=== This is the CP2K Performance-Test === Already up to date. Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb ################# 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 6.0.0, 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 (26.12.2022) # # \ 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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/01 job id: 44091206 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/02 job id: 44091207 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/03 job id: 44091208 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/04 job id: 44091209 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/05 job id: 44091210 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/06 job id: 44091211 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/07 job id: 44091212 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/08 job id: 44091213 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/09 job id: 44091214 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/10 job id: 44091215 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/11 job id: 44091216 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/12 job id: 44091217 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/13 job id: 44091218 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/14 job id: 44091219 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/15 job id: 44091220 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/16 job id: 44091221 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/17 job id: 44091222 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/18 job id: 44091223 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/19 job id: 44091224 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/20 job id: 44091225 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/21 job id: 44091226 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/22 job id: 44091227 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/23 job id: 44091228 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/24 job id: 44091229 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/25 job id: 44091230 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/26 job id: 44091231 --- 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 0.021 0.035 131.614 131.615 farming_run 1 2.0 131.110 131.111 131.586 131.590 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.454801E+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 228 1113141. 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 0.011 0.026 114.630 114.630 qs_energies 1 2.0 0.000 0.000 114.434 114.438 mp2_main 1 3.0 0.000 0.000 112.623 112.626 mp2_gpw_main 1 4.0 0.020 0.026 111.833 111.837 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 93.440 93.509 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.460 55.528 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.171 41.771 46.702 get_2c_integrals 1 6.0 0.000 0.002 37.369 37.808 integrate_v_rspace 273 8.0 0.438 0.455 25.109 29.753 pw_transfer 6555 10.6 0.370 0.384 27.385 27.835 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 26.071 26.300 grid_integrate_task_list 273 9.0 20.925 26.025 20.925 26.025 fft_wrap_pw1pw2_100 2178 12.4 1.198 1.428 23.635 23.889 compute_2c_integrals 1 7.0 0.005 0.006 19.562 19.563 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.902 19.297 mp2_eri_2c_integrate_gpw 1 9.0 2.380 2.436 18.898 19.296 rpa_ri_compute_en 1 5.0 0.001 0.001 18.287 18.432 cp_fm_cholesky_decompose 12 8.2 17.749 18.207 17.749 18.207 cholesky_decomp 1 7.0 0.000 0.000 16.652 17.109 fft3d_s 5443 13.4 16.172 16.374 16.194 16.396 ao_to_mo_and_store_B_mult_1 272 7.0 10.849 15.557 10.849 15.557 calculate_wavefunction 272 8.0 5.446 5.520 12.591 13.190 rpa_num_int 1 6.0 0.000 0.001 10.475 10.475 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.438 10.474 calc_mat_Q 8 8.0 0.000 0.000 9.291 9.410 contract_S_to_Q 8 9.0 0.000 0.000 8.711 8.833 calc_potential_gpw 544 9.5 0.005 0.006 8.271 8.658 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.230 8.519 parallel_gemm_fm 14 9.1 0.000 0.000 8.288 8.397 parallel_gemm_fm_cosma 14 10.1 8.288 8.397 8.288 8.397 potential_pw2rs 545 10.0 0.107 0.108 7.670 8.308 collocate_single_gaussian 272 10.0 0.040 0.043 7.468 7.705 create_integ_mat 1 6.0 0.022 0.028 7.647 7.647 array2fm 1 7.0 0.000 0.000 6.651 7.148 pw_scatter_s 2720 13.7 4.458 4.686 4.458 4.686 pw_gather_s 2722 13.2 3.877 4.322 3.877 4.322 array2fm_buffer_send 1 8.0 2.937 3.189 2.937 3.189 pw_poisson_solve 545 10.5 1.135 1.180 2.159 2.377 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.834993, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2727.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 0.029 0.037 397.260 397.262 farming_run 1 2.0 396.522 396.535 397.219 397.227 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.223954E+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 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 21636802. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.040 211.324 211.325 qs_energies 1 2.0 0.000 0.000 211.081 211.101 scf_env_do_scf 1 3.0 0.000 0.000 106.723 106.723 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.876 105.884 rebuild_ks_matrix 4 6.0 0.000 0.000 105.875 105.883 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.064 105.875 105.883 hfx_ks_matrix 4 8.0 0.001 0.001 105.502 105.505 integrate_four_center 4 9.0 0.144 0.456 105.501 105.504 mp2_main 1 3.0 0.004 0.020 104.078 104.098 mp2_gpw_main 1 4.0 0.073 0.178 103.258 103.268 integrate_four_center_main 4 10.0 0.090 0.609 96.729 99.669 integrate_four_center_bin 263 11.0 96.639 99.656 96.639 99.656 init_scf_loop 1 4.0 0.000 0.000 92.543 92.543 mp2_ri_gpw_compute_in 1 5.0 0.068 0.076 75.007 75.996 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.478 55.452 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.159 42.215 47.284 integrate_v_rspace 95 8.0 0.399 0.570 28.580 33.477 pw_transfer 2240 10.6 0.143 0.174 29.862 30.366 mp2_ri_gpw_compute_en 1 5.0 0.065 0.111 28.009 29.682 fft_wrap_pw1pw2 1868 11.4 0.017 0.021 28.876 29.390 grid_integrate_task_list 95 9.0 23.878 29.001 23.878 29.001 ao_to_mo_and_store_B_mult_1 91 7.0 10.576 28.864 10.576 28.864 fft_wrap_pw1pw2_100 730 12.4 1.266 1.440 26.594 27.114 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.843 1.897 26.287 26.298 get_2c_integrals 1 6.0 0.024 0.081 20.400 20.462 compute_2c_integrals 1 7.0 0.009 0.024 19.353 19.378 compute_2c_integrals_loop_lm 1 8.0 0.007 0.017 18.840 19.167 mp2_eri_2c_integrate_gpw 1 9.0 1.739 1.889 18.833 19.151 fft3d_s 1823 13.4 18.399 18.782 18.412 18.795 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.178 14.179 calculate_wavefunction 91 8.0 2.029 2.061 9.749 10.021 mp2_ri_gpw_compute_en_comm 22 7.0 0.467 0.487 9.178 9.783 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.583 8.873 9.651 potential_pw2rs 186 10.0 0.033 0.035 8.591 9.231 local_gemm 172 8.0 8.318 9.095 8.318 9.095 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.224 8.579 calc_potential_gpw 182 9.5 0.002 0.002 7.929 8.169 collocate_single_gaussian 91 10.0 0.019 0.055 7.863 8.078 mp_sendrecv_dm3 2068 8.0 7.306 7.895 7.306 7.895 mp2_ri_gpw_compute_en_ener 172 7.0 6.348 6.453 6.348 6.453 mp_sync 38 10.4 3.449 5.521 3.449 5.521 pw_gather_s 912 13.2 4.896 5.438 4.896 5.438 pw_scatter_s 910 13.7 3.947 4.436 3.947 4.436 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=103.254458, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.239936E+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 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 0.010 0.028 51.649 51.650 qs_mol_dyn_low 1 2.0 0.003 0.003 51.428 51.439 qs_forces 11 3.9 0.002 0.002 51.232 51.233 qs_energies 11 4.9 0.001 0.001 49.787 49.799 scf_env_do_scf 11 5.9 0.000 0.001 43.760 43.760 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 41.739 41.739 dbcsr_multiply_generic 2286 12.5 0.094 0.098 33.426 33.892 qs_scf_new_mos 108 7.5 0.000 0.001 32.018 32.294 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.018 32.293 ot_scf_mini 108 9.5 0.002 0.002 30.398 30.557 multiply_cannon 2286 13.5 0.179 0.186 26.103 27.782 multiply_cannon_loop 2286 14.5 1.503 1.579 25.432 27.128 velocity_verlet 10 3.0 0.001 0.001 25.318 25.320 ot_mini 108 10.5 0.001 0.001 19.219 19.471 qs_ot_get_derivative 108 11.5 0.001 0.001 16.258 16.434 mp_waitall_1 245248 16.5 8.424 15.018 8.424 15.018 multiply_cannon_metrocomm3 54864 15.5 0.067 0.073 6.041 13.146 multiply_cannon_multrec 54864 15.5 4.204 6.595 7.731 11.309 rebuild_ks_matrix 119 8.3 0.000 0.000 7.691 7.820 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.691 7.820 multiply_cannon_sync_h2d 54864 15.5 5.896 7.386 5.896 7.386 mp_sum_l 7207 12.9 5.407 7.020 5.407 7.020 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.781 6.898 qs_ot_get_p 119 10.4 0.001 0.001 6.492 6.822 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.504 5.974 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.474 5.575 init_scf_run 11 5.9 0.000 0.001 4.842 4.842 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.842 4.842 dbcsr_mm_accdrv_process 76910 16.1 1.182 1.818 3.449 4.784 sum_up_and_integrate 119 10.3 0.012 0.015 4.383 4.390 integrate_v_rspace 119 11.3 0.002 0.002 4.371 4.379 qs_rho_update_rho_low 119 7.7 0.000 0.001 3.935 4.049 calculate_rho_elec 119 8.7 0.011 0.016 3.935 4.048 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.692 3.752 calculate_dm_sparse 119 9.5 0.000 0.000 2.975 3.131 multiply_cannon_metrocomm1 54864 15.5 0.051 0.057 1.761 3.103 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.868 3.054 apply_single 119 13.6 0.000 0.000 2.867 3.054 jit_kernel_multiply 13 15.8 2.205 2.911 2.205 2.911 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.862 2.863 rs_pw_transfer 974 11.9 0.012 0.013 2.654 2.752 calculate_first_density_matrix 1 7.0 0.000 0.000 2.742 2.748 ot_diis_step 108 11.5 0.006 0.006 2.678 2.679 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.657 2.657 cp_fm_redistribute_end 50 14.0 2.418 2.635 2.423 2.636 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.492 2.566 cp_fm_diag_elpa_base 50 14.0 0.211 2.548 0.212 2.557 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.391 2.393 acc_transpose_blocks 54864 15.5 0.231 0.252 1.800 2.233 density_rs2pw 119 9.7 0.004 0.004 2.071 2.157 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.076 2.120 grid_integrate_task_list 119 12.3 2.035 2.119 2.035 2.119 wfi_extrapolate 11 7.9 0.001 0.001 2.039 2.039 init_scf_loop 11 6.9 0.000 0.000 2.004 2.004 mp_sum_d 4125 12.0 1.321 1.952 1.321 1.952 potential_pw2rs 119 12.3 0.004 0.004 1.765 1.780 make_m2s 4572 13.5 0.054 0.056 1.603 1.654 pw_transfer 1439 11.6 0.050 0.054 1.547 1.611 make_images 4572 14.5 0.133 0.139 1.520 1.571 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 1.473 1.538 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.511 1.534 mp_alltoall_d11v 2130 13.8 1.226 1.407 1.226 1.407 grid_collocate_task_list 119 9.7 1.289 1.369 1.289 1.369 mp_waitany 12084 13.8 1.226 1.348 1.226 1.348 fft3d_ps 1201 14.6 0.354 0.457 1.251 1.309 fft_wrap_pw1pw2_140 487 13.2 0.080 0.093 1.129 1.194 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.742 1.145 acc_transpose_blocks_kernels 54864 16.5 0.251 0.379 0.807 1.089 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=51.650000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.545455, yerr=0.655555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 487.530496E+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 1907156. 