=== 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: 3169bf56b7b69a13599bda2029f1ce159e56e4d8 ################# 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/01 job id: 43798233 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/02 job id: 43798234 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/03 job id: 43798235 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/04 job id: 43798236 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/05 job id: 43798237 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/06 job id: 43798238 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/07 job id: 43798239 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/08 job id: 43798240 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/09 job id: 43798241 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/10 job id: 43798242 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/11 job id: 43798243 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/12 job id: 43798244 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/13 job id: 43798245 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/14 job id: 43798247 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/15 job id: 43798248 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/16 job id: 43798251 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/17 job id: 43798252 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/18 job id: 43798254 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/19 job id: 43798255 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/20 job id: 43798256 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/21 job id: 43798257 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/22 job id: 43798258 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/23 job id: 43798259 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/24 job id: 43798260 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/25 job id: 43798261 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/26 job id: 43798262 --- 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 0.037 134.266 134.267 farming_run 1 2.0 133.682 133.682 134.238 134.242 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.460310E+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 0.028 115.992 115.992 qs_energies 1 2.0 0.000 0.000 115.784 115.787 mp2_main 1 3.0 0.000 0.000 113.758 113.761 mp2_gpw_main 1 4.0 0.019 0.022 112.777 112.780 mp2_ri_gpw_compute_in 1 5.0 0.173 0.179 93.719 94.247 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.497 56.027 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.170 41.795 47.293 get_2c_integrals 1 6.0 0.000 0.000 37.483 38.045 integrate_v_rspace 273 8.0 0.436 0.449 25.171 30.454 pw_transfer 6555 10.6 0.379 0.401 27.394 27.998 grid_integrate_task_list 273 9.0 20.975 26.763 20.975 26.763 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 26.070 26.689 fft_wrap_pw1pw2_100 2178 12.4 1.185 1.393 23.632 24.255 compute_2c_integrals 1 7.0 0.003 0.003 19.641 19.641 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.878 19.328 mp2_eri_2c_integrate_gpw 1 9.0 2.382 2.439 18.874 19.324 rpa_ri_compute_en 1 5.0 0.000 0.002 18.949 19.101 cp_fm_cholesky_decompose 12 8.2 17.847 18.404 17.847 18.404 cholesky_decomp 1 7.0 0.000 0.000 16.691 17.256 fft3d_s 5443 13.4 16.160 16.636 16.182 16.657 ao_to_mo_and_store_B_mult_1 272 7.0 10.860 15.591 10.860 15.591 calculate_wavefunction 272 8.0 5.401 5.565 12.542 13.155 rpa_num_int 1 6.0 0.000 0.001 10.661 10.661 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.584 10.631 calc_mat_Q 8 8.0 0.000 0.000 9.378 9.474 contract_S_to_Q 8 9.0 0.000 0.000 8.797 8.895 calc_potential_gpw 544 9.5 0.005 0.006 8.282 8.636 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.002 0.002 8.218 8.524 parallel_gemm_fm 14 9.1 0.000 0.000 8.369 8.481 parallel_gemm_fm_cosma 14 10.1 8.369 8.481 8.369 8.481 potential_pw2rs 545 10.0 0.106 0.108 7.673 8.294 collocate_single_gaussian 272 10.0 0.041 0.043 7.455 7.699 create_integ_mat 1 6.0 0.022 0.028 7.664 7.665 array2fm 1 7.0 0.000 0.000 6.627 7.070 pw_scatter_s 2720 13.7 4.452 4.604 4.452 4.604 pw_gather_s 2722 13.2 3.902 4.308 3.902 4.308 array2fm_buffer_send 1 8.0 2.960 3.149 2.960 3.149 pw_poisson_solve 545 10.5 1.128 1.183 2.153 2.365 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.779311, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2731.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 0.028 0.038 395.315 395.316 farming_run 1 2.0 394.518 394.528 395.269 395.273 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.223541E+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 3486442. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.042 209.016 209.016 qs_energies 1 2.0 0.000 0.000 208.791 208.799 scf_env_do_scf 1 3.0 0.000 0.000 105.820 105.820 qs_ks_update_qs_env 5 5.0 0.000 0.000 104.880 104.889 rebuild_ks_matrix 4 6.0 0.000 0.000 104.879 104.887 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.065 104.879 104.887 hfx_ks_matrix 4 8.0 0.001 0.001 104.503 104.506 integrate_four_center 4 9.0 0.143 0.448 104.502 104.505 mp2_main 1 3.0 0.000 0.000 102.675 102.684 mp2_gpw_main 1 4.0 0.031 0.044 101.701 101.711 integrate_four_center_main 4 10.0 0.086 0.419 96.729 98.764 integrate_four_center_bin 268 11.0 96.643 98.762 96.643 98.762 init_scf_loop 1 4.0 0.000 0.000 91.622 91.622 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.874 75.908 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.453 55.489 mp2_eri_3c_integrate_gpw 91 7.0 0.146 0.163 42.137 47.183 integrate_v_rspace 95 8.0 0.399 0.571 28.500 33.344 pw_transfer 2240 10.6 0.146 0.173 29.912 30.397 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.910 29.418 grid_integrate_task_list 95 9.0 23.788 28.856 23.788 28.856 mp2_ri_gpw_compute_en 1 5.0 0.056 0.064 26.677 28.347 ao_to_mo_and_store_B_mult_1 91 7.0 10.632 28.302 10.632 28.302 fft_wrap_pw1pw2_100 730 12.4 1.296 1.463 26.635 27.153 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.844 1.907 24.995 25.006 get_2c_integrals 1 6.0 0.000 0.000 20.303 20.356 compute_2c_integrals 