot_get_p 127 10.4 0.001 0.001 27.735 27.934 qs_ot_get_derivative 116 11.6 0.002 0.002 25.352 25.522 qs_ot_p2m_diag 82 11.4 0.868 0.874 23.483 23.517 cp_dbcsr_syevd 82 12.4 0.005 0.006 21.715 21.717 make_m2s 4970 13.5 0.075 0.077 20.136 21.383 make_images 4970 14.5 3.708 3.823 19.662 20.905 multiply_cannon_metrocomm3 9940 15.5 0.024 0.024 19.116 20.852 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.715 20.725 calculate_rho_elec 127 8.7 0.479 0.480 20.714 20.724 ot_diis_step 116 11.6 0.022 0.022 20.495 20.496 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.634 19.980 apply_single 127 13.6 0.001 0.001 19.634 19.980 sum_up_and_integrate 127 10.3 0.318 0.320 19.532 19.611 integrate_v_rspace 127 11.3 0.004 0.004 19.214 19.293 multiply_cannon_multrec 9940 15.5 10.639 12.507 18.128 18.510 cp_fm_diag_elpa 82 13.4 0.002 0.005 18.417 18.418 cp_fm_diag_elpa_base 82 14.4 14.056 15.684 18.408 18.408 multiply_cannon_sync_h2d 9940 15.5 15.535 15.544 15.535 15.544 init_scf_run 11 5.9 0.000 0.001 13.107 13.112 scf_env_initial_rho_setup 11 6.9 0.020 0.033 13.107 13.112 hybrid_alltoall_any 5155 16.4 1.294 3.038 10.998 12.972 make_images_data 4970 15.5 0.059 0.064 10.914 12.845 pw_transfer 1535 11.6 0.091 0.092 12.465 12.474 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 12.233 12.242 fft_wrap_pw1pw2_140 519 13.2 0.538 0.545 10.792 10.807 fft3d_ps 1281 14.7 2.726 2.743 10.281 10.295 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.718 9.834 cp_fm_cholesky_decompose 22 10.9 9.282 9.397 9.282 9.397 wfi_extrapolate 11 7.9 0.001 0.001 9.225 9.225 dbcsr_mm_accdrv_process 20590 16.0 3.810 5.734 7.251 9.027 mp_alltoall_d11v 2401 14.1 8.075 8.858 8.075 8.858 grid_integrate_task_list 127 12.3 8.514 8.674 8.514 8.674 density_rs2pw 127 9.7 0.005 0.005 8.412 8.533 qs_energies_init_hamiltonians 11 5.9 0.017 0.021 8.136 8.143 calculate_dm_sparse 127 9.5 0.001 0.001 6.685 6.756 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.424 6.498 grid_collocate_task_list 127 9.7 6.316 6.343 6.316 6.343 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.172 6.258 rs_scatter_matrices 138 9.7 3.553 4.370 5.946 6.175 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=305.767000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2677.272727, yerr=171.162916 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_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 1.262285E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 6805335. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.042 85.637 85.648 qs_energies 1 2.0 0.000 0.000 85.106 85.113 ls_scf 1 3.0 0.000 0.000 83.775 83.781 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.477 72.627 multiply_cannon 111 7.7 0.017 0.020 55.895 56.958 multiply_cannon_loop 111 8.7 0.207 0.227 52.428 53.639 ls_scf_main 1 4.0 0.000 0.000 52.349 52.349 density_matrix_trs4 2 5.0 0.002 0.003 46.807 46.899 ls_scf_init_scf 1 4.0 0.000 0.000 28.320 28.321 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.160 27.218 mp_waitall_1 11031 10.9 21.952 25.799 21.952 25.799 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.013 25.038 multiply_cannon_multrec 2664 9.7 8.185 8.974 15.588 17.369 multiply_cannon_sync_h2d 2664 9.7 13.843 15.845 13.843 15.845 make_m2s 222 7.7 0.008 0.011 12.968 13.559 make_images 222 8.7 0.099 0.109 12.946 13.539 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.321 12.434 make_images_data 222 9.7 0.004 0.005 7.549 8.236 dbcsr_mm_accdrv_process 4760 10.4 0.504 0.607 7.023 7.985 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.410 7.853 hybrid_alltoall_any 227 10.6 0.215 1.838 6.551 7.764 dbcsr_mm_accdrv_process_sort 4760 11.4 6.319 7.244 6.319 7.244 