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 0.017 0.037 38.225 38.226 qs_mol_dyn_low 1 2.0 0.003 0.004 37.952 37.961 qs_forces 11 3.9 0.001 0.002 37.880 37.881 qs_energies 11 4.9 0.002 0.005 36.183 36.186 scf_env_do_scf 11 5.9 0.001 0.002 31.122 31.123 scf_env_do_scf_inner_loop 108 6.5 0.003 0.012 28.640 28.641 dbcsr_multiply_generic 2286 12.5 0.100 0.107 21.169 21.519 qs_scf_new_mos 108 7.5 0.001 0.001 19.744 19.990 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.744 19.989 ot_scf_mini 108 9.5 0.003 0.003 18.845 19.019 velocity_verlet 10 3.0 0.002 0.003 17.997 17.998 multiply_cannon 2286 13.5 0.207 0.215 16.198 17.696 multiply_cannon_loop 2286 14.5 0.901 0.980 15.104 16.446 ot_mini 108 10.5 0.001 0.001 11.637 11.872 mp_waitall_1 200699 16.5 5.594 10.776 5.594 10.776 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.082 9.412 qs_ot_get_derivative 108 11.5 0.001 0.001 9.195 9.372 multiply_cannon_multrec 27432 15.5 1.980 4.535 5.853 8.803 rebuild_ks_matrix 119 8.3 0.000 0.000 7.127 7.268 qs_ks_build_kohn_sham_matrix 119 9.3 0.015 0.038 7.126 7.268 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.276 6.405 dbcsr_mm_accdrv_process 47894 16.0 3.002 5.003 3.804 5.639 qs_ot_get_p 119 10.4 0.001 0.001 4.332 4.568 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.547 4.389 sum_up_and_integrate 119 10.3 0.024 0.027 4.147 4.152 integrate_v_rspace 119 11.3 0.002 0.002 4.122 4.131 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.035 4.110 apply_single 119 13.6 0.000 0.000 3.034 4.110 mp_sum_l 7207 12.9 2.028 3.981 2.028 3.981 init_scf_run 11 5.9 0.000 0.001 3.831 3.832 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.831 3.831 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.563 3.600 calculate_rho_elec 119 8.7 0.021 0.024 3.562 3.599 multiply_cannon_sync_h2d 27432 15.5 2.201 2.919 2.201 2.919 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.897 2.916 make_m2s 4572 13.5 0.052 0.055 2.496 2.709 make_images 4572 14.5 0.203 0.256 2.407 2.618 rs_pw_transfer 974 11.9 0.010 0.011 2.526 2.609 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.468 2.468 init_scf_loop 11 6.9 0.001 0.003 2.461 2.463 ot_diis_step 108 11.5 0.011 0.011 2.394 2.395 calculate_first_density_matrix 1 7.0 0.001 0.002 2.341 2.343 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.108 2.201 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.142 2.143 calculate_dm_sparse 119 9.5 0.000 0.001 2.055 2.130 cp_fm_redistribute_end 50 14.0 1.774 2.119 1.778 2.120 cp_fm_diag_elpa_base 50 14.0 0.328 2.015 0.341 2.065 density_rs2pw 119 9.7 0.004 0.004 1.923 2.008 grid_integrate_task_list 119 12.3 1.840 1.933 1.840 1.933 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.897 1.899 potential_pw2rs 119 12.3 0.006 0.006 1.883 1.895 pw_transfer 1439 11.6 0.062 0.066 1.762 1.793 jit_kernel_multiply 9 16.2 0.751 1.762 0.751 1.762 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 1.692 1.734 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.673 1.706 prepare_preconditioner 11 7.9 0.000 0.000 1.550 1.577 make_preconditioner 11 8.9 0.000 0.001 1.550 1.577 make_images_data 4572 15.5 0.045 0.051 1.143 1.558 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.445 1.503 acc_transpose_blocks 27432 15.5 0.108 0.112 1.195 1.492 hybrid_alltoall_any 4725 16.4 0.050 0.111 0.990 1.452 wfi_extrapolate 11 7.9 0.001 0.001 1.440 1.440 fft3d_ps 1201 14.6 0.500 0.552 1.381 1.408 fft_wrap_pw1pw2_140 487 13.2 0.076 0.084 1.309 1.341 mp_alltoall_d11v 2130 13.8 1.183 1.319 1.183 1.319 grid_collocate_task_list 119 9.7 1.243 1.310 1.243 1.310 mp_allgather_i34 2286 14.5 0.539 1.288 0.539 1.288 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.263 1.270 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.163 1.214 mp_sum_d 4125 12.0 0.584 1.019 0.584 1.019 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.955 0.957 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.908 0.922 acc_transpose_blocks_kernels 27432 16.5 0.179 0.267 0.664 0.872 rs_pw_transfer_PW2RS_50 119 14.3 0.587 0.607 0.743 0.787 make_images_sizes 4572 15.5 0.005 0.005 0.532 0.784 mp_alltoall_i44 4572 16.5 0.528 0.780 0.528 0.780 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.226000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.363636, yerr=1.431638 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 522.825728E+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 1607811. 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 0.018 0.032 31.828 31.830 qs_mol_dyn_low 1 2.0 0.003 0.003 31.476 31.489 qs_forces 11 3.9 0.002 0.003 31.315 31.316 qs_energies 11 4.9 0.002 0.004 29.766 29.770 scf_env_do_scf 11 5.9 0.001 0.001 25.187 25.187 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.593 22.595 dbcsr_multiply_generic 2286 12.5 0.094 0.099 16.251 16.338 velocity_verlet 10 3.0 0.001 0.002 15.048 15.053 qs_scf_new_mos 108 7.5 0.001 0.001 14.768 14.790 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.767 14.789 ot_scf_mini 108 9.5 0.002 0.003 14.046 14.069 multiply_cannon 2286 13.5 0.194 0.201 12.956 13.699 multiply_cannon_loop 2286 14.5 0.636 0.664 12.174 12.979 ot_mini 108 10.5 0.001 0.001 8.699 8.719 qs_ot_get_derivative 108 11.5 0.001 0.001 7.198 7.219 multiply_cannon_multrec 18288 15.5 1.956 2.877 6.734 7.107 rebuild_ks_matrix 119 8.3 0.000 0.000 6.288 6.305 qs_ks_build_kohn_sham_matrix 119 9.3 0.022 0.054 6.287 6.304 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.560 5.578 dbcsr_mm_accdrv_process 38222 16.0 4.084 5.231 4.695 5.377 sum_up_and_integrate 119 10.3 0.030 0.031 3.883 3.889 integrate_v_rspace 119 11.3 0.002 0.002 3.853 3.862 mp_waitall_1 158411 16.6 2.584 3.817 2.584 3.817 init_scf_run 11 5.9 0.000 0.001 3.399 3.399 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.398 3.399 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.695 3.326 qs_ot_get_p 119 10.4 0.001 0.001 3.193 3.210 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.175 3.186 calculate_rho_elec 119 8.7 0.030 0.031 3.174 3.185 init_scf_loop 11 6.9 0.002 0.005 2.577 2.582 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.448 2.471 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.050 2.368 apply_single 119 13.6 0.000 0.000 2.050 2.368 rs_pw_transfer 974 11.9 0.009 0.010 2.212 2.327 calculate_first_density_matrix 1 7.0 0.001 0.002 2.199 2.200 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.115 2.121 make_m2s 4572 13.5 0.045 0.046 1.790 1.932 grid_integrate_task_list 119 12.3 1.800 1.893 1.800 1.893 density_rs2pw 119 9.7 0.004 0.004 1.777 1.884 jit_kernel_multiply 10 16.1 0.560 1.859 0.560 1.859 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.854 1.854 make_images 4572 14.5 0.191 0.204 1.705 1.845 calculate_dm_sparse 119 9.5 0.001 0.001 1.819 1.834 prepare_preconditioner 11 7.9 0.000 0.000 1.730 1.732 make_preconditioner 11 8.9 0.000 0.001 1.730 1.732 potential_pw2rs 119 12.3 0.007 0.008 1.681 1.686 pw_transfer 1439 11.6 0.063 0.066 1.670 1.681 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.584 1.665 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.649 1.651 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.624 1.625 cp_fm_redistribute_end 50 14.0 1.212 1.605 1.213 1.605 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.580 1.592 cp_fm_diag_elpa_base 50 14.0 0.377 1.522 0.391 1.573 multiply_cannon_sync_h2d 18288 15.5 1.410 1.556 1.410 1.556 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.543 1.549 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.504 1.510 ot_diis_step 108 11.5 0.011 0.011 1.479 1.479 mp_sum_l 7207 12.9 1.121 1.439 1.121 1.439 grid_collocate_task_list 119 9.7 1.203 1.292 1.203 1.292 fft3d_ps 1201 14.6 0.504 0.522 1.266 1.280 fft_wrap_pw1pw2_140 487 13.2 0.087 0.091 1.251 1.262 acc_transpose_blocks 18288 15.5 0.075 0.077 1.221 1.244 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.189 1.194 wfi_extrapolate 11 7.9 0.001 0.001 1.155 1.155 make_images_data 4572 15.5 0.044 0.047 0.798 0.957 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.937 0.939 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.876 0.900 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.690 0.864 mp_alltoall_d11v 2130 13.8 0.664 0.814 0.664 0.814 acc_transpose_blocks_kernels 18288 16.5 0.208 0.216 0.782 0.802 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.801 0.802 cp_fm_cholesky_invert 11 10.9 0.720 0.723 0.720 0.723 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.654 0.716 mp_alltoall_z22v 1201 16.6 0.608 0.673 0.608 0.673 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.830000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.363636, yerr=2.496278 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 554.700800E+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 0.024 0.045 35.021 35.023 qs_mol_dyn_low 1 2.0 0.003 0.004 34.691 34.699 qs_forces 11 3.9 0.002 0.003 34.477 34.479 qs_energies 11 4.9 0.009 0.019 32.784 32.787 scf_env_do_scf 11 5.9 0.002 0.004 27.796 27.797 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 24.355 24.355 dbcsr_multiply_generic 2286 12.5 0.101 0.102 18.120 18.292 velocity_verlet 10 3.0 0.002 0.002 17.682 17.689 qs_scf_new_mos 108 7.5 0.001 0.001 16.296 16.350 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.295 16.349 ot_scf_mini 108 9.5 0.003 0.003 15.365 15.423 multiply_cannon 2286 13.5 0.226 0.260 14.470 14.927 multiply_cannon_loop 2286 14.5 0.939 0.967 13.527 13.947 ot_mini 108 10.5 0.001 0.001 9.418 9.492 multiply_cannon_multrec 27432 15.5 2.333 3.137 8.589 9.041 qs_ot_get_derivative 108 11.5 0.001 0.001 7.616 7.672 dbcsr_mm_accdrv_process 47916 15.9 5.310 7.267 6.164 7.373 rebuild_ks_matrix 119 8.3 0.000 0.000 6.417 6.464 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.417 6.463 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.703 5.744 sum_up_and_integrate 119 10.3 0.035 0.038 3.716 3.723 integrate_v_rspace 119 11.3 0.002 0.002 3.681 3.689 init_scf_run 11 5.9 0.000 0.001 3.604 3.605 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.604 3.604 qs_ot_get_p 119 10.4 0.001 0.001 3.345 3.429 init_scf_loop 11 6.9 0.001 0.004 3.415 3.415 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.249 3.282 calculate_rho_elec 119 8.7 0.040 0.046 3.249 3.281 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.719 3.182 mp_waitall_1 137007 16.6 1.892 2.620 1.892 2.620 prepare_preconditioner 11 7.9 0.000 0.000 2.575 2.583 make_preconditioner 11 8.9 0.000 0.001 2.575 2.583 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.119 2.569 apply_single 119 13.6 0.000 0.000 2.119 2.569 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.189 2.508 make_m2s 4572 13.5 0.054 0.056 2.188 2.303 calculate_first_density_matrix 1 7.0 0.001 0.002 2.251 2.254 make_images 4572 14.5 0.272 0.332 2.080 2.193 calculate_dm_sparse 119 9.5 0.001 0.001 2.042 2.098 rs_pw_transfer 974 11.9 0.009 0.009 1.950 2.085 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.070 2.078 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.906 1.938 grid_integrate_task_list 119 12.3 1.824 1.933 1.824 1.933 density_rs2pw 119 9.7 0.004 0.004 1.742 1.882 jit_kernel_multiply 10 15.8 0.795 1.820 0.795 1.820 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.797 1.798 pw_transfer 1439 11.6 0.063 0.066 1.760 1.792 ot_diis_step 108 11.5 0.012 0.012 1.763 1.764 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.757 1.757 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.671 1.705 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.619 1.630 mp_sum_l 7207 12.9 1.043 1.552 1.043 1.552 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.517 1.518 potential_pw2rs 119 12.3 0.008 0.009 1.505 1.510 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.831 1.499 cp_fm_redistribute_end 50 14.0 1.003 1.491 1.004 1.491 cp_fm_diag_elpa_base 50 14.0 0.466 1.426 0.485 1.466 acc_transpose_blocks 27432 15.5 0.112 0.115 1.443 1.462 fft_wrap_pw1pw2_140 487 13.2 0.095 0.105 1.365 1.402 fft3d_ps 1201 14.6 0.534 0.589 1.341 1.361 wfi_extrapolate 11 7.9 0.001 0.001 1.306 1.306 grid_collocate_task_list 119 9.7 1.221 1.304 1.221 1.304 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.231 1.239 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.104 1.125 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.113 1.116 multiply_cannon_sync_h2d 27432 15.5 1.004 1.102 1.004 1.102 cp_fm_upper_to_full 72 13.5 0.785 1.089 0.785 1.089 dbcsr_complete_redistribute 329 12.2 0.121 0.145 0.784 1.050 make_images_data 4572 15.5 0.045 0.048 0.851 0.977 hybrid_alltoall_any 4725 16.4 0.061 0.151 0.741 0.911 mp_alltoall_d11v 2130 13.8 0.749 0.883 0.749 0.883 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.870 acc_transpose_blocks_kernels 27432 16.5 0.265 0.271 0.819 0.826 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.560 0.818 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.803 0.808 cp_fm_cholesky_invert 11 10.9 0.776 0.780 0.776 0.780 mp_alltoall_z22v 1201 16.6 0.689 0.719 0.689 0.719 mp_alltoall_i22 627 13.8 0.424 0.703 0.424 0.703 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.023000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.818182, yerr=3.712832 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 593.821696E+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 0.013 0.031 27.362 27.362 qs_mol_dyn_low 1 2.0 0.003 0.003 27.195 27.203 qs_forces 11 3.9 0.002 0.002 27.125 27.126 qs_energies 11 4.9 0.001 0.001 25.446 25.450 scf_env_do_scf 11 5.9 0.000 0.001 20.862 20.863 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.368 18.369 velocity_verlet 10 3.0 0.002 0.002 13.922 13.924 dbcsr_multiply_generic 2286 12.5 0.093 0.096 12.015 12.151 qs_scf_new_mos 108 7.5 0.001 0.001 10.994 11.021 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.993 11.020 ot_scf_mini 108 9.5 0.002 0.002 10.323 10.353 multiply_cannon 2286 13.5 0.230 0.236 9.533 10.035 multiply_cannon_loop 2286 14.5 0.330 0.340 8.636 8.886 multiply_cannon_multrec 9144 15.5 