1 7.0 0.002 0.003 19.287 19.290 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.928 19.156 mp2_eri_2c_integrate_gpw 1 9.0 1.736 1.984 18.927 19.154 fft3d_s 1823 13.4 18.405 18.675 18.418 18.688 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.196 14.196 calculate_wavefunction 91 8.0 2.029 2.066 9.752 9.999 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.590 8.753 9.403 potential_pw2rs 186 10.0 0.034 0.035 8.611 9.269 local_gemm 172 8.0 8.196 8.839 8.196 8.839 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.284 8.523 mp2_ri_gpw_compute_en_comm 22 7.0 0.492 0.511 8.010 8.433 calc_potential_gpw 182 9.5 0.002 0.002 7.977 8.258 collocate_single_gaussian 91 10.0 0.017 0.025 7.897 8.106 mp_sendrecv_dm3 2068 8.0 6.065 6.456 6.065 6.456 mp2_ri_gpw_compute_en_ener 172 7.0 6.342 6.453 6.342 6.453 pw_gather_s 912 13.2 4.882 5.456 4.882 5.456 mp_sync 38 10.4 2.481 4.960 2.481 4.960 pw_scatter_s 910 13.7 3.955 4.485 3.955 4.485 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.711346, 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 451.043328E+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 18780. 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 0.012 0.029 53.714 53.715 qs_mol_dyn_low 1 2.0 0.003 0.004 53.498 53.507 qs_forces 11 3.9 0.002 0.003 53.434 53.436 qs_energies 11 4.9 0.001 0.002 51.933 51.942 scf_env_do_scf 11 5.9 0.000 0.001 45.690 45.691 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.469 43.470 dbcsr_multiply_generic 2286 12.5 0.094 0.101 34.054 34.547 qs_scf_new_mos 108 7.5 0.000 0.001 32.838 33.131 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.838 33.130 ot_scf_mini 108 9.5 0.002 0.002 31.195 31.392 multiply_cannon 2286 13.5 0.183 0.191 26.281 27.787 multiply_cannon_loop 2286 14.5 1.505 1.577 25.452 26.915 velocity_verlet 10 3.0 0.001 0.002 26.815 26.816 ot_mini 108 10.5 0.001 0.001 19.725 19.965 qs_ot_get_derivative 108 11.5 0.001 0.002 16.740 16.920 mp_waitall_1 245248 16.5 8.691 14.889 8.691 14.889 multiply_cannon_metrocomm3 54864 15.5 0.067 0.072 5.996 13.374 multiply_cannon_multrec 54864 15.5 4.217 6.570 7.730 11.167 rebuild_ks_matrix 119 8.3 0.000 0.000 8.213 8.335 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.212 8.335 multiply_cannon_sync_h2d 54864 15.5 5.874 7.552 5.874 7.552 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.247 7.361 qs_ot_get_p 119 10.4 0.001 0.001 6.803 7.104 mp_sum_l 7207 12.9 5.300 6.908 5.300 6.908 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.801 6.272 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.532 5.660 init_scf_run 11 5.9 0.000 0.001 4.972 4.972 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.972 4.972 dbcsr_mm_accdrv_process 76910 16.1 1.181 1.828 3.436 4.769 sum_up_and_integrate 119 10.3 0.012 0.014 4.742 4.750 integrate_v_rspace 119 11.3 0.002 0.002 4.730 4.739 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.462 4.560 calculate_rho_elec 119 8.7 0.011 0.016 4.461 4.560 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.807 3.872 multiply_cannon_metrocomm1 54864 15.5 0.051 0.057 1.853 3.258 calculate_dm_sparse 119 9.5 0.000 0.000 2.980 3.149 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.938 3.138 apply_single 119 13.6 0.000 0.000 2.938 3.138 rs_pw_transfer 974 11.9 0.011 0.012 3.013 3.117 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.954 2.954 jit_kernel_multiply 13 15.8 2.192 2.855 2.192 2.855 ot_diis_step 108 11.5 0.006 0.006 2.726 2.727 calculate_first_density_matrix 1 7.0 0.001 0.001 2.697 2.703 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.678 2.678 cp_fm_redistribute_end 50 14.0 2.435 2.653 2.441 2.655 density_rs2pw 119 9.7 0.004 0.004 2.506 2.604 cp_fm_diag_elpa_base 50 14.0 0.212 2.556 0.213 2.569 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.449 2.504 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.415 2.418 make_m2s 4572 13.5 0.055 0.057 2.162 2.230 acc_transpose_blocks 54864 15.5 0.233 0.255 1.796 2.230 wfi_extrapolate 11 7.9 0.001 0.001 2.212 2.212 init_scf_loop 11 6.9 0.000 0.000 2.198 2.199 make_images 4572 14.5 0.133 0.138 2.077 2.144 grid_integrate_task_list 119 12.3 2.012 2.137 2.012 2.137 pw_transfer 1439 11.6 0.051 0.055 2.046 2.120 potential_pw2rs 119 12.3 0.004 0.004 2.083 2.117 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.059 2.114 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 1.970 2.046 mp_sum_d 4125 12.0 1.310 1.918 1.310 1.918 fft3d_ps 1201 14.6 0.360 0.466 1.749 1.820 mp_alltoall_d11v 2130 13.8 1.532 1.705 1.532 1.705 fft_wrap_pw1pw2_140 487 13.2 0.079 0.091 1.497 1.575 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.509 1.526 mp_waitany 12084 13.8 1.266 1.362 1.266 1.362 grid_collocate_task_list 119 9.7 1.290 1.348 1.290 1.348 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.709 1.125 make_images_sizes 4572 15.5 0.004 0.005 0.892 1.122 mp_alltoall_i44 4572 16.5 0.888 1.118 0.888 1.118 prepare_preconditioner 11 7.9 0.000 0.000 1.080 1.102 make_preconditioner 11 8.9 0.000 0.000 1.079 1.102 acc_transpose_blocks_kernels 54864 16.5 0.251 0.380 0.797 1.076 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.715000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.943616E+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 1185420. 