calculate_norms 4752 9.8 5.581 6.152 5.581 6.152 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.975 5.226 mp_sum_l 807 5.4 3.149 5.044 3.149 5.044 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.065 3.570 mp_irecv_dv 6231 10.9 2.046 3.537 2.046 3.537 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.254 3.457 arnoldi_extremal 4 6.8 0.000 0.000 3.336 3.375 arnoldi_normal_ev 4 7.8 0.001 0.003 3.336 3.375 build_subspace 16 8.4 0.009 0.012 3.233 3.236 make_images_sizes 222 9.7 0.000 0.000 0.692 3.196 mp_alltoall_i44 222 10.7 0.691 3.196 0.691 3.196 ls_scf_post 1 4.0 0.000 0.000 3.106 3.113 ls_scf_store_result 1 5.0 0.000 0.000 2.831 2.871 dbcsr_special_finalize 555 9.7 0.005 0.006 2.376 2.818 dbcsr_merge_single_wm 555 10.7 0.467 0.601 2.368 2.810 make_images_pack 222 9.7 2.205 2.618 2.207 2.620 dbcsr_matrix_vector_mult 304 9.0 0.005 0.012 2.357 2.594 dbcsr_sort_data 658 11.4 2.159 2.554 2.159 2.554 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.458 2.069 2.459 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.336 2.458 buffer_matrices_ensure_size 222 8.7 1.787 2.108 1.787 2.108 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.825 1.827 rebuild_ks_matrix 3 7.3 0.000 0.000 1.816 1.818 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.013 1.816 1.818 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.648000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_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 2.120901E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.078 0.115 92.261 92.264 qs_energies 1 2.0 0.000 0.000 91.479 91.497 ls_scf 1 3.0 0.000 0.001 90.045 90.062 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.413 75.795 multiply_cannon 111 7.7 0.028 0.041 53.223 57.174 ls_scf_main 1 4.0 0.000 0.000 55.211 55.217 multiply_cannon_loop 111 8.7 0.115 0.124 49.981 53.574 density_matrix_trs4 2 5.0 0.002 0.003 49.435 49.628 ls_scf_init_scf 1 4.0 0.000 0.001 31.341 31.343 ls_scf_init_matrix_S 1 5.0 0.000 0.001 30.043 30.156 mp_waitall_1 9105 10.9 21.110 29.509 21.110 29.509 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.613 27.626 multiply_cannon_multrec 1332 9.7 13.037 16.880 22.334 27.372 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.780 21.209 make_m2s 222 7.7 0.006 0.008 15.277 15.838 make_images 222 8.7 1.583 1.919 15.247 15.808 dbcsr_mm_accdrv_process 4041 10.4 0.267 0.456 8.908 10.462 dbcsr_mm_accdrv_process_sort 4041 11.4 8.495 10.006 8.495 10.006 make_images_data 222 9.7 0.004 0.004 8.740 9.602 hybrid_alltoall_any 227 10.6 0.523 2.456 8.185 9.091 mp_sum_l 807 5.4 5.656 8.369 5.656 8.369 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.227 7.924 mp_irecv_dv 3311 11.0 3.208 7.873 3.208 7.873 calculate_norms 2376 9.8 6.009 6.811 6.009 6.811 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.302 6.727 multiply_cannon_sync_h2d 1332 9.7 4.792 5.828 4.792 5.828 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.074 5.311 arnoldi_extremal 4 6.8 0.000 0.000 4.751 4.768 arnoldi_normal_ev 4 7.8 0.001 0.005 4.751 4.768 build_subspace 16 8.4 0.014 0.021 4.480 4.484 ls_scf_post 1 4.0 0.005 0.025 3.492 3.510 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.179 3.444 ls_scf_store_result 1 5.0 0.000 0.000 3.234 3.310 dbcsr_matrix_vector_mult_local 304 10.0 2.737 3.221 2.739 3.223 ls_scf_dm_to_ks 2 5.0 0.001 0.020 2.650 2.782 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.222 2.621 make_images_pack 222 9.7 2.033 2.431 2.035 2.433 mp_allgather_i34 111 8.7 0.943 2.253 0.943 2.253 dbcsr_sort_data 436 11.2 1.822 2.064 1.822 2.064 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.916 1.918 rebuild_ks_matrix 3 7.3 0.000 0.000 1.903 1.906 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.903 1.906 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.264000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1714.