1.590 1.886 5.748 5.951 ot_mini 108 10.5 0.001 0.001 5.752 5.783 rebuild_ks_matrix 119 8.3 0.000 0.000 5.678 5.700 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.678 5.700 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.060 5.081 qs_ot_get_derivative 108 11.5 0.001 0.001 4.465 4.494 dbcsr_mm_accdrv_process 12550 15.8 3.127 4.080 4.058 4.138 sum_up_and_integrate 119 10.3 0.037 0.041 3.469 3.474 integrate_v_rspace 119 11.3 0.002 0.003 3.431 3.437 init_scf_run 11 5.9 0.000 0.001 3.147 3.147 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.146 3.146 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.116 3.121 calculate_rho_elec 119 8.7 0.059 0.061 3.116 3.121 qs_ot_get_p 119 10.4 0.001 0.001 2.834 2.879 init_scf_loop 11 6.9 0.000 0.000 2.471 2.472 calculate_first_density_matrix 1 7.0 0.000 0.000 2.054 2.055 mp_waitall_1 115863 16.7 1.453 1.988 1.453 1.988 grid_integrate_task_list 119 12.3 1.845 1.918 1.845 1.918 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.900 1.903 make_m2s 4572 13.5 0.034 0.035 1.711 1.861 make_images 4572 14.5 0.268 0.301 1.621 1.769 jit_kernel_multiply 10 15.7 0.893 1.751 0.893 1.751 prepare_preconditioner 11 7.9 0.000 0.000 1.736 1.741 make_preconditioner 11 8.9 0.000 0.000 1.736 1.741 calculate_dm_sparse 119 9.5 0.000 0.000 1.706 1.725 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.705 1.706 rs_pw_transfer 974 11.9 0.008 0.008 1.569 1.668 pw_transfer 1439 11.6 0.063 0.065 1.645 1.653 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.624 1.651 density_rs2pw 119 9.7 0.003 0.004 1.548 1.646 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.554 1.562 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.492 1.510 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.507 1.508 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.412 1.412 cp_fm_redistribute_end 50 14.0 0.705 1.390 0.706 1.391 cp_fm_diag_elpa_base 50 14.0 0.640 1.317 0.684 1.375 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.357 1.366 grid_collocate_task_list 119 9.7 1.272 1.355 1.272 1.355 fft_wrap_pw1pw2_140 487 13.2 0.083 0.087 1.266 1.275 ot_diis_step 108 11.5 0.012 0.013 1.275 1.275 potential_pw2rs 119 12.3 0.010 0.010 1.264 1.267 fft3d_ps 1201 14.6 0.536 0.547 1.221 1.229 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.220 1.222 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.149 1.168 apply_single 119 13.6 0.000 0.000 1.149 1.168 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.139 1.154 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.131 1.137 wfi_extrapolate 11 7.9 0.001 0.001 1.049 1.049 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.774 1.008 make_images_data 4572 15.5 0.039 0.042 0.788 0.969 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.865 0.922 acc_transpose_blocks 9144 15.5 0.038 0.039 0.888 0.897 cp_fm_cholesky_invert 11 10.9 0.875 0.877 0.875 0.877 mp_alltoall_d11v 2130 13.8 0.780 0.873 0.780 0.873 multiply_cannon_sync_h2d 9144 15.5 0.709 0.795 0.709 0.795 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.739 0.743 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.677 0.727 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.403 0.725 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.713 0.721 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 0.647 0.652 mp_allgather_i34 2286 14.5 0.232 0.628 0.232 0.628 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.277 0.598 mp_alltoall_z22v 1201 16.6 0.560 0.593 0.560 0.593 qs_create_task_list 11 7.9 0.010 0.012 0.555 0.579 generate_qs_task_list 11 8.9 0.190 0.213 0.545 0.568 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.362000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=563.454545, yerr=4.075729 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 759.889920E+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 0.025 0.095 39.986 39.986 qs_mol_dyn_low 1 2.0 0.003 0.003 39.680 39.687 qs_forces 11 3.9 0.002 0.002 39.607 39.608 qs_energies 11 4.9 0.001 0.001 37.692 37.697 scf_env_do_scf 11 5.9 0.001 0.001 32.044 32.044 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.362 24.363 velocity_verlet 10 3.0 0.002 0.002 22.543 22.549 dbcsr_multiply_generic 2286 12.5 0.100 0.101 17.295 17.453 qs_scf_new_mos 108 7.5 0.001 0.001 15.648 15.744 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.647 15.744 ot_scf_mini 108 9.5 0.002 0.002 14.583 14.682 multiply_cannon 2286 13.5 0.300 0.305 13.506 14.345 multiply_cannon_loop 2286 14.5 0.341 0.347 12.251 13.099 ot_mini 108 10.5 0.001 0.001 8.761 8.875 multiply_cannon_multrec 9144 15.5 3.351 4.667 8.494 8.550 init_scf_loop 11 6.9 0.000 0.000 7.655 7.659 rebuild_ks_matrix 119 8.3 0.000 0.000 6.705 6.845 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.704 6.845 qs_ot_get_derivative 108 11.5 0.001 0.001 6.729 6.829 prepare_preconditioner 11 7.9 0.000 0.000 6.747 6.759 make_preconditioner 11 8.9 0.000 0.000 6.747 6.759 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.332 6.639 dbcsr_mm_accdrv_process 12550 15.8 4.039 5.254 5.021 6.298 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.066 6.194 cp_fm_upper_to_full 72 14.2 3.147 4.502 3.147 4.502 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.596 3.600 calculate_rho_elec 119 8.7 0.118 0.121 3.596 3.600 sum_up_and_integrate 119 10.3 0.064 0.065 3.586 3.593 init_scf_run 11 5.9 0.000 0.001 3.574 3.574 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.574 3.574 integrate_v_rspace 119 11.3 0.003 0.003 3.522 3.530 qs_ot_get_p 119 10.4 0.001 0.001 3.108 3.238 mp_waitall_1 94719 16.7 2.306 3.233 2.306 3.233 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.447 2.858 dbcsr_complete_redistribute 329 12.2 0.288 0.292 1.956 2.760 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.187 2.472 apply_single 119 13.6 0.000 0.000 2.186 2.471 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.647 2.453 make_m2s 4572 13.5 0.038 0.038 2.220 2.402 make_images 4572 14.5 0.351 0.380 2.098 2.281 mp_alltoall_i22 627 13.8 1.442 2.251 1.442 2.251 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.410 2.210 calculate_first_density_matrix 1 7.0 0.000 0.000 2.185 2.190 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.333 2.155 calculate_dm_sparse 119 9.5 0.000 0.000 2.133 2.145 grid_integrate_task_list 119 12.3 1.994 2.010 1.994 2.010 ot_diis_step 108 11.5 0.014 0.014 2.004 2.004 pw_transfer 1439 11.6 0.066 0.066 1.978 1.979 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.907 1.907 qs_ot_p2m_diag 50 11.0 0.042 0.043 1.899 1.900 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.882 1.885 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 1.809 1.810 density_rs2pw 119 9.7 0.003 0.003 1.716 1.726 mp_sum_l 7207 12.9 1.033 1.696 1.033 1.696 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.645 1.696 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.630 1.630 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.595 1.619 fft_wrap_pw1pw2_140 487 13.2 0.088 0.090 1.583 1.586 fft3d_ps 1201 14.6 0.565 0.576 1.516 1.520 grid_collocate_task_list 119 9.7 1.438 1.452 1.438 1.452 cp_fm_cholesky_invert 11 10.9 1.424 1.427 1.424 1.427 jit_kernel_multiply 8 15.7 0.956 1.405 0.956 1.405 hybrid_alltoall_any 4725 16.4 0.087 0.146 1.099 1.358 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.342 1.343 cp_fm_diag_elpa_base 50 14.0 1.202 1.253 1.341 1.341 rs_pw_transfer 974 11.9 0.009 0.009 1.304 1.328 wfi_extrapolate 11 7.9 0.001 0.001 1.326 1.326 make_images_data 4572 15.5 0.043 0.046 1.043 1.272 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.216 1.220 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.147 1.163 potential_pw2rs 119 12.3 0.014 0.014 1.153 1.154 mp_alltoall_d11v 2130 13.8 1.138 1.148 1.138 1.148 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.120 1.142 multiply_cannon_sync_h2d 9144 15.5 1.041 1.045 1.041 1.045 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.979 1.024 qs_create_task_list 11 7.9 0.029 0.047 0.985 0.999 generate_qs_task_list 11 8.9 0.372 0.392 0.956 0.977 acc_transpose_blocks 9144 15.5 0.038 0.038 0.891 0.898 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.849 0.862 mp_alltoall_z22v 1201 16.6 0.816 0.835 0.816 0.835 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=39.986000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=706.272727, yerr=19.945794 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 501.284864E+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 1462452. 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 0.048 0.074 82.566 82.568 qs_mol_dyn_low 1 2.0 0.003 0.006 82.053 82.064 qs_forces 11 3.9 0.002 0.002 81.941 81.942 qs_energies 11 4.9 0.001 0.002 79.172 79.181 scf_env_do_scf 11 5.9 0.000 0.001 70.261 70.264 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.896 64.896 dbcsr_multiply_generic 2055 12.4 0.112 0.118 52.337 52.602 qs_scf_new_mos 99 7.5 0.000 0.001 48.263 48.374 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.262 48.374 ot_scf_mini 99 9.5 0.002 0.002 45.813 45.916 multiply_cannon 2055 13.4 0.174 0.183 42.704 43.554 multiply_cannon_loop 2055 14.4 1.553 1.592 41.781 42.645 velocity_verlet 10 3.0 0.001 0.002 41.468 41.469 ot_mini 99 10.5 0.001 0.001 27.588 27.690 qs_ot_get_derivative 99 11.5 0.001 0.001 20.730 20.831 multiply_cannon_multrec 49320 15.4 12.107 12.878 17.128 17.807 rebuild_ks_matrix 110 8.3 0.000 0.001 14.229 14.415 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.229 14.414 mp_waitall_1 220248 16.4 11.685 12.955 11.685 12.955 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.512 12.677 multiply_cannon_sync_h2d 49320 15.4 9.972 10.603 9.972 10.603 qs_ot_get_p 110 10.4 0.001 0.001 9.595 9.711 multiply_cannon_metrocomm3 49320 15.4 0.078 0.084 7.191 8.689 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.158 7.623 apply_single 110 13.6 0.000 0.001 7.158 7.623 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.025 7.555 init_scf_run 11 5.9 0.000 0.001 6.922 6.922 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.922 6.922 sum_up_and_integrate 110 10.3 0.036 0.042 6.870 6.883 integrate_v_rspace 110 11.3 0.002 0.003 6.834 6.855 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.704 6.754 ot_diis_step 99 11.5 0.005 0.006 6.544 6.544 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.298 6.331 mp_sum_l 6514 12.8 5.405 6.183 5.405 6.183 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.014 6.149 calculate_rho_elec 110 8.6 0.020 0.024 6.014 6.148 init_scf_loop 11 6.9 0.000 0.000 5.339 5.339 cp_dbcsr_syevd 48 12.0 0.002 0.002 5.302 5.302 dbcsr_mm_accdrv_process 87628 16.1 2.077 2.181 4.901 5.229 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.847 4.848 cp_fm_redistribute_end 48 14.0 4.232 4.824 4.236 4.825 cp_fm_diag_elpa_base 48 14.0 0.583 4.682 0.587 4.706 calculate_dm_sparse 110 9.5 0.000 0.001 3.931 4.033 multiply_cannon_metrocomm1 49320 15.4 0.060 0.063 2.705 4.011 wfi_extrapolate 11 7.9 0.001 0.001 3.951 3.951 make_m2s 4110 13.4 0.061 0.067 3.773 3.885 make_images 4110 14.4 0.177 0.191 3.676 3.793 rs_pw_transfer 902 11.9 0.012 0.013 3.582 3.753 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.678 3.681 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.387 3.462 grid_integrate_task_list 110 12.3 3.248 3.419 3.248 3.419 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.284 3.357 density_rs2pw 110 9.6 0.004 0.004 3.096 3.280 prepare_preconditioner 11 7.9 0.000 0.000 3.223 3.243 make_preconditioner 11 8.9 0.000 0.000 3.223 3.243 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.021 3.067 calculate_first_density_matrix 1 7.0 0.000 0.001 2.886 2.892 pw_transfer 1331 11.6 0.053 0.064 2.702 2.755 jit_kernel_multiply 13 15.9 2.541 2.678 2.541 2.678 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.615 2.671 mp_alltoall_d11v 2046 13.8 2.054 2.619 2.054 2.619 potential_pw2rs 110 12.3 0.006 0.007 2.478 2.497 fft_wrap_pw1pw2_140 451 13.1 0.170 0.190 2.224 2.283 acc_transpose_blocks 49320 15.4 0.223 0.235 2.160 2.262 grid_collocate_task_list 110 9.6 2.087 2.201 2.087 2.201 fft3d_ps 1111 14.6 0.753 0.839 2.151 2.194 mp_waitany 14300 13.8 1.804 2.045 1.804 2.045 mp_sum_d 3879 11.9 1.489 1.973 1.489 1.973 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.855 1.879 make_images_data 4110 15.4 0.042 0.045 1.721 1.854 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.749 1.777 hybrid_alltoall_any 4261 16.3 0.081 0.479 1.500 1.763 cp_fm_cholesky_invert 11 10.9 1.745 1.749 1.745 1.749 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.568000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.818182, yerr=2.036851 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 585.773056E+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 2154033. 