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 0.038 0.103 39.595 39.596 qs_mol_dyn_low 1 2.0 0.003 0.005 39.183 39.191 qs_forces 11 3.9 0.002 0.005 39.104 39.105 qs_energies 11 4.9 0.002 0.008 37.415 37.419 scf_env_do_scf 11 5.9 0.001 0.007 31.936 31.937 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 29.374 29.374 dbcsr_multiply_generic 2286 12.5 0.101 0.107 21.774 22.155 qs_scf_new_mos 108 7.5 0.001 0.001 20.339 20.595 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.338 20.595 ot_scf_mini 108 9.5 0.002 0.003 19.424 19.608 velocity_verlet 10 3.0 0.001 0.002 18.516 18.518 multiply_cannon 2286 13.5 0.212 0.252 16.384 18.032 multiply_cannon_loop 2286 14.5 0.904 0.978 15.170 16.763 ot_mini 108 10.5 0.001 0.001 11.967 12.212 mp_waitall_1 200699 16.5 5.890 11.155 5.890 11.155 qs_ot_get_derivative 108 11.5 0.001 0.001 9.477 9.662 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.191 9.661 multiply_cannon_multrec 27432 15.5 1.968 4.566 5.843 8.780 rebuild_ks_matrix 119 8.3 0.000 0.000 7.188 7.332 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.187 7.332 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.352 6.485 dbcsr_mm_accdrv_process 47894 16.0 3.022 5.352 3.807 5.635 qs_ot_get_p 119 10.4 0.001 0.001 4.536 4.771 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.635 4.490 sum_up_and_integrate 119 10.3 0.025 0.027 4.210 4.217 integrate_v_rspace 119 11.3 0.002 0.002 4.186 4.194 init_scf_run 11 5.9 0.000 0.001 4.184 4.185 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.184 4.184 mp_sum_l 7207 12.9 2.168 4.169 2.168 4.169 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.071 4.146 apply_single 119 13.6 0.000 0.000 3.071 4.146 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.644 3.685 calculate_rho_elec 119 8.7 0.021 0.024 3.644 3.684 make_m2s 4572 13.5 0.054 0.056 2.775 3.064 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.985 3.005 make_images 4572 14.5 0.202 0.248 2.684 2.971 multiply_cannon_sync_h2d 27432 15.5 2.167 2.774 2.167 2.774 rs_pw_transfer 974 11.9 0.010 0.011 2.638 2.729 calculate_first_density_matrix 1 7.0 0.001 0.006 2.620 2.622 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.534 2.535 init_scf_loop 11 6.9 0.001 0.004 2.533 2.534 ot_diis_step 108 11.5 0.011 0.012 2.435 2.435 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.199 2.295 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.182 2.185 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.158 2.159 cp_fm_redistribute_end 50 14.0 1.787 2.134 1.791 2.136 calculate_dm_sparse 119 9.5 0.001 0.001 2.053 2.131 density_rs2pw 119 9.7 0.004 0.004 2.003 2.088 cp_fm_diag_elpa_base 50 14.0 0.330 2.028 0.343 2.078 potential_pw2rs 119 12.3 0.006 0.006 1.955 1.970 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.904 1.950 grid_integrate_task_list 119 12.3 1.836 1.939 1.836 1.939 pw_transfer 1439 11.6 0.063 0.066 1.805 1.839 make_images_data 4572 15.5 0.045 0.051 1.283 1.833 jit_kernel_multiply 9 16.1 0.733 1.764 0.733 1.764 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.716 1.750 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.106 1.702 prepare_preconditioner 11 7.9 0.000 0.000 1.611 1.638 make_preconditioner 11 8.9 0.000 0.001 1.611 1.638 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.510 1.569 wfi_extrapolate 11 7.9 0.001 0.001 1.512 1.512 mp_allgather_i34 2286 14.5 0.651 1.502 0.651 1.502 acc_transpose_blocks 27432 15.5 0.109 0.114 1.188 1.489 fft3d_ps 1201 14.6 0.498 0.552 1.424 1.455 mp_alltoall_d11v 2130 13.8 1.251 1.359 1.251 1.359 fft_wrap_pw1pw2_140 487 13.2 0.075 0.082 1.305 1.337 grid_collocate_task_list 119 9.7 1.243 1.335 1.243 1.335 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.247 1.254 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.196 1.243 mp_sum_d 4125 12.0 0.638 1.072 0.638 1.072 qs_energies_init_hamiltonians 11 5.9 0.002 0.013 1.009 1.010 make_images_sizes 4572 15.5 0.005 0.005 0.629 0.919 mp_alltoall_i44 4572 16.5 0.624 0.914 0.624 0.914 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.893 0.906 acc_transpose_blocks_kernels 27432 16.5 0.182 0.270 0.655 0.860 rs_pw_transfer_PW2RS_50 119 14.3 0.589 0.609 0.773 0.805 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.596000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.454545, yerr=1.616035 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 521.715712E+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 0.054 0.093 32.693 32.694 qs_mol_dyn_low 1 2.0 0.003 0.004 32.214 32.221 qs_forces 11 3.9 0.002 0.004 32.125 32.125 qs_energies 11 4.9 0.002 0.006 30.542 30.545 scf_env_do_scf 11 5.9 0.001 0.002 25.868 25.868 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.189 23.189 dbcsr_multiply_generic 2286 12.5 0.094 0.097 16.535 16.659 velocity_verlet 10 3.0 0.001 0.002 15.385 15.388 qs_scf_new_mos 108 7.5 0.001 0.001 15.169 15.184 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.168 15.183 ot_scf_mini 108 9.5 0.002 0.003 14.429 14.445 multiply_cannon 2286 13.5 0.195 0.204 13.055 13.797 multiply_cannon_loop 2286 14.5 0.637 0.665 12.242 12.978 ot_mini 108 10.5 0.001 0.001 8.863 8.878 qs_ot_get_derivative 108 11.5 0.001 0.001 7.329 7.345 multiply_cannon_multrec 18288 15.5 1.937 2.965 6.732 7.075 rebuild_ks_matrix 119 8.3 0.000 0.000 6.429 6.457 qs_ks_build_kohn_sham_matrix 119 9.3 0.016 0.043 6.429 6.456 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.690 5.715 dbcsr_mm_accdrv_process 38222 16.0 4.103 5.092 4.712 5.523 sum_up_and_integrate 119 10.3 0.030 0.031 4.026 4.030 integrate_v_rspace 119 11.3 0.002 0.003 3.996 4.003 mp_waitall_1 158411 16.6 2.861 3.800 2.861 3.800 init_scf_run 11 5.9 0.000 0.001 3.492 3.493 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.492 3.492 qs_ot_get_p 119 10.4 0.001 0.001 3.356 3.381 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.740 3.372 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.219 3.222 calculate_rho_elec 119 8.7 0.030 0.031 3.218 3.222 init_scf_loop 11 6.9 0.001 0.006 2.660 2.660 rs_pw_transfer 974 11.9 0.009 0.009 2.329 2.453 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.068 2.385 apply_single 119 13.6 0.000 0.000 2.068 2.385 calculate_first_density_matrix 1 7.0 0.001 0.004 2.273 2.274 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.249 2.255 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.433 2.233 make_m2s 4572 13.5 0.046 0.047 1.912 2.077 make_images 4572 14.5 0.189 0.202 1.825 1.989 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.977 1.978 density_rs2pw 119 9.7 0.004 0.004 1.810 1.918 grid_integrate_task_list 119 12.3 1.803 1.883 1.803 1.883 jit_kernel_multiply 10 16.1 0.558 1.859 0.558 1.859 calculate_dm_sparse 119 9.5 0.001 0.001 1.830 1.847 prepare_preconditioner 11 7.9 0.000 0.000 1.836 1.844 make_preconditioner 11 8.9 0.001 0.008 1.836 1.844 potential_pw2rs 119 12.3 0.007 0.008 1.806 1.817 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.686 1.778 pw_transfer 1439 11.6 