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_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 2.716725E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.137 95.076 95.077 qs_energies 1 2.0 0.000 0.000 94.302 94.315 ls_scf 1 3.0 0.010 0.070 92.564 92.572 dbcsr_multiply_generic 111 6.7 0.015 0.016 77.197 77.495 ls_scf_main 1 4.0 0.000 0.001 58.209 58.235 multiply_cannon 111 7.7 0.040 0.088 53.003 56.710 multiply_cannon_loop 111 8.7 0.099 0.109 49.468 53.858 density_matrix_trs4 2 5.0 0.002 0.003 52.134 52.371 mp_waitall_1 7281 11.0 24.203 34.515 24.203 34.515 ls_scf_init_scf 1 4.0 0.000 0.002 30.763 30.838 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.552 29.624 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.144 27.156 multiply_cannon_multrec 888 9.7 12.643 15.217 21.235 24.398 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.067 23.627 make_m2s 222 7.7 0.006 0.007 17.323 18.579 make_images 222 8.7 1.971 2.301 17.284 18.540 hybrid_alltoall_any 227 10.6 0.616 2.820 9.529 11.092 make_images_data 222 9.7 0.003 0.004 9.962 11.028 dbcsr_mm_accdrv_process 3754 10.4 0.241 0.416 8.141 9.386 dbcsr_mm_accdrv_process_sort 3754 11.4 7.770 8.970 7.770 8.970 mp_sum_l 807 5.4 5.322 8.884 5.322 8.884 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.791 7.464 multiply_cannon_sync_h2d 888 9.7 5.980 7.110 5.980 7.110 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.452 6.985 mp_irecv_dv 2335 11.1 2.437 6.930 2.437 6.930 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.060 6.897 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.940 5.187 arnoldi_extremal 4 6.8 0.000 0.000 5.061 5.076 arnoldi_normal_ev 4 7.8 0.001 0.006 5.060 5.076 build_subspace 16 8.4 0.014 0.020 4.750 4.756 calculate_norms 1584 9.8 4.320 4.731 4.320 4.731 mp_allgather_i34 111 8.7 1.373 3.896 1.373 3.896 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.418 3.747 ls_scf_post 1 4.0 0.000 0.003 3.582 3.592 dbcsr_matrix_vector_mult_local 304 10.0 3.013 3.582 3.015 3.584 ls_scf_store_result 1 5.0 0.000 0.000 3.316 3.405 ls_scf_dm_to_ks 2 5.0 0.003 0.034 2.865 2.958 dbcsr_sort_data 325 11.1 1.893 2.173 1.893 2.173 make_images_sizes 222 9.7 0.000 0.000 1.037 2.155 mp_alltoall_i44 222 10.7 1.037 2.154 1.037 2.154 make_images_pack 222 9.7 1.809 2.114 1.812 2.116 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.965 2.039 rebuild_ks_matrix 3 7.3 0.000 0.000 1.948 2.021 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.008 1.948 2.021 dbcsr_data_release 9322 10.9 1.288 1.962 1.288 1.962 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.077000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2180.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_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 3.370557E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.062 0.089 98.453 98.454 qs_energies 1 2.0 0.000 0.000 97.675 97.680 ls_scf 1 3.0 0.000 0.004 95.874 95.880 dbcsr_multiply_generic 111 6.7 0.016 0.017 79.560 79.833 ls_scf_main 1 4.0 0.000 0.000 59.541 59.542 multiply_cannon 111 7.7 0.052 0.117 52.329 56.721 density_matrix_trs4 2 5.0 0.002 0.003 53.350 53.496 multiply_cannon_loop 111 8.7 0.114 0.125 47.256 50.740 ls_scf_init_scf 1 4.0 0.000 0.002 33.030 33.032 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.747 31.857 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.267 29.277 mp_waitall_1 6369 11.0 23.328 28.702 23.328 28.702 multiply_cannon_multrec 1332 9.7 14.187 17.357 22.199 25.349 make_m2s 222 7.7 0.007 