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 0.022 0.044 68.301 68.302 qs_mol_dyn_low 1 2.0 0.003 0.005 67.946 67.956 qs_forces 11 3.9 0.002 0.002 67.877 67.879 qs_energies 11 4.9 0.002 0.005 64.585 64.590 scf_env_do_scf 11 5.9 0.000 0.001 56.166 56.170 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.544 48.544 dbcsr_multiply_generic 2055 12.4 0.114 0.118 37.491 37.680 velocity_verlet 10 3.0 0.002 0.002 35.895 35.896 qs_scf_new_mos 99 7.5 0.001 0.001 32.731 32.848 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.731 32.848 multiply_cannon 2055 13.4 0.222 0.242 30.900 31.925 ot_scf_mini 99 9.5 0.003 0.003 31.077 31.193 multiply_cannon_loop 2055 14.4 0.924 0.944 29.607 30.366 ot_mini 99 10.5 0.001 0.001 18.136 18.257 multiply_cannon_multrec 24660 15.4 7.629 10.209 13.795 16.316 rebuild_ks_matrix 110 8.3 0.000 0.001 13.597 13.770 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.596 13.770 qs_ot_get_derivative 99 11.5 0.001 0.001 12.317 12.437 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.991 12.146 mp_waitall_1 176588 16.5 7.624 10.040 7.624 10.040 multiply_cannon_sync_h2d 24660 15.4 7.037 8.311 7.037 8.311 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 5.133 7.670 init_scf_loop 11 6.9 0.000 0.000 7.583 7.584 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.545 7.286 apply_single 110 13.6 0.000 0.001 6.544 7.286 sum_up_and_integrate 110 10.3 0.053 0.060 6.517 6.529 integrate_v_rspace 110 11.3 0.002 0.002 6.464 6.479 qs_ot_get_p 110 10.4 0.001 0.001 6.304 6.444 dbcsr_mm_accdrv_process 52282 16.1 4.663 5.404 6.007 6.316 init_scf_run 11 5.9 0.000 0.001 6.058 6.058 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.057 6.058 ot_diis_step 99 11.5 0.010 0.010 5.772 5.772 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.743 5.752 calculate_rho_elec 110 8.6 0.039 0.047 5.743 5.752 prepare_preconditioner 11 7.9 0.000 0.000 5.570 5.589 make_preconditioner 11 8.9 0.000 0.000 5.570 5.589 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.644 5.375 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.169 5.327 make_m2s 4110 13.4 0.057 0.061 4.202 4.672 make_images 4110 14.4 0.399 0.445 4.092 4.558 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.365 4.386 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.940 3.940 wfi_extrapolate 11 7.9 0.001 0.001 3.488 3.488 pw_transfer 1331 11.6 0.065 0.070 3.274 3.421 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.404 3.406 cp_fm_redistribute_end 48 14.0 2.542 3.375 2.544 3.375 cp_fm_diag_elpa_base 48 14.0 0.796 3.229 0.827 3.319 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.168 3.316 grid_integrate_task_list 110 12.3 3.159 3.300 3.159 3.300 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.166 3.229 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.217 3.220 density_rs2pw 110 9.6 0.004 0.004 3.031 3.204 rs_pw_transfer 902 11.9 0.012 0.013 2.923 3.132 calculate_dm_sparse 110 9.5 0.001 0.001 2.950 2.976 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.877 2.936 make_images_data 4110 15.4 0.046 0.050 2.372 2.842 fft_wrap_pw1pw2_140 451 13.1 0.201 0.221 2.661 2.807 hybrid_alltoall_any 4261 16.3 0.101 0.443 2.080 2.794 fft3d_ps 1111 14.6 1.073 1.261 2.524 2.659 cp_fm_cholesky_invert 11 10.9 2.601 2.609 2.601 2.609 calculate_first_density_matrix 1 7.0 0.000 0.002 2.486 2.489 mp_sum_l 6514 12.8 1.769 2.486 1.769 2.486 potential_pw2rs 110 12.3 0.008 0.008 2.369 2.383 grid_collocate_task_list 110 9.6 2.122 2.268 2.122 2.268 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.952 1.966 jit_kernel_multiply 10 16.4 0.996 1.942 0.996 1.942 mp_alltoall_d11v 2046 13.8 1.760 1.937 1.760 1.937 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.790 1.898 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.844 1.846 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.768 1.784 mp_irecv_dv 57340 16.2 0.667 1.682 0.667 1.682 mp_allgather_i34 2055 14.4 0.623 1.669 0.623 1.669 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.533 1.546 acc_transpose_blocks 24660 15.4 0.110 0.113 1.507 1.523 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.367 1.479 dbcsr_complete_redistribute 325 12.2 0.242 0.298 1.150 1.411 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.302000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.181818, yerr=6.739547 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 656.326656E+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 0.018 0.033 59.004 59.005 qs_mol_dyn_low 1 2.0 0.003 0.004 58.594 58.604 qs_forces 11 3.9 0.002 0.004 58.527 58.528 qs_energies 11 4.9 0.002 0.006 55.357 55.360 scf_env_do_scf 11 5.9 0.001 0.001 47.535 47.535 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.111 39.112 velocity_verlet 10 3.0 0.001 0.002 32.149 32.150 dbcsr_multiply_generic 2055 12.4 0.108 0.111 28.306 28.597 qs_scf_new_mos 99 7.5 0.001 0.001 24.767 24.886 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.767 24.885 ot_scf_mini 99 9.5 0.002 0.003 23.548 23.671 multiply_cannon 2055 13.4 0.213 0.221 22.070 23.296 multiply_cannon_loop 2055 14.4 0.615 0.630 20.849 22.067 ot_mini 99 10.5 0.001 0.001 13.516 13.650 rebuild_ks_matrix 110 8.3 0.000 0.000 12.076 12.247 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 12.076 12.247 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.652 10.807 multiply_cannon_multrec 16440 15.4 3.810 4.814 9.531 10.402 mp_waitall_1 139946 16.5 7.111 10.358 7.111 10.358 qs_ot_get_derivative 99 11.5 0.001 0.001 8.995 9.121 init_scf_loop 11 6.9 0.001 0.003 8.388 8.390 multiply_cannon_metrocomm3 16440 15.4 0.043 0.045 4.492 7.544 prepare_preconditioner 11 7.9 0.000 0.000 6.646 6.666 make_preconditioner 11 8.9 0.000 0.001 6.646 6.666 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.021 6.380 sum_up_and_integrate 110 10.3 0.061 0.062 6.336 6.352 integrate_v_rspace 110 11.3 0.002 0.002 6.275 6.292 dbcsr_mm_accdrv_process 34862 16.1 4.566 5.250 5.576 5.722 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.039 5.531 apply_single 110 13.6 0.000 0.000 5.039 5.531 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.446 5.455 calculate_rho_elec 110 8.6 0.058 0.058 5.446 5.455 init_scf_run 11 5.9 0.000 0.001 5.422 5.422 scf_env_initial_rho_setup 11 6.9 0.000 0.002 5.422 5.422 qs_ot_get_p 110 10.4 0.001 0.001 5.224 5.391 make_m2s 4110 13.4 0.050 0.052 4.129 4.522 ot_diis_step 99 11.5 0.010 0.011 4.491 4.491 make_images 4110 14.4 0.392 0.510 4.013 4.405 multiply_cannon_sync_h2d 16440 15.4 3.648 4.168 3.648 4.168 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.075 3.720 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.629 3.632 grid_integrate_task_list 110 12.3 3.180 3.378 3.180 3.378 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.294 3.295 pw_transfer 1331 11.6 0.064 0.071 3.137 3.142 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.031 3.038 density_rs2pw 110 9.6 0.004 0.004 2.741 3.010 wfi_extrapolate 11 7.9 0.001 0.001 2.940 2.940 make_images_data 4110 15.4 0.043 0.048 2.409 2.904 rs_pw_transfer 902 11.9 0.011 0.011 2.588 2.880 hybrid_alltoall_any 4261 16.3 0.105 0.375 2.115 2.809 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.778 2.780 cp_fm_redistribute_end 48 14.0 1.735 2.755 1.736 2.755 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.753 2.755 cp_fm_diag_elpa_base 48 14.0 0.958 2.614 1.015 2.714 cp_fm_cholesky_invert 11 10.9 2.638 2.644 2.638 2.644 fft_wrap_pw1pw2_140 451 13.1 0.210 0.215 2.592 2.600 calculate_dm_sparse 110 9.5 0.001 0.001 2.495 2.528 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.896 2.514 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.422 2.484 calculate_first_density_matrix 1 7.0 0.001 0.006 2.402 2.403 mp_irecv_dv 48980 15.7 0.826 2.391 0.826 2.391 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.302 2.367 grid_collocate_task_list 110 9.6 2.177 2.366 2.177 2.366 fft3d_ps 1111 14.6 1.065 1.076 2.346 2.359 mp_sum_l 6514 12.8 1.506 2.194 1.506 2.194 potential_pw2rs 110 12.3 0.011 0.011 2.169 2.174 mp_alltoall_d11v 2046 13.8 1.696 1.963 1.696 1.963 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.960 1.962 dbcsr_complete_redistribute 325 12.2 0.325 0.353 1.413 1.858 cp_fm_upper_to_full 70 13.6 1.373 1.830 1.373 1.830 mp_allgather_i34 2055 14.4 0.548 1.650 0.548 1.650 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.623 1.641 cp_fm_cholesky_decompose 22 10.9 1.558 1.576 1.558 1.576 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.378 1.501 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.460 1.472 mp_waitany 17072 13.8 1.114 1.384 1.114 1.384 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.927 1.367 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.256 1.262 acc_transpose_blocks 16440 15.4 0.070 0.071 1.194 1.205 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.005000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=621.090909, yerr=8.229156 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 728.014848E+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 0.048 0.071 63.566 63.567 qs_mol_dyn_low 1 2.0 0.003 0.004 63.210 63.219 qs_forces 11 3.9 0.002 0.003 63.142 63.143 qs_energies 11 4.9 0.001 0.004 59.799 59.803 scf_env_do_scf 11 5.9 0.001 0.001 51.646 51.649 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.256 40.256 velocity_verlet 10 3.0 0.003 0.003 36.054 36.057 dbcsr_multiply_generic 2055 12.4 0.115 0.118 29.340 29.535 qs_scf_new_mos 99 7.5 0.001 0.001 26.070 26.172 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.070 26.172 ot_scf_mini 99 9.5 0.003 0.003 24.455 24.551 multiply_cannon 2055 13.4 0.242 0.258 22.438 23.399 multiply_cannon_loop 2055 14.4 0.885 0.904 21.103 21.672 ot_mini 99 10.5 0.001 0.001 14.069 14.191 multiply_cannon_multrec 24660 15.4 4.209 6.791 12.611 13.874 rebuild_ks_matrix 110 8.3 0.000 0.000 11.785 11.896 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.784 11.896 init_scf_loop 11 6.9 0.000 0.001 11.350 11.350 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.438 10.540 qs_ot_get_derivative 99 11.5 0.001 0.001 9.937 10.039 prepare_preconditioner 11 7.9 0.000 0.000 9.643 9.660 make_preconditioner 11 8.9 0.000 0.001 9.643 9.660 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.906 9.343 dbcsr_mm_accdrv_process 52304 16.0 6.807 8.272 8.256 9.219 mp_waitall_1 121746 16.5 4.387 6.448 4.387 6.448 sum_up_and_integrate 110 10.3 0.067 0.071 6.223 6.242 integrate_v_rspace 110 11.3 0.002 0.003 6.156 6.173 make_m2s 4110 13.4 0.060 0.062 5.327 5.603 qs_ot_get_p 110 10.4 0.001 0.001 5.404 5.558 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.504 5.511 calculate_rho_elec 110 8.6 0.077 0.080 5.503 5.511 make_images 4110 14.4 0.574 0.697 5.185 5.458 init_scf_run 11 5.9 0.000 0.001 5.404 5.404 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.404 5.404 cp_fm_upper_to_full 70 13.8 3.228 4.588 3.228 4.588 ot_diis_step 99 11.5 0.011 0.011 4.097 4.097 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.018 4.091 apply_single 110 13.6 0.000 0.000 4.017 4.090 dbcsr_complete_redistribute 325 12.2 0.426 0.478 2.598 3.718 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.669 3.684 grid_integrate_task_list 110 12.3 3.284 3.441 3.284 3.441 multiply_cannon_sync_h2d 24660 15.4 3.205 3.388 3.205 3.388 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.225 3.277 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.201 3.201 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.081 3.186 pw_transfer 1331 11.6 0.064 0.071 3.163 3.185 multiply_cannon_metrocomm3 24660 15.4 0.036 0.036 1.466 3.179 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.058 3.082 make_images_data 4110 15.4 0.046 0.050 2.709 3.039 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.972 3.030 hybrid_alltoall_any 4261 16.3 0.119 0.461 2.294 2.953 calculate_dm_sparse 110 9.5 0.001 0.001 2.901 2.927 wfi_extrapolate 11 7.9 0.001 0.001 2.922 2.923 density_rs2pw 110 9.6 0.004 0.004 2.679 2.822 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.727 2.818 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.771 2.772 mp_alltoall_i22 605 13.7 1.619 2.762 1.619 2.762 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.678 2.680 cp_fm_redistribute_end 48 14.0 1.340 2.652 1.341 2.653 fft_wrap_pw1pw2_140 451 13.1 0.202 0.214 2.610 2.638 cp_fm_diag_elpa_base 48 14.0 1.232 2.520 1.309 2.626 cp_fm_cholesky_invert 11 10.9 2.607 2.615 2.607 2.615 rs_pw_transfer 902 11.9 0.010 0.011 2.334 2.496 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.406 2.446 calculate_first_density_matrix 1 7.0 0.000 0.002 2.383 2.387 grid_collocate_task_list 110 9.6 2.226 2.375 2.226 2.375 fft3d_ps 1111 14.6 1.062 1.094 2.361 2.374 qs_energies_init_hamiltonians 11 5.9 0.002 0.008 2.261 2.262 jit_kernel_multiply 12 15.8 1.120 2.194 1.120 2.194 potential_pw2rs 110 12.3 0.012 0.012 2.008 2.016 