0.063 0.067 1.729 1.741 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.723 1.724 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.721 1.722 cp_fm_redistribute_end 50 14.0 1.284 1.701 1.285 1.701 cp_fm_diag_elpa_base 50 14.0 0.399 1.618 0.414 1.662 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.638 1.650 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.608 1.616 multiply_cannon_sync_h2d 18288 15.5 1.376 1.607 1.376 1.607 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.553 1.557 mp_sum_l 7207 12.9 1.193 1.511 1.193 1.511 ot_diis_step 108 11.5 0.011 0.011 1.509 1.510 fft3d_ps 1201 14.6 0.507 0.522 1.326 1.339 grid_collocate_task_list 119 9.7 1.204 1.310 1.204 1.310 fft_wrap_pw1pw2_140 487 13.2 0.086 0.090 1.290 1.302 acc_transpose_blocks 18288 15.5 0.075 0.077 1.227 1.255 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.185 1.191 wfi_extrapolate 11 7.9 0.001 0.001 1.171 1.171 multiply_cannon_metrocomm1 18288 15.5 0.028 0.029 0.440 1.063 make_images_data 4572 15.5 0.044 0.049 0.877 1.059 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.953 0.954 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.751 0.933 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.902 0.926 mp_alltoall_d11v 2130 13.8 0.726 0.834 0.726 0.834 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.810 0.814 acc_transpose_blocks_kernels 18288 16.5 0.209 0.219 0.787 0.801 cp_fm_cholesky_invert 11 10.9 0.774 0.778 0.774 0.778 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.664 0.736 mp_alltoall_z22v 1201 16.6 0.663 0.728 0.663 0.728 mp_allgather_i34 2286 14.5 0.276 0.691 0.276 0.691 mp_waitany 9880 13.7 0.529 0.661 0.529 0.661 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.694000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.909091, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 559.484928E+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 0.015 0.033 35.388 35.389 qs_mol_dyn_low 1 2.0 0.003 0.004 35.214 35.223 qs_forces 11 3.9 0.002 0.002 35.154 35.156 qs_energies 11 4.9 0.002 0.006 33.440 33.446 scf_env_do_scf 11 5.9 0.001 0.004 28.369 28.371 scf_env_do_scf_inner_loop 108 6.5 0.003 0.012 24.806 24.807 dbcsr_multiply_generic 2286 12.5 0.099 0.101 18.243 18.359 velocity_verlet 10 3.0 0.002 0.002 18.016 18.018 qs_scf_new_mos 108 7.5 0.001 0.001 16.488 16.541 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.487 16.541 ot_scf_mini 108 9.5 0.003 0.004 15.541 15.595 multiply_cannon 2286 13.5 0.233 0.271 14.486 14.892 multiply_cannon_loop 2286 14.5 0.940 0.970 13.492 13.909 ot_mini 108 10.5 0.001 0.001 9.398 9.467 multiply_cannon_multrec 27432 15.5 2.323 3.012 8.579 8.931 qs_ot_get_derivative 108 11.5 0.001 0.001 7.564 7.620 dbcsr_mm_accdrv_process 47916 15.9 5.254 7.091 6.164 7.343 rebuild_ks_matrix 119 8.3 0.000 0.000 6.586 6.634 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.024 6.585 6.633 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.855 5.898 sum_up_and_integrate 119 10.3 0.035 0.037 3.853 3.860 integrate_v_rspace 119 11.3 0.002 0.003 3.818 3.825 init_scf_run 11 5.9 0.000 0.001 3.655 3.656 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.655 3.655 qs_ot_get_p 119 10.4 0.001 0.001 3.492 3.567 init_scf_loop 11 6.9 0.001 0.004 3.537 3.537 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.388 3.421 calculate_rho_elec 119 8.7 0.040 0.046 3.388 3.420 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.855 3.256 prepare_preconditioner 11 7.9 0.000 0.000 2.672 2.682 make_preconditioner 11 8.9 0.001 0.009 2.672 2.682 mp_waitall_1 137007 16.6 2.024 2.648 2.024 2.648 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.279 2.607 make_m2s 4572 13.5 0.055 0.057 2.323 2.457 make_images 4572 14.5 0.271 0.332 2.213 2.346 rs_pw_transfer 974 11.9 0.009 0.009 2.142 2.263 calculate_first_density_matrix 1 7.0 0.001 0.002 2.246 2.249 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.065 2.207 apply_single 119 13.6 0.000 0.000 2.065 2.207 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.172 2.181 calculate_dm_sparse 119 9.5 0.000 0.000 2.045 2.105 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.980 2.007 density_rs2pw 119 9.7 0.004 0.004 1.869 1.983 grid_integrate_task_list 119 12.3 1.828 1.899 1.828 1.899 pw_transfer 1439 11.6 0.063 0.067 1.835 1.868 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.849 1.850 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.839 1.840 jit_kernel_multiply 10 15.5 0.850 1.822 0.850 1.822 ot_diis_step 108 11.5 0.012 0.012 1.793 1.793 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.745 1.780 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.628 1.639 potential_pw2rs 119 12.3 0.008 0.009 1.630 1.635 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.594 1.595 cp_fm_redistribute_end 50 14.0 1.054 1.565 1.055 1.566 cp_fm_diag_elpa_base 50 14.0 0.488 1.495 0.508 1.535 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.798 1.490 acc_transpose_blocks 27432 15.5 0.112 0.115 1.456 1.478 fft3d_ps 1201 14.6 0.537 0.589 1.426 1.449 fft_wrap_pw1pw2_140 487 13.2 0.084 0.093 1.400 1.438 wfi_extrapolate 11 7.9 0.001 0.001 1.360 1.360 mp_sum_l 7207 12.9 1.013 1.332 1.013 1.332 grid_collocate_task_list 119 9.7 1.220 1.305 1.220 1.305 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.219 1.229 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.152 1.153 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.130 1.147 cp_fm_upper_to_full 72 13.5 0.800 1.109 0.800 1.109 dbcsr_complete_redistribute 329 12.2 0.124 0.163 0.821 1.097 make_images_data 4572 15.5 0.045 0.048 0.925 1.071 multiply_cannon_sync_h2d 27432 15.5 0.989 1.054 0.989 1.054 hybrid_alltoall_any 4725 16.4 0.061 0.151 0.794 0.988 mp_alltoall_d11v 2130 13.8 0.785 0.900 0.785 0.900 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.872 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.600 0.869 cp_fm_cholesky_invert 11 10.9 0.842 0.845 0.842 0.845 acc_transpose_blocks_kernels 27432 16.5 0.268 0.278 0.829 0.837 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.821 0.825 mp_alltoall_z22v 1201 16.6 0.771 0.800 0.771 0.800 