0.008 21.185 22.616 make_images 222 8.7 3.139 3.586 21.135 22.567 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.605 17.611 make_images_data 222 9.7 0.004 0.004 11.840 13.532 hybrid_alltoall_any 227 10.6 0.798 3.791 11.001 12.813 dbcsr_mm_accdrv_process 3641 10.4 0.267 0.663 7.642 9.107 mp_sum_l 807 5.4 4.478 8.985 4.478 8.985 dbcsr_mm_accdrv_process_sort 3641 11.4 7.194 8.693 7.194 8.693 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.456 7.236 multiply_cannon_sync_h2d 1332 9.7 5.437 6.038 5.437 6.038 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.075 6.001 mp_irecv_dv 3229 10.9 2.051 5.957 2.051 5.957 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.714 5.610 arnoldi_extremal 4 6.8 0.000 0.000 5.242 5.262 arnoldi_normal_ev 4 7.8 0.001 0.004 5.242 5.262 build_subspace 16 8.4 0.014 0.021 4.908 4.918 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.607 4.783 calculate_norms 2376 9.8 4.217 4.635 4.217 4.635 mp_allgather_i34 111 8.7 2.196 4.556 2.196 4.556 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.594 3.888 dbcsr_matrix_vector_mult_local 304 10.0 3.179 3.688 3.181 3.690 dbcsr_sort_data 658 11.4 3.019 3.315 3.019 3.315 ls_scf_post 1 4.0 0.000 0.001 3.303 3.308 ls_scf_dm_to_ks 2 5.0 0.000 0.002 3.045 3.105 ls_scf_store_result 1 5.0 0.000 0.000 3.026 3.104 dbcsr_special_finalize 555 9.7 0.006 0.007 2.764 3.088 dbcsr_merge_single_wm 555 10.7 0.528 0.646 2.756 3.079 dbcsr_data_release 10477 10.7 1.590 2.385 1.590 2.385 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.985 1.986 dbcsr_finalize 304 7.8 0.049 0.061 1.797 1.972 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.454000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2770.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_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 4.627743E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.088 0.095 99.295 99.296 qs_energies 1 2.0 0.000 0.000 98.332 98.333 ls_scf 1 3.0 0.000 0.000 96.379 96.385 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.544 77.786 ls_scf_main 1 4.0 0.000 0.000 62.043 62.043 multiply_cannon 111 7.7 0.093 0.155 55.200 60.745 density_matrix_trs4 2 5.0 0.002 0.003 54.801 54.905 multiply_cannon_loop 111 8.7 0.071 0.085 50.564 52.302 mp_waitall_1 5436 11.0 26.129 32.371 26.129 32.371 ls_scf_init_scf 1 4.0 0.000 0.000 30.748 30.753 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.372 29.423 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.177 27.184 multiply_cannon_multrec 444 9.7 13.972 16.190 21.093 24.103 make_m2s 222 7.7 0.004 0.005 17.547 20.160 make_images 222 8.7 3.721 4.432 17.485 20.099 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.168 15.820 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.941 14.633 make_images_data 222 9.7 0.003 0.004 9.818 12.331 hybrid_alltoall_any 227 10.6 0.787 3.747 9.616 12.143 multiply_cannon_sync_h2d 444 9.7 6.538 8.376 6.538 8.376 dbcsr_mm_accdrv_process 3003 10.4 0.175 0.344 6.819 7.948 dbcsr_mm_accdrv_process_sort 3003 11.4 6.503 7.602 6.503 7.602 mp_allgather_i34 111 8.7 2.865 7.109 2.865 7.109 arnoldi_extremal 4 6.8 0.000 0.000 5.931 5.942 arnoldi_normal_ev 4 7.8 0.002 0.005 5.931 5.942 mp_sum_l 807 5.4 2.846 5.606 2.846 5.606 build_subspace 16 8.4 0.015 0.019 5.525 5.537 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.598 4.727 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.240 4.440 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.974 4.373 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.641 4.191 mp_irecv_dv 1241 11.2 1.621 4.175 1.621 4.175 dbcsr_matrix_vector_mult_local 304 10.0 3.675 4.146 3.677 4.148 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.611 