mp_alltoall_d11v 2046 13.8 1.713 1.878 1.713 1.878 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.683 1.719 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.598 1.709 cp_fm_cholesky_decompose 22 10.9 1.616 1.670 1.616 1.670 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.591 1.603 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.521 1.536 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.858 1.507 acc_transpose_blocks 24660 15.4 0.105 0.107 1.491 1.506 mp_sum_l 6514 12.8 0.894 1.503 0.894 1.503 mp_irecv_dv 62702 16.1 0.759 1.432 0.759 1.432 qs_env_update_s_mstruct 11 6.9 0.001 0.007 1.214 1.316 mp_waitany 13376 13.8 1.054 1.303 1.054 1.303 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=63.567000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=688.363636, yerr=8.657868 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 818.950144E+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 0.014 0.031 53.866 53.867 qs_mol_dyn_low 1 2.0 0.003 0.003 53.635 53.646 qs_forces 11 3.9 0.002 0.002 53.569 53.570 qs_energies 11 4.9 0.003 0.004 49.973 49.978 scf_env_do_scf 11 5.9 0.000 0.001 41.861 41.861 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 34.217 34.218 velocity_verlet 10 3.0 0.002 0.002 30.302 30.305 dbcsr_multiply_generic 2055 12.4 0.105 0.107 22.660 22.841 qs_scf_new_mos 99 7.5 0.001 0.001 20.231 20.276 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.230 20.276 ot_scf_mini 99 9.5 0.002 0.002 19.004 19.029 multiply_cannon 2055 13.4 0.246 0.259 17.184 18.739 multiply_cannon_loop 2055 14.4 0.320 0.332 15.812 16.226 rebuild_ks_matrix 110 8.3 0.000 0.000 11.410 11.445 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.410 11.444 ot_mini 99 10.5 0.001 0.001 10.281 10.301 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.170 10.198 multiply_cannon_multrec 8220 15.4 3.211 4.401 7.408 8.358 mp_waitall_1 103326 16.6 6.001 7.737 6.001 7.737 init_scf_loop 11 6.9 0.000 0.000 7.596 7.597 qs_ot_get_derivative 99 11.5 0.001 0.001 6.500 6.523 sum_up_and_integrate 110 10.3 0.080 0.082 6.120 6.134 integrate_v_rspace 110 11.3 0.003 0.003 6.040 6.055 prepare_preconditioner 11 7.9 0.000 0.000 5.957 5.961 make_preconditioner 11 8.9 0.000 0.000 5.957 5.961 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.541 5.613 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.529 5.540 calculate_rho_elec 110 8.6 0.115 0.115 5.529 5.539 dbcsr_mm_accdrv_process 17442 15.9 2.804 4.309 4.068 4.966 init_scf_run 11 5.9 0.000 0.001 4.953 4.953 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.953 4.953 qs_ot_get_p 110 10.4 0.001 0.001 4.787 4.814 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 2.988 4.542 make_m2s 4110 13.4 0.039 0.041 4.224 4.506 make_images 4110 14.4 0.636 0.694 4.093 4.376 ot_diis_step 99 11.5 0.012 0.012 3.761 3.762 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.679 3.700 apply_single 110 13.6 0.000 0.000 3.678 3.700 grid_integrate_task_list 110 12.3 3.366 3.509 3.366 3.509 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.460 3.464 pw_transfer 1331 11.6 0.064 0.070 3.193 3.211 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.164 3.164 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.086 3.108 multiply_cannon_sync_h2d 8220 15.4 2.918 3.048 2.918 3.048 make_images_data 4110 15.4 0.038 0.043 2.441 2.900 cp_fm_cholesky_invert 11 10.9 2.864 2.867 2.864 2.867 hybrid_alltoall_any 4261 16.3 0.200 0.866 2.332 2.793 density_rs2pw 110 9.6 0.004 0.004 2.581 2.736 qs_energies_init_hamiltonians 11 5.9 0.063 0.067 2.705 2.705 fft_wrap_pw1pw2_140 451 13.1 0.213 0.218 2.657 2.682 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.661 2.662 wfi_extrapolate 11 7.9 0.001 0.001 2.655 2.655 cp_fm_redistribute_end 48 14.0 0.677 2.635 0.680 2.635 cp_fm_diag_elpa_base 48 14.0 1.770 2.443 1.947 2.607 grid_collocate_task_list 110 9.6 2.328 2.503 2.328 2.503 calculate_dm_sparse 110 9.5 0.001 0.001 2.452 2.492 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.471 2.473 fft3d_ps 1111 14.6 1.122 1.167 2.345 2.369 rs_pw_transfer 902 11.9 0.010 0.010 2.100 2.236 calculate_first_density_matrix 1 7.0 0.000 0.000 2.213 2.214 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.114 2.131 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.769 1.987 potential_pw2rs 110 12.3 0.015 0.015 1.880 1.885 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.866 1.883 mp_alltoall_d11v 2046 13.8 1.599 1.807 1.599 1.807 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.759 1.770 jit_kernel_multiply 10 15.7 0.958 1.737 0.958 1.737 mp_allgather_i34 2055 14.4 0.558 1.706 0.558 1.706 cp_fm_cholesky_decompose 22 10.9 1.661 1.676 1.661 1.676 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.489 1.607 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.571 1.576 dbcsr_complete_redistribute 325 12.2 0.565 0.596 1.432 1.527 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.424 1.436 qs_create_task_list 11 7.9 0.000 0.001 1.221 1.321 generate_qs_task_list 11 8.9 0.379 0.446 1.221 1.321 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.738 1.178 mp_waitany 9240 13.8 1.032 1.149 1.032 1.149 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.112 1.137 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.483 1.137 mp_irecv_dv 24056 15.7 0.459 1.095 0.459 1.095 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=53.867000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=772.363636, yerr=10.619645 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 1.343042E+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 0.018 0.037 85.165 85.165 qs_mol_dyn_low 1 2.0 0.003 0.003 84.814 84.823 qs_forces 11 3.9 0.002 0.003 84.745 84.746 qs_energies 11 4.9 0.001 0.001 80.661 80.663 scf_env_do_scf 11 5.9 0.001 0.001 70.861 70.861 velocity_verlet 10 3.0 0.003 0.003 54.450 54.457 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.608 42.609 dbcsr_multiply_generic 2055 12.4 0.119 0.122 28.441 28.495 init_scf_loop 11 6.9 0.000 0.000 28.181 28.182 prepare_preconditioner 11 7.9 0.000 0.000 26.236 26.242 make_preconditioner 11 8.9 0.000 0.000 26.236 26.242 qs_scf_new_mos 99 7.5 0.001 0.001 25.941 25.981 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.940 25.980 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.521 25.708 ot_scf_mini 99 9.5 0.002 0.002 24.187 24.218 multiply_cannon 2055 13.4 0.342 0.365 21.380 22.091 multiply_cannon_loop 2055 14.4 0.339 0.342 19.577 19.984 cp_fm_upper_to_full 70 14.2 12.702 18.130 12.702 18.130 ot_mini 99 10.5 0.001 0.001 13.477 13.505 rebuild_ks_matrix 110 8.3 0.001 0.001 13.058 13.107 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.058 13.106 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.878 11.922 dbcsr_complete_redistribute 325 12.2 1.025 1.044 7.303 10.424 multiply_cannon_multrec 8220 15.4 4.357 4.560 9.544 9.659 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.266 9.392 qs_ot_get_derivative 99 11.5 0.001 0.001 8.877 8.908 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.699 8.800 mp_alltoall_i22 605 13.7 5.314 8.452 5.314 8.452 mp_waitall_1 84994 16.7 7.521 8.352 7.521 8.352 sum_up_and_integrate 110 10.3 0.150 0.151 6.569 6.585 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.400 6.437 calculate_rho_elec 110 8.6 0.227 0.227 6.400 6.437 integrate_v_rspace 110 11.3 0.003 0.003 6.419 6.434 make_m2s 4110 13.4 0.043 0.044 5.336 5.865 make_images 4110 14.4 0.872 0.918 5.147 5.676 init_scf_run 11 5.9 0.000 0.001 5.591 5.591 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.590 5.590 qs_ot_get_p 110 10.4 0.001 0.001 5.517 5.547 cp_fm_cholesky_invert 11 10.9 5.319 5.323 5.319 5.323 dbcsr_mm_accdrv_process 11614 15.7 3.259 3.669 5.044 5.313 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.722 5.182 apply_single 110 13.6 0.000 0.000 4.721 5.182 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.661 5.070 ot_diis_step 99 11.5 0.015 0.015 4.577 4.577 qs_ot_p2m_diag 48 11.0 0.150 0.155 4.011 4.018 multiply_cannon_sync_h2d 8220 15.4 3.954 3.957 3.954 3.957 hybrid_alltoall_any 4261 16.3 0.254 0.550 2.996 3.793 grid_integrate_task_list 110 12.3 3.654 3.714 3.654 3.714 pw_transfer 1331 11.6 0.073 0.074 3.699 3.706 make_images_data 4110 15.4 0.041 0.044 2.994 3.659 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.624 3.625 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.583 3.589 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.573 3.574 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.834 3.295 wfi_extrapolate 11 7.9 0.001 0.001 3.188 3.188 calculate_dm_sparse 110 9.5 0.001 0.001 3.166 3.184 fft_wrap_pw1pw2_140 451 13.1 0.215 0.218 3.114 3.121 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.983 2.983 cp_fm_diag_elpa_base 48 14.0 2.452 2.644 2.981 2.981 density_rs2pw 110 9.6 0.004 0.004 2.906 2.925 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.858 2.864 fft3d_ps 1111 14.6 1.264 1.273 2.792 2.797 grid_collocate_task_list 110 9.6 2.624 2.650 2.624 2.650 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.369 2.387 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.357 2.373 calculate_first_density_matrix 1 7.0 0.000 0.000 2.299 2.299 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.099 2.236 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.163 2.221 rs_pw_transfer 902 11.9 0.011 0.011 2.074 2.123 mp_alltoall_d11v 2046 13.8 2.010 2.059 2.010 2.059 potential_pw2rs 110 12.3 0.021 0.021 2.025 2.031 cp_fm_cholesky_decompose 22 10.9 1.952 1.965 1.952 1.965 qs_create_task_list 11 7.9 0.001 0.001 1.897 1.941 generate_qs_task_list 11 8.9 0.738 0.791 1.897 1.941 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.767 1.797 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.785 1.789 jit_kernel_multiply 10 15.3 1.586 1.754 1.586 1.754 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.705 1.720 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=85.165000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1204.272727, yerr=64.610850 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 628.781056E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 2268968. 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 0.040 0.068 202.072 202.073 qs_mol_dyn_low 1 2.0 0.003 0.004 201.510 201.523 qs_forces 11 3.9 0.003 0.009 201.420 201.424 qs_energies 11 4.9 0.002 0.008 195.936 195.951 scf_env_do_scf 11 5.9 0.001 0.002 179.549 179.553 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 159.004 159.006 dbcsr_multiply_generic 2507 12.6 0.177 0.183 123.876 124.689 velocity_verlet 10 3.0 0.001 0.002 121.728 121.729 qs_scf_new_mos 117 7.6 0.001 0.001 120.586 120.845 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.585 120.844 ot_scf_mini 117 9.6 0.003 0.004 114.027 114.274 multiply_cannon 2507 13.6 0.235 0.248 100.825 102.891 multiply_cannon_loop 2507 14.6 2.083 2.191 98.611 100.565 ot_mini 117 10.6 0.001 0.001 65.511 65.758 multiply_cannon_multrec 60168 15.6 33.501 36.194 41.690 44.017 qs_ot_get_derivative 117 11.6 0.001 0.001 40.602 40.854 rebuild_ks_matrix 128 8.3 0.001 0.001 33.152 33.610 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.027 33.151 33.609 mp_waitall_1 267128 16.5 27.932 31.135 27.932 31.135 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.820 30.240 multiply_cannon_sync_h2d 60168 15.6 27.558 29.712 27.558 29.712 qs_ot_get_p 128 10.4 0.001 0.001 26.703 26.980 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.354 25.386 apply_single 128 13.6 0.001 0.001 24.354 25.386 ot_diis_step 117 11.6 0.007 0.008 24.551 24.552 init_scf_loop 11 6.9 0.001 0.006 20.470 20.472 qs_ot_p2m_diag 83 11.4 0.078 0.091 20.033 20.091 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.778 18.977 multiply_cannon_metrocomm3 60168 15.6 0.115 0.121 15.640 18.134 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.538 17.539 prepare_preconditioner 11 7.9 0.000 0.000 15.848 15.886 make_preconditioner 11 8.9 0.000 0.002 15.848 15.886 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.079 15.275 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.731 14.736 cp_fm_redistribute_end 83 14.4 11.646 14.662 11.660 14.667 cp_fm_diag_elpa_base 83 14.4 2.962 14.287 2.994 14.414 make_m2s 5014 13.6 0.104 0.111 13.588 13.986 sum_up_and_integrate 128 10.3 0.089 0.106 13.824 13.840 make_images 5014 14.6 0.399 0.418 13.407 13.814 integrate_v_rspace 128 11.3 0.003 0.004 13.734 13.751 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.430 12.483 calculate_rho_elec 128 8.7 0.045 0.063 12.430 12.482 init_scf_run 11 5.9 0.000 0.001 12.231 12.231 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.230 12.231 mp_sum_l 7870 13.0 8.207 9.435 8.207 9.435 cp_fm_cholesky_invert 11 10.9 9.085 9.093 9.085 9.093 wfi_extrapolate 11 7.9 0.001 0.001 9.048 9.048 dbcsr_mm_accdrv_process 124484 16.2 3.171 3.486 7.750 8.479 calculate_dm_sparse 128 9.5 0.001 0.002 8.359 8.438 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.048 8.183 multiply_cannon_metrocomm1 60168 15.6 0.089 0.094 5.864 7.884 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.779 7.863 make_images_data 5014 15.6 0.066 0.072 6.708 7.636 grid_integrate_task_list 128 12.3 7.058 7.485 7.058 7.485 hybrid_alltoall_any 5200 16.5 0.290 2.266 5.842 7.139 pw_transfer 1547 11.6 0.074 0.105 6.659 6.943 density_rs2pw 128 9.7 0.006 0.007 6.254 6.857 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.457 6.716 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.635 6.644 rs_pw_transfer 1046 11.9 0.017 0.019 5.477 6.092 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.796 5.903 fft_wrap_pw1pw2_140 523 13.2 0.446 0.515 5.604 5.766 fft3d_ps 1291 14.7 2.099 2.584 5.251 5.467 mp_alltoall_d11v 2415 14.1 4.175 5.314 4.175 5.314 grid_collocate_task_list 128 9.7 4.721 5.075 4.721 5.075 cp_fm_cholesky_decompose 22 10.9 4.585 4.598 4.585 4.598 mp_sum_d 4455 12.2 3.595 4.563 3.595 4.563 potential_pw2rs 128 12.3 0.009 0.011 4.329 4.352 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=202.073000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.090909, yerr=6.788420 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 827.084800E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57203. MP_Allreduce 11252 946. MP_Sync 170 MP_Alltoall 1969 5154435. 