mp_alltoall_i22 627 13.8 0.437 0.735 0.437 0.735 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.389000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=532.181818, yerr=3.270201 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 594.714624E+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 0.018 0.034 30.061 30.062 qs_mol_dyn_low 1 2.0 0.003 0.003 29.875 29.883 qs_forces 11 3.9 0.003 0.004 29.812 29.813 qs_energies 11 4.9 0.001 0.001 28.044 28.046 scf_env_do_scf 11 5.9 0.001 0.002 23.193 23.193 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 20.256 20.257 velocity_verlet 10 3.0 0.002 0.002 15.126 15.128 dbcsr_multiply_generic 2286 12.5 0.092 0.094 12.857 12.956 qs_scf_new_mos 108 7.5 0.001 0.001 12.229 12.259 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.228 12.258 ot_scf_mini 108 9.5 0.002 0.002 11.537 11.563 multiply_cannon 2286 13.5 0.230 0.235 9.906 10.489 multiply_cannon_loop 2286 14.5 0.330 0.343 8.925 9.145 ot_mini 108 10.5 0.001 0.001 6.338 6.372 rebuild_ks_matrix 119 8.3 0.000 0.000 6.256 6.286 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.256 6.285 multiply_cannon_multrec 9144 15.5 1.613 2.034 5.766 6.148 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.581 5.607 qs_ot_get_derivative 108 11.5 0.001 0.001 5.010 5.037 dbcsr_mm_accdrv_process 12550 15.8 2.966 4.068 4.051 4.138 sum_up_and_integrate 119 10.3 0.037 0.041 3.849 3.868 integrate_v_rspace 119 11.3 0.003 0.003 3.812 3.831 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.418 3.425 calculate_rho_elec 119 8.7 0.059 0.061 3.418 3.424 qs_ot_get_p 119 10.4 0.001 0.001 3.346 3.389 init_scf_run 11 5.9 0.000 0.001 3.356 3.357 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.356 3.356 init_scf_loop 11 6.9 0.000 0.000 2.911 2.912 mp_waitall_1 115863 16.7 2.047 2.598 2.047 2.598 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.237 2.239 make_m2s 4572 13.5 0.036 0.037 1.987 2.138 rs_pw_transfer 974 11.9 0.008 0.008 2.007 2.101 calculate_first_density_matrix 1 7.0 0.001 0.001 2.099 2.100 make_images 4572 14.5 0.266 0.301 1.895 2.046 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.017 2.018 prepare_preconditioner 11 7.9 0.000 0.000 2.009 2.016 make_preconditioner 11 8.9 0.000 0.000 2.009 2.016 density_rs2pw 119 9.7 0.003 0.004 1.854 1.953 grid_integrate_task_list 119 12.3 1.850 1.935 1.850 1.935 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.894 1.928 pw_transfer 1439 11.6 0.064 0.065 1.874 1.897 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.783 1.806 calculate_dm_sparse 119 9.5 0.000 0.000 1.725 1.742 jit_kernel_multiply 10 15.5 1.047 1.741 1.047 1.741 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.709 1.709 cp_fm_redistribute_end 50 14.0 0.852 1.676 0.853 1.677 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.651 1.669 cp_fm_diag_elpa_base 50 14.0 0.777 1.593 0.822 1.651 potential_pw2rs 119 12.3 0.010 0.010 1.588 1.609 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.579 1.580 fft3d_ps 1201 14.6 0.539 0.551 1.451 1.473 fft_wrap_pw1pw2_140 487 13.2 0.081 0.084 1.384 1.408 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.373 1.381 grid_collocate_task_list 119 9.7 1.277 1.361 1.277 1.361 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.299 1.316 ot_diis_step 108 11.5 0.012 0.013 1.308 1.309 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.264 1.264 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.167 1.197 apply_single 119 13.6 0.000 0.000 1.166 1.197 wfi_extrapolate 11 7.9 0.001 0.001 1.189 1.189 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.135 1.142 make_images_data 4572 15.5 0.039 0.042 0.909 1.132 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.827 1.099 cp_fm_cholesky_invert 11 10.9 1.062 1.066 1.062 1.066 mp_alltoall_d11v 2130 13.8 0.849 0.987 0.849 0.987 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.922 acc_transpose_blocks 9144 15.5 0.038 0.039 0.895 0.904 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.450 0.885 mp_alltoall_z22v 1201 16.6 0.786 0.815 0.786 0.815 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.797 0.806 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.802 0.805 multiply_cannon_sync_h2d 9144 15.5 0.706 0.778 0.706 0.778 mp_allgather_i34 2286 14.5 0.316 0.764 0.316 0.764 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.702 0.749 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.498 0.745 mp_sum_l 7207 12.9 0.542 0.680 0.542 0.680 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 0.653 0.659 dbcsr_complete_redistribute 329 12.2 0.172 0.179 0.627 0.657 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.062000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=565.181818, yerr=2.822577 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 739.094528E+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 0.029 0.068 40.856 40.856 qs_mol_dyn_low 1 2.0 0.003 0.004 40.607 40.613 qs_forces 11 3.9 0.016 0.032 40.543 40.544 qs_energies 11 4.9 0.001 0.002 38.575 38.590 scf_env_do_scf 11 5.9 0.001 0.005 32.894 32.894 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.983 24.984 velocity_verlet 10 3.0 0.002 0.002 23.111 23.116 dbcsr_multiply_generic 2286 12.5 0.100 0.101 17.568 17.755 qs_scf_new_mos 108 7.5 0.001 0.001 16.013 16.102 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.012 16.101 ot_scf_mini 108 9.5 0.002 0.002 14.922 15.016 multiply_cannon 2286 13.5 0.304 0.311 13.675 14.586 multiply_cannon_loop 2286 14.5 0.344 0.350 12.389 13.294 ot_mini 108 10.5 0.001 0.001 8.858 8.970 multiply_cannon_multrec 9144 15.5 3.367 4.683 8.523 8.635 init_scf_loop 11 6.9 0.000 0.000 7.880 7.882 rebuild_ks_matrix 119 8.3 0.000 0.000 6.918 7.056 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.918 7.055 prepare_preconditioner 11 7.9 0.000 0.000 6.931 6.945 make_preconditioner 11 8.9 0.000 0.000 6.931 6.945 qs_ot_get_derivative 108 11.5 0.001 0.001 6.811 6.904 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.490 6.824 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.245 6.368 dbcsr_mm_accdrv_process 12550 15.8 4.115 5.602 5.032 6.336 cp_fm_upper_to_full 