3.734 calculate_norms 792 9.8 3.537 3.654 3.537 3.654 ls_scf_post 1 4.0 0.000 0.000 3.588 3.590 ls_scf_store_result 1 5.0 0.000 0.000 3.340 3.406 make_images_sizes 222 9.7 0.000 0.000 1.086 3.352 mp_alltoall_i44 222 10.7 1.085 3.351 1.085 3.351 dbcsr_finalize 304 7.8 0.063 0.078 2.197 2.273 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.185 2.187 rebuild_ks_matrix 3 7.3 0.000 0.000 2.153 2.154 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.153 2.154 dbcsr_merge_all 275 8.9 0.475 0.523 2.046 2.110 dbcsr_data_release 10123 10.8 1.322 1.991 1.322 1.991 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.296000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3660.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d51b7226a3a06d954c541917aa996a97b60aada9_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 8.668791E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.175 0.189 108.881 108.882 qs_energies 1 2.0 0.000 0.000 106.992 107.005 ls_scf 1 3.0 0.000 0.000 104.037 104.049 dbcsr_multiply_generic 111 6.7 0.023 0.027 76.949 77.092 ls_scf_main 1 4.0 0.000 0.000 65.650 65.650 density_matrix_trs4 2 5.0 0.002 0.003 56.302 56.373 multiply_cannon 111 7.7 0.156 0.293 49.414 51.564 multiply_cannon_loop 111 8.7 0.067 0.071 46.048 46.669 ls_scf_init_scf 1 4.0 0.000 0.000 34.706 34.707 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.899 32.907 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.102 30.112 make_m2s 222 7.7 0.005 0.005 23.832 24.703 mp_waitall_1 4527 11.1 21.266 24.648 21.266 24.648 make_images 222 8.7 4.596 4.976 23.726 24.594 multiply_cannon_multrec 444 9.7 17.889 18.617 22.544 23.317 hybrid_alltoall_any 227 10.6 1.662 3.631 12.784 15.628 make_images_data 222 9.7 0.003 0.003 13.014 15.378 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.690 10.748 multiply_cannon_sync_h2d 444 9.7 8.844 8.884 8.844 8.884 arnoldi_extremal 4 6.8 0.000 0.000 7.601 7.615 arnoldi_normal_ev 4 7.8 0.003 0.009 7.601 7.615 build_subspace 16 8.4 0.026 0.037 6.998 7.009 dbcsr_matrix_vector_mult 304 9.0 0.016 0.032 5.604 5.765 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.302 5.389 dbcsr_matrix_vector_mult_local 304 10.0 4.991 5.296 4.993 5.298 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.998 5.238 dbcsr_mm_accdrv_process 1814 10.4 0.230 0.316 4.484 4.609 dbcsr_mm_accdrv_process_sort 1814 11.4 4.184 4.313 4.184 4.313 ls_scf_post 1 4.0 0.000 0.000 3.681 3.694 make_images_sizes 222 9.7 0.000 0.000 1.469 3.669 mp_alltoall_i44 222 10.7 1.468 3.668 1.468 3.668 ls_scf_store_result 1 5.0 0.000 0.000 3.388 3.399 calculate_norms 792 9.8 3.244 3.289 3.244 3.289 dbcsr_finalize 304 7.8 0.082 0.089 3.085 3.121 mp_allgather_i34 111 8.7 0.989 2.980 0.989 2.980 dbcsr_complete_redistribute 5 7.6 1.443 1.497 2.777 2.927 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.925 2.925 dbcsr_merge_all 275 8.9 0.892 0.916 2.871 2.903 dbcsr_data_release 12724 10.6 2.341 2.862 2.341 2.862 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.625 2.626 matrix_ls_to_qs 2 6.0 0.000 0.000 2.420 2.574 rebuild_ks_matrix 3 7.3 0.000 0.000 2.558 2.560 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.558 2.560 dbcsr_sort_data 325 11.1 2.442 2.501 2.442 2.501 dbcsr_new_transposed 4 7.5 0.243 0.251 2.298 2.315 dbcsr_frobenius_norm 74 6.6 2.058 2.143 2.199 2.229 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.205 dbcsr_add_anytype 103 7.2 0.859 0.891 2.132 2.205 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.882000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6817.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: d51b7226a3a06d954c541917aa996a97b60aada9 Summary: empty Status: OK