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 0.021 0.051 188.055 188.059 qs_mol_dyn_low 1 2.0 0.003 0.006 187.616 187.630 qs_forces 11 3.9 0.003 0.003 187.517 187.519 qs_energies 11 4.9 0.002 0.005 180.845 180.857 scf_env_do_scf 11 5.9 0.001 0.001 164.535 164.545 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 131.891 131.893 velocity_verlet 10 3.0 0.001 0.002 118.144 118.145 dbcsr_multiply_generic 2507 12.6 0.186 0.192 96.879 98.122 qs_scf_new_mos 117 7.6 0.001 0.002 93.683 94.181 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.682 94.181 ot_scf_mini 117 9.6 0.004 0.005 88.886 89.460 multiply_cannon 2507 13.6 0.480 0.534 76.968 81.635 multiply_cannon_loop 2507 14.6 1.250 1.288 73.852 76.311 ot_mini 117 10.6 0.001 0.001 49.642 50.155 mp_waitall_1 214728 16.6 24.847 38.496 24.847 38.496 multiply_cannon_multrec 30084 15.6 21.674 25.648 31.254 35.636 rebuild_ks_matrix 128 8.3 0.001 0.001 32.458 33.031 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 32.458 33.030 init_scf_loop 11 6.9 0.003 0.021 32.555 32.556 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.246 29.770 multiply_cannon_metrocomm3 30084 15.6 0.094 0.100 16.120 28.419 prepare_preconditioner 11 7.9 0.000 0.000 28.187 28.257 make_preconditioner 11 8.9 0.001 0.009 28.187 28.257 qs_ot_get_derivative 117 11.6 0.001 0.002 27.629 28.205 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.851 27.372 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.108 23.320 apply_single 128 13.6 0.001 0.001 22.107 23.320 ot_diis_step 117 11.6 0.014 0.015 21.835 21.837 qs_ot_get_p 128 10.4 0.001 0.001 21.119 21.765 multiply_cannon_sync_h2d 30084 15.6 18.980 21.402 18.980 21.402 qs_ot_p2m_diag 83 11.4 0.187 0.216 16.405 16.438 cp_fm_cholesky_invert 11 10.9 16.410 16.423 16.410 16.423 make_m2s 5014 13.6 0.090 0.096 14.213 15.651 make_images 5014 14.6 1.152 1.343 14.002 15.443 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.263 15.264 sum_up_and_integrate 128 10.3 0.116 0.133 14.219 14.246 integrate_v_rspace 128 11.3 0.003 0.003 14.104 14.133 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.788 12.822 calculate_rho_elec 128 8.7 0.087 0.105 12.788 12.821 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.212 12.224 cp_fm_redistribute_end 83 14.4 7.132 12.154 7.143 12.156 cp_fm_diag_elpa_base 83 14.4 4.777 11.665 4.992 12.020 init_scf_run 11 5.9 0.000 0.001 11.556 11.558 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.556 11.558 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 10.855 11.257 multiply_cannon_metrocomm4 27577 15.6 0.095 0.109 3.843 10.649 make_images_data 5014 15.6 0.064 0.071 8.697 10.429 mp_irecv_dv 69486 16.3 3.649 10.261 3.649 10.261 hybrid_alltoall_any 5200 16.5 0.341 1.494 7.512 9.857 dbcsr_mm_accdrv_process 62242 16.2 4.495 5.068 9.038 9.622 wfi_extrapolate 11 7.9 0.001 0.001 8.277 8.277 pw_transfer 1547 11.6 0.084 0.099 7.688 7.758 grid_integrate_task_list 128 12.3 7.184 7.531 7.184 7.531 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.465 7.519 density_rs2pw 128 9.7 0.006 0.006 6.602 7.126 cp_fm_cholesky_decompose 22 10.9 6.841 6.930 6.841 6.930 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.185 6.907 fft_wrap_pw1pw2_140 523 13.2 0.470 0.515 6.553 6.605 calculate_dm_sparse 128 9.5 0.001 0.001 6.445 6.579 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.187 6.197 rs_pw_transfer 1046 11.9 0.014 0.017 5.416 6.072 mp_sum_l 7870 13.0 3.952 5.927 3.952 5.927 fft3d_ps 1291 14.7 2.774 2.925 5.840 5.877 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.340 5.406 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.212 5.388 grid_collocate_task_list 128 9.7 4.913 5.273 4.913 5.273 mp_allgather_i34 2507 14.6 1.692 4.893 1.692 4.893 potential_pw2rs 128 12.3 0.015 0.018 4.687 4.703 mp_alltoall_d11v 2415 14.1 4.115 4.690 4.115 4.690 dbcsr_complete_redistribute 395 12.7 0.823 0.952 3.258 4.097 mp_sum_d 4460 12.2 2.602 4.029 2.602 4.029 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=188.059000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.272727, yerr=2.299838 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 935.632896E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172905552 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57194. MP_Allreduce 11251 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_comm_split 82 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.051 172.994 172.996 qs_mol_dyn_low 1 2.0 0.003 0.004 172.579 172.592 qs_forces 11 3.9 0.003 0.004 172.465 172.470 qs_energies 11 4.9 0.002 0.004 165.986 165.996 scf_env_do_scf 11 5.9 0.001 0.002 150.542 150.546 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.601 115.602 velocity_verlet 10 3.0 0.002 0.002 110.695 110.697 dbcsr_multiply_generic 2485 12.5 0.181 0.188 80.966 82.077 qs_scf_new_mos 116 7.6 0.001 0.001 79.588 79.944 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.587 79.943 ot_scf_mini 116 9.6 0.004 0.004 75.476 75.857 multiply_cannon 2485 13.5 0.499 0.530 61.409 65.515 multiply_cannon_loop 2485 14.5 0.846 0.874 58.253 60.705 ot_mini 116 10.6 0.001 0.001 42.123 42.515 init_scf_loop 11 6.9 0.000 0.002 34.842 34.843 mp_waitall_1 169034 16.6 25.112 34.337 25.112 34.337 prepare_preconditioner 11 7.9 0.000 0.000 30.876 30.932 make_preconditioner 11 8.9 0.000 0.001 30.876 30.932 rebuild_ks_matrix 127 8.3 0.001 0.001 29.730 30.202 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.017 29.729 30.202 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.499 29.953 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.818 27.260 multiply_cannon_multrec 19880 15.5 13.329 16.165 21.951 24.751 multiply_cannon_metrocomm3 19880 15.5 0.060 0.063 15.438 24.703 qs_ot_get_derivative 116 11.6 0.001 0.002 22.325 22.697 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.934 21.106 apply_single 127 13.6 0.001 0.001 19.934 21.105 ot_diis_step 116 11.6 0.018 0.018 19.695 19.696 qs_ot_get_p 127 10.4 0.001 0.001 18.814 19.310 multiply_cannon_sync_h2d 19880 15.5 13.969 15.689 13.969 15.689 make_m2s 4970 13.5 0.081 0.087 14.622 15.648 make_images 4970 14.5 1.175 1.250 14.390 15.412 qs_ot_p2m_diag 82 11.4 0.261 0.268 14.696 14.706 cp_fm_cholesky_invert 11 10.9 14.363 14.371 14.363 14.371 sum_up_and_integrate 127 10.3 0.131 0.143 13.943 13.968 integrate_v_rspace 127 11.3 0.003 0.004 13.812 13.838 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.707 13.708 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.806 12.845 calculate_rho_elec 127 8.7 0.131 0.145 12.806 12.844 cp_fm_diag_elpa 82 13.4 0.000 0.001 10.647 10.648 cp_fm_redistribute_end 82 14.4 4.026 10.590 4.041 10.593 make_images_data 4970 15.5 0.060 0.067 9.092 10.478 cp_fm_diag_elpa_base 82 14.4 6.134 10.019 6.533 10.469 init_scf_run 11 5.9 0.000 0.001 10.402 10.403 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.402 10.402 hybrid_alltoall_any 5155 16.4 0.427 1.951 7.921 9.552 multiply_cannon_metrocomm4 17395 15.5 0.061 0.072 3.471 9.458 mp_irecv_dv 49801 16.2 3.349 9.207 3.349 9.207 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 8.687 8.958 dbcsr_mm_accdrv_process 41158 16.2 4.382 5.271 8.085 8.180 pw_transfer 1535 11.6 0.083 0.103 7.611 7.714 grid_integrate_task_list 127 12.3 7.217 7.681 7.217 7.681 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.390 7.499 cp_fm_upper_to_full 104 14.5 5.753 7.411 5.753 7.411 cp_fm_cholesky_decompose 22 10.9 7.305 7.342 7.305 7.342 wfi_extrapolate 11 7.9 0.001 0.001 7.323 7.323 density_rs2pw 127 9.7 0.005 0.006 6.312 6.671 fft_wrap_pw1pw2_140 519 13.2 0.474 0.526 6.529 6.639 dbcsr_complete_redistribute 393 12.7 1.177 1.208 4.619 6.373 calculate_dm_sparse 127 9.5 0.001 0.001 5.708 5.815 fft3d_ps 1281 14.7 2.675 2.877 5.673 5.735 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.551 5.562 grid_collocate_task_list 127 9.7 5.033 5.401 5.033 5.401 rs_pw_transfer 1038 11.9 0.014 0.015 4.947 5.352 copy_fm_to_dbcsr 208 11.6 0.002 0.003 3.444 5.186 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.514 5.183 mp_alltoall_d11v 2401 14.1 4.284 4.843 4.284 4.843 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.606 4.744 mp_allgather_i34 2485 14.5 1.691 4.702 1.691 4.702 mp_sum_l 7804 13.0 3.139 4.628 3.139 4.628 potential_pw2rs 127 12.3 0.020 0.022 4.399 4.423 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.358 4.075 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.968 4.001 mp_alltoall_i22 712 14.1 1.961 3.849 1.961 3.849 qs_energies_init_hamiltonians 11 5.9 0.002 0.013 3.776 3.777 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.418 3.461 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=172.996000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=888.363636, yerr=9.911176 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.152840E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379345520 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57332. MP_Allreduce 11226 1068. MP_Sync 170 MP_Alltoall 1712 12503107. 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 0.026 0.050 183.687 183.689 qs_mol_dyn_low 1 2.0 0.003 0.004 183.259 183.273 qs_forces 11 3.9 0.003 0.005 183.145 183.158 qs_energies 11 4.9 0.006 0.022 176.159 176.175 scf_env_do_scf 11 5.9 0.001 0.002 159.385 159.395 velocity_verlet 10 3.0 0.002 0.002 121.888 121.891 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 113.552 113.553 dbcsr_multiply_generic 2507 12.6 0.191 0.195 78.762 79.524 qs_scf_new_mos 117 7.6 0.001 0.001 78.250 78.546 qs_scf_loop_do_ot 117 8.6 0.001 0.001 78.249 78.545 ot_scf_mini 117 9.6 0.003 0.004 73.773 74.077 multiply_cannon 2507 13.6 0.556 0.582 54.458 57.663 multiply_cannon_loop 2507 14.6 1.181 1.209 50.939 52.635 init_scf_loop 11 6.9 0.001 0.004 45.708 45.709 ot_mini 117 10.6 0.001 0.001 41.811 42.110 prepare_preconditioner 11 7.9 0.000 0.000 41.677 41.708 make_preconditioner 11 8.9 0.000 0.001 41.677 41.708 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.386 40.217 multiply_cannon_multrec 30084 15.6 13.951 19.141 25.811 30.783 rebuild_ks_matrix 128 8.3 0.001 0.001 28.847 29.108 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 28.846 29.107 mp_waitall_1 147882 16.7 17.343 26.781 17.343 26.781 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.081 26.308 qs_ot_get_derivative 117 11.6 0.001 0.002 22.263 22.571 make_m2s 5014 13.6 0.096 0.105 20.088 20.927 make_images 5014 14.6 1.958 2.257 19.779 20.616 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.935 19.527 apply_single 128 13.6 0.001 0.001 18.935 19.526 ot_diis_step 117 11.6 0.017 0.019 19.430 19.432 qs_ot_get_p 128 10.4 0.001 0.001 18.195 18.511 cp_fm_upper_to_full 105 14.7 10.893 16.055 10.893 16.055 cp_fm_cholesky_invert 11 10.9 15.975 15.984 15.975 15.984 multiply_cannon_metrocomm3 30084 15.6 0.047 0.049 6.697 15.488 qs_ot_p2m_diag 83 11.4 0.342 0.389 14.203 14.255 sum_up_and_integrate 128 10.3 0.140 0.152 13.960 13.983 integrate_v_rspace 128 11.3 0.003 0.004 13.820 13.845 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.993 13.023 calculate_rho_elec 128 8.7 0.175 0.190 12.993 13.022 cp_dbcsr_syevd 83 12.4 0.005 0.005 12.938 12.940 multiply_cannon_sync_h2d 30084 15.6 11.731 12.726 11.731 12.726 dbcsr_complete_redistribute 395 12.7 1.554 1.674 8.917 12.555 make_images_data 5014 15.6 0.064 0.067 10.813 12.440 dbcsr_mm_accdrv_process 62264 16.2 7.342 8.219 11.433 