72 14.2 3.177 4.563 3.177 4.563 sum_up_and_integrate 119 10.3 0.064 0.065 3.733 3.740 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.705 3.711 calculate_rho_elec 119 8.7 0.118 0.121 3.705 3.710 integrate_v_rspace 119 11.3 0.003 0.003 3.668 3.676 init_scf_run 11 5.9 0.000 0.001 3.650 3.650 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.650 3.650 mp_waitall_1 94719 16.7 2.520 3.540 2.520 3.540 qs_ot_get_p 119 10.4 0.001 0.001 3.296 3.425 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.519 2.908 dbcsr_complete_redistribute 329 12.2 0.286 0.293 2.003 2.795 make_m2s 4572 13.5 0.039 0.040 2.346 2.541 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.176 2.477 apply_single 119 13.6 0.000 0.000 2.175 2.476 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.674 2.466 make_images 4572 14.5 0.350 0.382 2.223 2.419 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.415 2.314 mp_alltoall_i22 627 13.8 1.477 2.279 1.477 2.279 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.436 2.220 calculate_first_density_matrix 1 7.0 0.000 0.001 2.214 2.215 calculate_dm_sparse 119 9.5 0.000 0.000 2.176 2.197 pw_transfer 1439 11.6 0.067 0.067 2.088 2.091 grid_integrate_task_list 119 12.3 2.017 2.035 2.017 2.035 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.028 2.029 ot_diis_step 108 11.5 0.014 0.014 2.014 2.014 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.992 1.994 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.917 1.918 density_rs2pw 119 9.7 0.003 0.003 1.799 1.819 mp_sum_l 7207 12.9 1.036 1.781 1.036 1.781 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.768 1.780 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.704 1.752 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.748 1.749 jit_kernel_multiply 10 15.5 0.891 1.731 0.891 1.731 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.654 1.658 fft3d_ps 1201 14.6 0.567 0.576 1.627 1.631 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.600 1.622 cp_fm_cholesky_invert 11 10.9 1.515 1.519 1.515 1.519 grid_collocate_task_list 119 9.7 1.457 1.475 1.457 1.475 rs_pw_transfer 974 11.9 0.009 0.009 1.425 1.468 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.157 1.448 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.443 1.444 cp_fm_diag_elpa_base 50 14.0 1.297 1.350 1.440 1.440 make_images_data 4572 15.5 0.043 0.046 1.127 1.380 wfi_extrapolate 11 7.9 0.001 0.001 1.377 1.377 potential_pw2rs 119 12.3 0.014 0.014 1.253 1.256 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.232 1.240 mp_alltoall_d11v 2130 13.8 1.172 1.190 1.172 1.190 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.158 1.178 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.103 1.125 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 0.985 1.018 qs_create_task_list 11 7.9 0.000 0.000 0.944 0.955 generate_qs_task_list 11 8.9 0.372 0.391 0.943 0.954 mp_alltoall_z22v 1201 16.6 0.923 0.946 0.923 0.946 acc_transpose_blocks 9144 15.5 0.038 0.038 0.911 0.924 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.884 0.897 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.856000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=695.363636, yerr=11.601510 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 500.178944E+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 1383689. 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 0.017 0.038 83.171 83.173 qs_mol_dyn_low 1 2.0 0.003 0.004 82.874 82.883 qs_forces 11 3.9 0.002 0.002 82.804 82.805 qs_energies 11 4.9 0.002 0.006 79.995 80.006 scf_env_do_scf 11 5.9 0.000 0.001 71.206 71.208 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.545 65.545 dbcsr_multiply_generic 2055 12.4 0.107 0.112 52.278 52.677 qs_scf_new_mos 99 7.5 0.000 0.001 48.553 48.724 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.552 48.724 ot_scf_mini 99 9.5 0.002 0.003 46.129 46.209 multiply_cannon 2055 13.4 0.177 0.184 42.621 43.693 multiply_cannon_loop 2055 14.4 1.555 1.606 41.650 42.766 velocity_verlet 10 3.0 0.001 0.002 41.993 41.993 ot_mini 99 10.5 0.001 0.001 27.775 27.875 qs_ot_get_derivative 99 11.5 0.001 0.001 20.914 21.030 multiply_cannon_multrec 49320 15.4 12.139 12.837 17.214 17.910 rebuild_ks_matrix 110 8.3 0.000 0.001 14.434 14.666 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.433 14.666 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.690 12.902 mp_waitall_1 220248 16.4 11.582 12.611 11.582 12.611 multiply_cannon_sync_h2d 49320 15.4 9.984 10.505 9.984 10.505 qs_ot_get_p 110 10.4 0.001 0.001 9.742 9.858 multiply_cannon_metrocomm3 49320 15.4 0.078 0.082 6.961 8.516 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.167 7.709 apply_single 110 13.6 0.000 0.000 7.167 7.709 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.065 7.595 sum_up_and_integrate 110 10.3 0.036 0.043 7.043 7.056 integrate_v_rspace 110 11.3 0.002 0.003 7.007 7.030 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.835 6.900 init_scf_run 11 5.9 0.000 0.001 6.754 6.754 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.754 6.754 ot_diis_step 99 11.5 0.006 0.006 6.576 6.577 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.425 6.460 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.245 6.391 calculate_rho_elec 110 8.6 0.020 0.024 6.244 6.390 mp_sum_l 6514 12.8 5.296 6.203 5.296 6.203 init_scf_loop 11 6.9 0.001 0.003 5.631 5.632 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.438 5.438 dbcsr_mm_accdrv_process 87628 16.1 2.078 2.166 4.954 5.250 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.953 4.954 cp_fm_redistribute_end 48 14.0 4.324 4.928 4.328 4.929 cp_fm_diag_elpa_base 48 14.0 0.595 4.786 0.599 4.809 make_m2s 4110 13.4 0.062 0.067 3.919 4.020 wfi_extrapolate 11 7.9 0.001 0.001 3.984 3.984 calculate_dm_sparse 110 9.5 0.000 0.001 3.844 3.964 rs_pw_transfer 902 11.9 0.012 0.013 3.790 3.935 make_images 4110 14.4 0.177 0.192 3.821 3.925 multiply_cannon_metrocomm1 49320 15.4 0.059 0.063 2.683 3.785 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.542 3.547 prepare_preconditioner 11 7.9 0.000 0.000 3.478 3.499 make_preconditioner 11 8.9 0.000 0.001 