11.900 hybrid_alltoall_any 5200 16.5 0.524 2.200 9.748 11.605 copy_fm_to_dbcsr 209 11.7 0.002 0.003 7.556 11.186 init_scf_run 11 5.9 0.000 0.001 10.689 10.690 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.688 10.690 cp_fm_diag_elpa 83 13.4 0.000 0.001 9.897 9.899 cp_fm_redistribute_end 83 14.4 1.711 9.833 1.724 9.838 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.267 9.828 cp_fm_diag_elpa_base 83 14.4 7.512 9.238 8.084 9.718 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.198 9.419 mp_alltoall_i22 716 14.1 5.408 9.089 5.408 9.089 pw_transfer 1547 11.6 0.084 0.099 7.835 7.923 grid_integrate_task_list 128 12.3 7.532 7.871 7.532 7.871 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.611 7.703 wfi_extrapolate 11 7.9 0.001 0.001 7.568 7.568 cp_fm_cholesky_decompose 22 10.9 7.437 7.533 7.437 7.533 multiply_cannon_metrocomm4 25070 15.6 0.075 0.087 2.821 7.139 mp_irecv_dv 76098 16.2 2.675 6.874 2.675 6.874 fft_wrap_pw1pw2_140 523 13.2 0.482 0.497 6.752 6.835 density_rs2pw 128 9.7 0.006 0.006 6.143 6.680 calculate_dm_sparse 128 9.5 0.001 0.001 6.153 6.240 fft3d_ps 1291 14.7 2.796 2.885 5.843 5.916 grid_collocate_task_list 128 9.7 5.196 5.649 5.196 5.649 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.490 5.553 mp_alltoall_d11v 2415 14.1 4.821 5.148 4.821 5.148 rs_pw_transfer 1046 11.9 0.013 0.015 4.568 5.133 qs_energies_init_hamiltonians 11 5.9 0.009 0.034 4.580 4.590 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.368 4.458 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.295 4.381 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.177 4.226 potential_pw2rs 128 12.3 0.022 0.023 4.178 4.187 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=183.689000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1095.272727, yerr=14.948949 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.493631E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58279. MP_Allreduce 10990 1175. MP_Sync 86 MP_Alltoall 1700 18828162. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.032 167.359 167.359 qs_mol_dyn_low 1 2.0 0.003 0.003 166.965 166.977 qs_forces 11 3.9 0.002 0.003 166.845 166.855 qs_energies 11 4.9 0.002 0.002 159.598 159.616 scf_env_do_scf 11 5.9 0.001 0.001 142.665 142.680 velocity_verlet 10 3.0 0.002 0.003 109.693 109.697 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 107.272 107.273 dbcsr_multiply_generic 2485 12.5 0.182 0.187 72.784 73.204 qs_scf_new_mos 116 7.6 0.001 0.001 72.507 72.569 qs_scf_loop_do_ot 116 8.6 0.001 0.001 72.506 72.568 ot_scf_mini 116 9.6 0.003 0.004 68.176 68.234 multiply_cannon 2485 13.5 0.581 0.606 53.550 57.860 multiply_cannon_loop 2485 14.5 0.441 0.451 49.014 49.747 ot_mini 116 10.6 0.001 0.001 38.875 38.937 init_scf_loop 11 6.9 0.000 0.000 35.243 35.244 mp_waitall_1 124680 16.7 25.069 32.541 25.069 32.541 prepare_preconditioner 11 7.9 0.000 0.000 31.435 31.453 make_preconditioner 11 8.9 0.000 0.000 31.435 31.453 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.261 29.520 rebuild_ks_matrix 127 8.3 0.001 0.001 27.979 28.055 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 27.979 28.054 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.513 25.583 multiply_cannon_multrec 9940 15.5 10.386 14.342 17.612 20.409 ot_diis_step 116 11.6 0.019 0.020 19.714 19.714 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.294 19.688 apply_single 127 13.6 0.001 0.001 19.294 19.688 qs_ot_get_derivative 116 11.6 0.002 0.002 19.091 19.142 multiply_cannon_metrocomm3 9940 15.5 0.022 0.024 12.191 19.126 make_m2s 4970 13.5 0.066 0.070 15.749 18.049 cp_fm_cholesky_invert 11 10.9 17.897 17.902 17.897 17.902 make_images 4970 14.5 2.328 2.864 15.444 17.746 qs_ot_get_p 127 10.4 0.001 0.001 16.570 16.661 sum_up_and_integrate 127 10.3 0.180 0.190 13.747 13.794 integrate_v_rspace 127 11.3 0.003 0.004 13.567 13.622 qs_ot_p2m_diag 82 11.4 0.489 0.494 13.011 13.027 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.811 12.846 calculate_rho_elec 127 8.7 0.256 0.267 12.810 12.846 multiply_cannon_sync_h2d 9940 15.5 11.533 12.092 11.533 12.092 cp_dbcsr_syevd 82 12.4 0.005 0.005 11.912 11.913 make_images_data 4970 15.5 0.052 0.059 9.557 11.872 hybrid_alltoall_any 5155 16.4 0.806 3.533 9.388 11.817 init_scf_run 11 5.9 0.000 0.001 10.119 10.119 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.119 10.119 cp_fm_diag_elpa 82 13.4 0.000 0.000 8.986 8.999 cp_fm_diag_elpa_base 82 14.4 8.748 8.822 8.976 8.988 grid_integrate_task_list 127 12.3 7.713 8.127 7.713 8.127 cp_fm_cholesky_decompose 22 10.9 7.825 7.934 7.825 7.934 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.596 7.637 pw_transfer 1535 11.6 0.082 0.088 7.488 7.501 dbcsr_mm_accdrv_process 20590 16.1 2.677 3.605 6.857 7.484 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.270 7.286 wfi_extrapolate 11 7.9 0.001 0.001 7.209 7.209 mp_allgather_i34 2485 14.5 2.762 7.011 2.762 7.011 multiply_cannon_metrocomm1 9940 15.5 0.029 0.029 4.078 6.683 fft_wrap_pw1pw2_140 519 13.2 0.499 0.518 6.424 6.444 calculate_dm_sparse 127 9.5 0.001 0.001 5.973 6.078 density_rs2pw 127 9.7 0.005 0.006 5.733 5.938 grid_collocate_task_list 127 9.7 5.443 5.756 5.443 5.756 dbcsr_complete_redistribute 393 12.7 2.126 2.181 5.182 5.575 fft3d_ps 1281 14.7 2.703 2.772 5.442 5.470 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 5.291 5.300 mp_alltoall_d11v 2401 14.1 4.585 5.222 4.585 5.222 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.121 5.130 rs_pw_transfer 1038 11.9 0.012 0.013 3.981 4.216 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.062 4.092 multiply_cannon_metrocomm4 7455 15.5 0.023 0.026 1.851 4.014 mp_irecv_dv 28618 15.9 1.814 3.944 1.814 3.944 potential_pw2rs 127 12.3 0.026 0.027 3.905 3.917 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.560 3.837 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.435 3.735 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.647 3.680 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.477 3.552 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.354 3.377 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.367 3.372 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=167.359000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1385.727273, yerr=37.207260 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 3.029455E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+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 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 58592. MP_Allreduce 11099 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.057 287.472 287.473 qs_mol_dyn_low 1 2.0 0.003 0.003 286.931 286.943 qs_forces 11 3.9 0.010 0.012 286.829 286.841 qs_energies 11 4.9 0.002 0.002 278.352 278.368 scf_env_do_scf 11 5.9 0.001 0.001 256.717 256.731 velocity_verlet 10 3.0 0.002 0.003 208.193 208.201 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 132.682 132.683 init_scf_loop 11 6.9 0.000 0.000 123.784 123.786 prepare_preconditioner 11 7.9 0.000 0.000 119.081 119.117 make_preconditioner 11 8.9 0.000 0.000 119.081 119.117 make_full_inverse_cholesky 11 9.9 0.000 0.000 94.925 116.326 qs_scf_new_mos 118 7.6 0.001 0.001 90.552 90.684 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.551 90.683 ot_scf_mini 118 9.6 0.004 0.004 85.764 85.869 dbcsr_multiply_generic 2535 12.6 0.216 0.223 82.934 83.598 cp_fm_upper_to_full 106 14.8 52.800 76.101 52.800 76.101 multiply_cannon 2535 13.6 0.722 0.748 58.453 59.116 multiply_cannon_loop 2535 14.6 0.475 0.488 54.839 56.379 ot_mini 118 10.6 0.001 0.001 44.166 44.274 dbcsr_complete_redistribute 397 12.7 3.977 4.022 29.807 42.652 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.433 39.256 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.111 36.835 mp_alltoall_i22 720 14.1 21.944 34.676 21.944 34.676 rebuild_ks_matrix 129 8.3 0.001 0.001 33.662 33.748 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 33.661 33.747 cp_fm_cholesky_invert 11 10.9 33.136 33.142 33.136 33.142 mp_waitall_1 104820 16.8 27.669 31.745 27.669 31.745 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.483 31.569 qs_ot_get_p 129 10.4 0.001 0.001 26.112 26.205 qs_ot_get_derivative 118 11.6 0.002 0.002 24.070 24.176 qs_ot_p2m_diag 84 11.4 0.889 0.896 22.052 22.079 make_m2s 5070 13.6 0.077 0.078 19.753 20.640 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.320 20.321 make_images 5070 14.6 3.791 3.896 19.269 20.159 ot_diis_step 118 11.6 0.022 0.022 20.063 20.063 multiply_cannon_metrocomm3 10140 15.6 0.024 0.024 18.481 19.835 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.245 19.467 apply_single 129 13.6 0.001 0.001 19.245 19.467 multiply_cannon_multrec 10140 15.6 10.612 12.383 18.045 18.117 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.111 17.111 cp_fm_diag_elpa_base 84 14.4 12.700 14.321 17.106 17.107 multiply_cannon_sync_h2d 10140 15.6 15.784 15.792 15.784 15.792 sum_up_and_integrate 129 10.3 0.324 0.326 15.681 15.782 integrate_v_rspace 129 11.3 0.004 0.004 15.357 15.456 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.957 14.975 calculate_rho_elec 129 8.7 0.486 0.487 14.957 14.974 hybrid_alltoall_any 5257 16.5 1.309 3.048 10.688 12.437 make_images_data 5070 15.6 0.061 0.065 10.470 12.128 init_scf_run 11 5.9 0.000 0.001 11.702 11.702 scf_env_initial_rho_setup 11 6.9 0.000 0.000 11.702 11.702 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.452 9.539 dbcsr_mm_accdrv_process 20958 16.1 3.776 5.592 7.194 9.048 cp_fm_cholesky_decompose 22 10.9 8.776 8.842 8.776 8.842 grid_integrate_task_list 129 12.3 8.602 8.790 8.602 8.790 wfi_extrapolate 11 7.9 0.001 0.001 8.633 8.633 pw_transfer 1559 11.6 0.091 0.091 8.266 8.275 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.032 8.039 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 7.951 7.968 fft_wrap_pw1pw2_140 527 13.2 0.545 0.548 7.129 7.136 mp_alltoall_d11v 2429 14.1 6.848 7.006 6.848 7.006 calculate_dm_sparse 129 9.5 0.001 0.001 6.645 6.716 grid_collocate_task_list 129 9.7 6.386 6.429 6.386 6.429 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.328 6.405 fft3d_ps 1301 14.7 2.775 2.779 6.064 6.079 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.008 6.055 density_rs2pw 129 9.7 0.005 0.005 5.974 6.007 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=287.473000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2667.545455, yerr=195.175615 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 1.261339E+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 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 0.019 0.036 84.759 84.761 qs_energies 1 2.0 0.000 0.000 84.324 84.331 ls_scf 1 3.0 0.000 0.000 83.412 83.419 dbcsr_multiply_generic 111 6.7 0.014 0.016 72.590 72.739 multiply_cannon 111 7.7 0.017 0.021 55.845 57.090 multiply_cannon_loop 111 8.7 0.207 0.221 52.398 53.795 ls_scf_main 1 4.0 0.000 0.000 52.151 52.154 density_matrix_trs4 2 5.0 0.002 0.003 46.696 46.776 ls_scf_init_scf 1 4.0 0.000 0.000 28.230 28.233 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.157 27.211 mp_waitall_1 11031 10.9 22.554 26.291 22.554 26.291 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.983 25.002 multiply_cannon_multrec 2664 9.7 8.144 9.022 15.413 17.240 multiply_cannon_sync_h2d 2664 9.7 13.517 15.705 13.517 15.705 make_m2s 222 7.7 0.009 0.011 12.952 13.472 make_images 222 8.7 0.099 0.108 12.930 13.452 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.768 13.048 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.565 8.390 make_images_data 222 9.7 0.004 0.005 7.546 8.101 hybrid_alltoall_any 227 10.6 0.214 1.828 6.482 7.832 dbcsr_mm_accdrv_process 4760 10.4 0.508 0.619 6.887 7.828 dbcsr_mm_accdrv_process_sort 4760 11.4 6.180 7.039 6.180 7.039 calculate_norms 4752 9.8 5.483 6.124 5.483 6.124 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.140 5.265 mp_sum_l 807 5.4 3.258 4.826 3.258 4.826 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.414 3.732 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.072 3.591 mp_irecv_dv 6231 10.9 2.054 3.569 2.054 3.569 make_images_sizes 222 9.7 0.000 0.000 0.642 3.268 mp_alltoall_i44 222 10.7 0.642 3.267 0.642 3.267 arnoldi_extremal 4 6.8 0.000 0.000 3.211 3.241 arnoldi_normal_ev 4 7.8 0.001 0.003 3.211 3.241 build_subspace 16 8.4 0.009 0.012 3.119 3.121 ls_scf_post 1 4.0 0.000 0.000 3.031 3.037 ls_scf_store_result 1 5.0 0.000 0.000 2.854 2.896 dbcsr_special_finalize 555 9.7 0.005 0.006 2.410 2.758 dbcsr_merge_single_wm 555 10.7 0.467 0.595 2.402 2.750 make_images_pack 222 9.7 2.207 2.633 2.208 2.635 dbcsr_matrix_vector_mult 304 9.0 0.004 0.010 2.325 2.560 dbcsr_sort_data 658 11.4 2.194 2.497 2.194 2.497 dbcsr_matrix_vector_mult_local 304 10.0 2.065 2.462 2.067 2.464 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.186 2.255 buffer_matrices_ensure_size 222 8.7 1.762 2.104 1.762 2.104 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.728 1.731 rebuild_ks_matrix 3 7.3 0.000 0.000 1.719 1.721 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.006 1.719 1.721 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.761000, 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/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 2.072502E+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 0.032 0.047 91.713 91.715 qs_energies 1 2.0 0.000 0.000 91.179 91.183 ls_scf 1 3.0 0.000 0.000 89.817 89.832 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.948 76.379 multiply_cannon 111 7.7 0.029 0.043 53.364 57.588 ls_scf_main 1 4.0 0.000 0.000 55.443 55.450 multiply_cannon_loop 111 8.7 0.116 0.122 50.057 53.467 density_matrix_trs4 2 5.0 0.002 0.003 49.804 50.029 ls_scf_init_scf 1 4.0 0.000 0.001 30.823 30.838 mp_waitall_1 9105 10.9 21.540 30.356 21.540 30.356 ls_scf_init_matrix_S 1 5.0 0.001 0.005 29.680 29.789 multiply_cannon_multrec 1332 9.7 13.064 16.933 22.229 27.215 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.191 27.211 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.991 20.991 make_m2s 222 7.7 0.006 0.007 15.561 16.220 make_images 222 8.7 1.573 1.961 15.531 16.190 dbcsr_mm_accdrv_process 4041 10.4 0.304 0.454 8.759 10.316 make_images_data 222 9.7 0.004 0.004 8.989 9.940 dbcsr_mm_accdrv_process_sort 4041 11.4 8.352 9.862 8.352 9.862 hybrid_alltoall_any 227 10.6 0.523 2.483 8.430 9.562 mp_sum_l 807 5.4 5.716 9.175 5.716 9.175 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.234 7.623 mp_irecv_dv 3311 11.0 3.213 7.569 3.213 7.569 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.435 7.279 calculate_norms 2376 9.8 6.000 6.747 6.000 6.747 multiply_cannon_sync_h2d 1332 9.7 4.843 6.093 4.843 6.093 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.025 5.218 arnoldi_extremal 4 6.8 0.000 0.000 4.605 4.641 arnoldi_normal_ev 4 7.8 0.001 0.004 4.605 4.641 build_subspace 16 8.4 0.014 0.021 4.339 4.342 ls_scf_post 1 4.0 0.000 0.000 3.550 3.555 ls_scf_store_result 1 5.0 0.000 0.000 3.219 3.374 dbcsr_matrix_vector_mult 304 9.0 0.005 0.018 3.108 3.346 dbcsr_matrix_vector_mult_local 304 10.0 2.726 3.213 2.728 3.215 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.194 2.788 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.509 2.597 mp_allgather_i34 111 8.7 1.004 2.494 1.004 2.494 make_images_pack 222 9.7 2.025 2.435 2.027 2.437 dbcsr_sort_data 436 11.2 1.860 2.112 1.860 2.112 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.715000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1747.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 2.695512E+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 0.044 0.106 94.223 94.225 qs_energies 1 2.0 0.000 0.000 93.446 93.449 ls_scf 1 3.0 0.000 0.001 91.977 91.980 dbcsr_multiply_generic 111 6.7 0.016 0.020 76.909 77.261 ls_scf_main 1 4.0 0.000 0.000 57.694 57.700 multiply_cannon 111 7.7 0.040 0.072 52.986 57.159 multiply_cannon_loop 111 8.7 0.099 0.105 49.442 53.491 density_matrix_trs4 2 5.0 0.002 0.003 51.794 51.992 mp_waitall_1 7281 11.0 24.336 34.405 24.336 34.405 ls_scf_init_scf 1 4.0 0.000 0.001 30.722 30.724 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.576 29.673 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.183 27.197 multiply_cannon_multrec 888 9.7 12.506 14.987 20.975 23.975 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.434 23.466 make_m2s 222 7.7 0.006 0.007 17.216 18.436 make_images 222 8.7 1.958 2.298 17.178 18.396 hybrid_alltoall_any 227 10.6 0.621 2.859 9.561 10.922 make_images_data 222 9.7 0.003 0.004 9.887 10.907 dbcsr_mm_accdrv_process 3754 10.4 0.239 0.413 8.006 9.219 dbcsr_mm_accdrv_process_sort 3754 11.4 7.636 8.806 7.636 8.806 mp_sum_l 807 5.4 5.155 8.687 5.155 8.687 multiply_cannon_sync_h2d 888 9.7 6.057 7.570 6.057 7.570 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.477 7.105 mp_irecv_dv 2335 11.1 2.461 7.049 2.461 7.049 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.962 6.722 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.635 6.263 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.942 5.160 arnoldi_extremal 4 6.8 0.000 0.000 5.071 5.096 arnoldi_normal_ev 4 7.8 0.001 0.005 5.071 5.096 build_subspace 16 8.4 0.014 0.020 4.763 4.769 calculate_norms 1584 9.8 4.255 4.637 4.255 4.637 mp_allgather_i34 111 8.7 1.378 3.921 1.378 3.921 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.417 3.752 dbcsr_matrix_vector_mult_local 304 10.0 2.996 3.570 2.998 3.572 ls_scf_post 1 4.0 0.000 0.002 3.561 3.564 ls_scf_store_result 1 5.0 0.000 0.000 3.299 3.388 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.741 2.832 dbcsr_sort_data 325 11.1 1.882 2.179 1.882 2.179 make_images_pack 222 9.7 1.816 2.146 1.819 2.148 make_images_sizes 222 9.7 0.000 0.000 1.024 2.075 mp_alltoall_i44 222 10.7 1.024 2.075 1.024 2.075 dbcsr_finalize 304 7.8 0.026 0.032 1.607 1.910 dbcsr_data_release 9322 10.9 1.298 1.893 1.298 1.893 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.225000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2198.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 3.342307E+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 0.031 0.047 97.541 97.542 qs_energies 1 2.0 0.000 0.000 96.942 96.948 ls_scf 1 3.0 0.000 0.000 95.290 95.296 dbcsr_multiply_generic 111 6.7 0.016 0.018 79.296 79.575 ls_scf_main 1 4.0 0.000 0.000 59.369 59.370 multiply_cannon 111 7.7 0.056 0.109 51.872 56.291 density_matrix_trs4 2 5.0 0.002 0.003 53.345 53.480 multiply_cannon_loop 111 8.7 0.114 0.127 46.836 50.457 ls_scf_init_scf 1 4.0 0.000 0.000 32.627 32.628 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.462 31.561 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.999 29.010 mp_waitall_1 6369 11.0 23.164 28.837 23.164 28.837 multiply_cannon_multrec 1332 9.7 14.191 17.693 21.999 24.451 make_m2s 222 7.7 0.007 0.008 21.308 22.723 make_images 222 8.7 3.130 3.592 21.257 22.675 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.362 17.529 make_images_data 222 9.7 0.004 0.004 11.911 13.421 hybrid_alltoall_any 227 10.6 0.798 3.837 11.257 13.065 dbcsr_mm_accdrv_process 3641 10.4 0.211 0.407 7.443 8.967 mp_sum_l 807 5.4 4.547 8.776 4.547 8.776 dbcsr_mm_accdrv_process_sort 3641 11.4 7.070 8.545 7.070 8.545 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.593 7.129 multiply_cannon_sync_h2d 1332 9.7 5.516 6.205 5.516 6.205 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.116 6.166 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.723 6.111 mp_irecv_dv 3229 10.9 2.093 6.081 2.093 6.081 arnoldi_extremal 4 6.8 0.000 0.000 5.164 5.184 arnoldi_normal_ev 4 7.8 0.001 0.004 5.164 5.184 build_subspace 16 8.4 0.014 0.021 4.827 4.834 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.567 4.744 calculate_norms 2376 9.8 4.192 4.554 4.192 4.554 mp_allgather_i34 111 8.7 2.164 4.400 2.164 4.400 dbcsr_matrix_vector_mult 304 9.0 0.006 0.017 3.540 3.842 dbcsr_matrix_vector_mult_local 304 10.0 3.152 3.650 3.154 3.652 dbcsr_sort_data 658 11.4 3.083 3.591 3.083 3.591 ls_scf_post 1 4.0 0.000 0.000 3.294 3.301 dbcsr_special_finalize 555 9.7 0.006 0.006 2.839 3.287 dbcsr_merge_single_wm 555 10.7 0.539 0.674 2.830 3.280 ls_scf_store_result 1 5.0 0.000 0.000 3.018 3.099 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.887 2.931 dbcsr_data_release 10477 10.7 1.578 2.415 1.578 2.415 dbcsr_finalize 304 7.8 0.049 0.061 1.791 2.047 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.542000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2709.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 4.625281E+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 0.057 0.073 98.740 98.742 qs_energies 1 2.0 0.000 0.000 97.960 97.964 ls_scf 1 3.0 0.000 0.000 96.049 96.054 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.886 78.083 ls_scf_main 1 4.0 0.000 0.000 61.873 61.876 multiply_cannon 111 7.7 0.098 0.197 55.579 61.059 density_matrix_trs4 2 5.0 0.002 0.003 54.882 54.985 multiply_cannon_loop 111 8.7 0.069 0.077 50.966 52.638 mp_waitall_1 5436 11.0 26.625 31.881 26.625 31.881 ls_scf_init_scf 1 4.0 0.000 0.000 30.504 30.509 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.319 29.366 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.154 27.164 multiply_cannon_multrec 444 9.7 13.944 16.333 20.939 23.449 make_m2s 222 7.7 0.004 0.005 17.496 20.040 make_images 222 8.7 3.721 4.433 17.434 19.982 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.321 16.486 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.351 14.251 make_images_data 222 9.7 0.003 0.004 9.754 12.255 hybrid_alltoall_any 227 10.6 0.787 3.769 9.550 12.076 multiply_cannon_sync_h2d 444 9.7 6.536 8.143 6.536 8.143 dbcsr_mm_accdrv_process 3003 10.4 0.175 0.331 6.700 7.826 dbcsr_mm_accdrv_process_sort 3003 11.4 6.385 7.479 6.385 7.479 mp_allgather_i34 111 8.7 2.851 7.038 2.851 7.038 arnoldi_extremal 4 6.8 0.000 0.000 5.749 5.761 arnoldi_normal_ev 4 7.8 0.002 0.005 5.749 5.761 build_subspace 16 8.4 0.015 0.019 5.362 5.370 mp_sum_l 807 5.4 2.861 4.788 2.861 4.788 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.603 4.775 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.637 4.328 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.110 4.316 mp_irecv_dv 1241 11.2 1.617 4.300 1.617 4.300 dbcsr_matrix_vector_mult_local 304 10.0 3.636 4.106 3.638 4.109 ls_scf_post 1 4.0 0.000 0.000 3.673 3.677 calculate_norms 792 9.8 3.545 3.671 3.545 3.671 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.958 3.662 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.403 3.512 ls_scf_store_result 1 5.0 0.000 0.000 3.448 3.495 make_images_sizes 222 9.7 0.000 0.000 1.095 3.302 mp_alltoall_i44 222 10.7 1.095 3.302 1.095 3.302 dbcsr_finalize 304 7.8 0.062 0.077 2.191 2.256 dbcsr_merge_all 275 8.9 0.472 0.522 2.041 2.095 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.742000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3627.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb_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 8.720540E+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 0.076 0.094 107.801 107.802 qs_energies 1 2.0 0.000 0.000 106.405 106.418 ls_scf 1 3.0 0.000 0.000 103.543 103.556 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.920 78.017 ls_scf_main 1 4.0 0.000 0.000 65.921 65.922 density_matrix_trs4 2 5.0 0.002 0.003 57.087 57.167 multiply_cannon 111 7.7 0.196 0.361 50.292 52.061 multiply_cannon_loop 111 8.7 0.068 0.078 46.880 47.617 ls_scf_init_scf 1 4.0 0.000 0.000 33.954 33.955 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.637 32.648 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.802 29.816 mp_waitall_1 4527 11.1 22.741 26.381 22.741 26.381 make_m2s 222 7.7 0.005 0.005 23.967 24.868 make_images 222 8.7 4.577 4.959 23.859 24.760 multiply_cannon_multrec 444 9.7 17.877 18.814 22.465 23.186 hybrid_alltoall_any 227 10.6 1.658 3.617 12.977 15.828 make_images_data 222 9.7 0.003 0.003 13.187 15.552 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.964 11.593 multiply_cannon_sync_h2d 444 9.7 8.846 8.880 8.846 8.880 arnoldi_extremal 4 6.8 0.000 0.000 7.277 7.296 arnoldi_normal_ev 4 7.8 0.003 0.009 7.277 7.296 build_subspace 16 8.4 0.026 0.036 6.730 6.741 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.367 5.516 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.044 5.302 dbcsr_matrix_vector_mult_local 304 10.0 4.904 5.206 4.907 5.209 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.766 4.859 dbcsr_mm_accdrv_process 1814 10.4 0.257 0.324 4.413 4.541 dbcsr_mm_accdrv_process_sort 1814 11.4 4.109 4.236 4.109 4.236 ls_scf_post 1 4.0 0.000 0.000 3.667 3.681 make_images_sizes 222 9.7 0.000 0.000 1.467 3.629 mp_alltoall_i44 222 10.7 1.467 3.629 1.467 3.629 ls_scf_store_result 1 5.0 0.000 0.000 3.417 3.427 mp_allgather_i34 111 8.7 1.064 3.299 1.064 3.299 calculate_norms 792 9.8 3.242 3.278 3.242 3.278 dbcsr_finalize 304 7.8 0.082 0.089 3.084 3.135 dbcsr_merge_all 275 8.9 0.892 0.912 2.871 2.910 dbcsr_complete_redistribute 5 7.6 1.430 1.461 2.771 2.884 dbcsr_data_release 12724 10.6 2.326 2.850 2.326 2.850 qs_energies_init_hamiltonians 1 3.0 0.002 0.002 2.832 2.832 matrix_ls_to_qs 2 6.0 0.000 0.000 2.408 2.530 dbcsr_sort_data 325 11.1 2.440 2.498 2.440 2.498 dbcsr_new_transposed 4 7.5 0.243 0.252 2.276 2.290 dbcsr_frobenius_norm 74 6.6 2.055 2.132 2.185 2.220 dbcsr_add_d 103 6.2 0.000 0.000 2.134 2.207 dbcsr_add_anytype 103 7.2 0.860 0.895 2.134 2.206 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.187 2.189 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.802000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6799.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: ebdeb2876a9d0cf37cfa08b2b47b3dfed26951eb Summary: empty Status: OK