3.478 3.499 density_rs2pw 110 9.6 0.004 0.004 3.278 3.436 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.355 3.424 grid_integrate_task_list 110 12.3 3.245 3.389 3.245 3.389 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.237 3.286 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.203 3.255 pw_transfer 1331 11.6 0.054 0.066 2.854 2.915 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.766 2.831 calculate_first_density_matrix 1 7.0 0.001 0.003 2.684 2.689 potential_pw2rs 110 12.3 0.006 0.007 2.631 2.661 mp_alltoall_d11v 2046 13.8 2.200 2.637 2.200 2.637 jit_kernel_multiply 13 15.9 2.592 2.609 2.592 2.609 fft_wrap_pw1pw2_140 451 13.1 0.169 0.190 2.328 2.394 fft3d_ps 1111 14.6 0.751 0.835 2.304 2.358 acc_transpose_blocks 49320 15.4 0.225 0.236 2.189 2.254 grid_collocate_task_list 110 9.6 2.089 2.216 2.089 2.216 mp_sum_d 3879 11.9 1.518 2.121 1.518 2.121 mp_waitany 14300 13.8 1.854 2.120 1.854 2.120 make_images_data 4110 15.4 0.042 0.045 1.780 1.907 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.885 1.903 cp_fm_cholesky_invert 11 10.9 1.864 1.867 1.864 1.867 hybrid_alltoall_any 4261 16.3 0.081 0.480 1.567 1.816 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.696 1.723 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.173000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.181818, yerr=1.991718 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 585.809920E+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 2993538. 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 0.014 0.031 70.573 70.574 qs_mol_dyn_low 1 2.0 0.003 0.006 70.321 70.331 qs_forces 11 3.9 0.003 0.006 70.247 70.248 qs_energies 11 4.9 0.001 0.002 66.877 66.882 scf_env_do_scf 11 5.9 0.000 0.001 58.225 58.228 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 50.037 50.038 dbcsr_multiply_generic 2055 12.4 0.114 0.119 38.404 38.619 velocity_verlet 10 3.0 0.001 0.002 37.245 37.247 qs_scf_new_mos 99 7.5 0.001 0.001 33.588 33.734 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.587 33.733 multiply_cannon 2055 13.4 0.222 0.245 31.296 32.414 ot_scf_mini 99 9.5 0.003 0.003 31.908 32.040 multiply_cannon_loop 2055 14.4 0.930 0.951 29.903 30.868 ot_mini 99 10.5 0.002 0.016 18.774 18.903 multiply_cannon_multrec 24660 15.4 7.604 9.319 13.792 15.469 rebuild_ks_matrix 110 8.3 0.000 0.001 14.022 14.130 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 14.021 14.130 qs_ot_get_derivative 99 11.5 0.001 0.001 12.882 13.016 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.371 12.467 mp_waitall_1 176588 16.5 8.143 10.522 8.143 10.522 init_scf_loop 11 6.9 0.000 0.000 8.145 8.146 multiply_cannon_sync_h2d 24660 15.4 7.067 8.004 7.067 8.004 multiply_cannon_metrocomm3 24660 15.4 0.069 0.072 5.338 7.873 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.592 7.293 apply_single 110 13.6 0.000 0.001 6.592 7.293 sum_up_and_integrate 110 10.3 0.053 0.058 6.846 6.857 integrate_v_rspace 110 11.3 0.002 0.002 6.793 6.806 qs_ot_get_p 110 10.4 0.001 0.001 6.409 6.591 dbcsr_mm_accdrv_process 52282 16.1 4.721 5.624 6.031 6.319 init_scf_run 11 5.9 0.000 0.001 6.265 6.266 scf_env_initial_rho_setup 11 6.9 0.006 0.007 6.265 6.266 prepare_preconditioner 11 7.9 0.000 0.000 6.070 6.085 make_preconditioner 11 8.9 0.000 0.000 6.070 6.085 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.070 6.079 calculate_rho_elec 110 8.6 0.039 0.047 6.070 6.079 ot_diis_step 99 11.5 0.011 0.017 5.831 5.833 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.626 5.784 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.820 5.545 make_m2s 4110 13.4 0.058 0.062 4.490 4.917 make_images 4110 14.4 0.397 0.441 4.378 4.801 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.370 4.393 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.914 3.915 pw_transfer 1331 11.6 0.066 0.072 3.570 3.733 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.463 3.627 density_rs2pw 110 9.6 0.004 0.004 3.334 3.591 wfi_extrapolate 11 7.9 0.001 0.001 3.564 3.564 rs_pw_transfer 902 11.9 0.012 0.013 3.261 3.549 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.382 3.456 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.368 3.370 grid_integrate_task_list 110 12.3 3.145 3.319 3.145 3.319 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.288 3.290 cp_fm_redistribute_end 48 14.0 2.454 3.252 2.456 3.253 cp_fm_diag_elpa_base 48 14.0 0.762 3.097 0.793 3.185 calculate_dm_sparse 110 9.5 0.001 0.001 3.001 3.039 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.007 3.036 fft3d_ps 1111 14.6 1.074 1.267 2.821 2.973 fft_wrap_pw1pw2_140 451 13.1 0.199 0.217 2.807 2.967 make_images_data 4110 15.4 0.047 0.051 2.494 2.927 cp_fm_cholesky_invert 11 10.9 2.864 2.872 2.864 2.872 hybrid_alltoall_any 4261 16.3 0.101 0.440 2.210 2.866 potential_pw2rs 110 12.3 0.008 0.009 2.653 2.677 mp_sum_l 6514 12.8 1.999 2.624 1.999 2.624 calculate_first_density_matrix 1 7.0 0.003 0.004 2.597 2.599 grid_collocate_task_list 110 9.6 2.118 2.290 2.118 2.290 mp_alltoall_d11v 2046 13.8 1.991 2.238 1.991 2.238 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.015 2.034 jit_kernel_multiply 10 16.3 0.959 1.944 0.959 1.944 multiply_cannon_metrocomm4 22605 15.4 0.072 0.076 0.777 1.872 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.856 1.857 mp_allgather_i34 2055 14.4 0.716 1.830 0.716 1.830 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.816 1.826 mp_irecv_dv 57340 16.2 0.656 1.656 0.656 1.656 acc_transpose_blocks 24660 15.4 0.110 0.112 1.546 1.576 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.563 1.572 mp_alltoall_z22v 1111 16.6 1.462 1.540 1.462 1.540 cp_fm_cholesky_decompose 22 10.9 1.521 1.528 1.521 1.528 dbcsr_complete_redistribute 325 12.2 0.238 0.297 1.238 1.513 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.369 1.479 mp_waitany 10164 13.8 1.208 1.445 1.208 1.445 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.574000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.090909, yerr=6.881524 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 659.898368E+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 0.024 0.042 63.175 63.176 qs_mol_dyn_low 1 2.0 0.003 0.004 62.824 62.834 qs_forces 11 3.9 0.002 0.004 62.345 62.345 qs_energies 11 4.9 0.002 0.006 59.100 59.103 scf_env_do_scf 11 5.9 0.001 0.001 51.111 51.112 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 41.935 41.936 velocity_verlet 10 3.0 0.002 0.002 34.209 34.224 dbcsr_multiply_generic 2055 12.4 0.110 0.124 29.225 29.564 qs_scf_new_mos 99 7.5 0.001 0.001 26.912 27.024 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.911 27.023 ot_scf_mini 99 9.5 0.003 0.003 25.650 25.788 multiply_cannon 2055 13.4 0.210 0.222 22.324 23.731 multiply_cannon_loop 2055 14.4 0.617 0.635 21.046 21.986 ot_mini 99 10.5 0.001 0.001 14.243 14.385 rebuild_ks_matrix 110 8.3 0.000 0.000 12.539 12.716 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.538 12.716 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.035 11.196 mp_waitall_1 139946 16.5 7.782 10.945 7.782 10.945 multiply_cannon_multrec 16440 15.4 3.751 4.525 9.473 10.165 qs_ot_get_derivative 99 11.5 0.001 0.001 9.642 9.778 init_scf_loop 11 6.9 0.001 0.002 9.135 9.138 multiply_cannon_metrocomm3 16440 15.4 0.043 0.045 4.772 7.742 prepare_preconditioner 11 7.9 0.000 0.000 7.316 7.335 make_preconditioner 11 8.9 0.000 0.001 7.316 7.335 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.653 7.009 sum_up_and_integrate 110 10.3 0.061 0.062 6.709 6.729 integrate_v_rspace 110 11.3 0.002 0.003 6.648 6.667 qs_ot_get_p 110 10.4 0.001 0.001 6.450 6.634 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.805 5.816 calculate_rho_elec 110 8.6 0.058 0.058 5.804 5.816 dbcsr_mm_accdrv_process 34862 16.1 4.560 5.269 5.580 5.790 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.086 5.579 apply_single 110 13.6 0.000 0.000 5.085 5.579 init_scf_run 11 5.9 0.000 0.001 5.554 5.554 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.554 5.554 make_m2s 4110 13.4 0.051 0.053 4.490 4.886 make_images 4110 14.4 0.390 0.503 4.373 4.767 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.581 4.589 ot_diis_step 99 11.5 0.010 0.011 4.554 4.555 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.193 4.196 multiply_cannon_sync_h2d 16440 15.4 3.569 4.146 3.569 4.146 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.289 3.963 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.570 3.573 cp_fm_redistribute_end 48 14.0 2.220 3.525 2.223 3.526 cp_fm_diag_elpa_base 48 14.0 1.235 3.358 1.295 3.465 pw_transfer 1331 11.6 0.065 0.072 3.422 3.443 grid_integrate_task_list 110 12.3 3.178 3.419 3.178 3.419 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.315 3.338 density_rs2pw 110 9.6 0.004 0.004 3.039 3.235 rs_pw_transfer 902 11.9 0.011 0.012 2.953 3.125 make_images_data 4110 15.4 0.043 0.048 2.599 3.058 wfi_extrapolate 11 7.9 0.001 0.001 3.055 3.055 cp_fm_cholesky_invert 11 10.9 3.027 3.035 3.027 3.035 hybrid_alltoall_any 4261 16.3 0.105 0.374 2.332 2.939 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.804 2.805 fft_wrap_pw1pw2_140 451 13.1 0.208 0.211 2.774 2.797 fft3d_ps 1111 14.6 1.065 1.076 2.632 2.652 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.913 2.599 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.521 2.594 calculate_dm_sparse 110 9.5 0.001 0.001 2.546 2.579 potential_pw2rs 110 12.3 0.011 0.011 2.490 2.513 mp_sum_l 6514 12.8 1.803 2.497 1.803 2.497 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.434 2.491 mp_irecv_dv 48980 15.7 0.843 2.469 0.843 2.469 calculate_first_density_matrix 1 7.0 0.001 0.002 2.411 2.413 grid_collocate_task_list 110 9.6 2.174 2.321 2.174 2.321 mp_alltoall_d11v 2046 13.8 2.013 2.251 2.013 2.251 dbcsr_complete_redistribute 325 12.2 0.315 0.340 1.566 2.038 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.958 1.958 cp_fm_upper_to_full 70 13.6 1.391 1.868 1.391 1.868 mp_allgather_i34 2055 14.4 0.607 1.768 0.607 1.768 cp_fm_cholesky_decompose 22 10.9 1.739 1.760 1.739 1.760 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.706 1.727 jit_kernel_multiply 8 16.4 0.632 1.564 0.632 1.564 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.476 1.492 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.026 1.481 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.361 1.468 mp_alltoall_z22v 1111 16.6 1.353 1.382 1.353 1.382 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.366 1.373 mp_waitany 17072 13.8 1.147 1.364 1.147 1.364 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.176000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.272727, yerr=9.845081 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 731.308032E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.038 66.930 66.931 qs_mol_dyn_low 1 2.0 0.003 0.004 66.622 66.631 qs_forces 11 3.9 0.003 0.004 66.553 66.554 qs_energies 11 4.9 0.005 0.025 63.127 63.130 scf_env_do_scf 11 5.9 0.001 0.001 54.748 54.751 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 42.585 42.585 velocity_verlet 10 3.0 0.002 0.002 37.644 37.647 dbcsr_multiply_generic 2055 12.4 0.116 0.118 29.997 30.241 qs_scf_new_mos 99 7.5 0.001 0.001 27.847 27.958 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.847 27.958 ot_scf_mini 99 9.5 0.003 0.003 26.194 26.306 multiply_cannon 2055 13.4 0.242 0.262 22.709 23.695 multiply_cannon_loop 2055 14.4 0.889 0.911 21.218 21.851 ot_mini 99 10.5 0.001 0.001 14.703 14.839 multiply_cannon_multrec 24660 15.4 4.212 6.660 12.642 13.894 rebuild_ks_matrix 110 8.3 0.000 0.000 12.152 12.251 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.152 12.250 init_scf_loop 11 6.9 0.002 0.011 12.116 12.117 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.740 10.826 qs_ot_get_derivative 99 11.5 0.001 0.001 10.495 10.611 prepare_preconditioner 11 7.9 0.000 0.000 10.360 10.377 make_preconditioner 11 8.9 0.000 0.001 10.360 10.377 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.577 10.004 dbcsr_mm_accdrv_process 52304 16.0 6.767 7.980 8.284 9.240 mp_waitall_1 121746 16.5 4.743 6.853 4.743 6.853 qs_ot_get_p 110 10.4 0.001 0.001 6.424 6.584 sum_up_and_integrate 110 10.3 0.067 0.071 6.533 6.548 integrate_v_rspace 110 11.3 0.002 0.003 6.466 6.483 make_m2s 4110 13.4 0.061 0.064 5.590 5.978 make_images 4110 14.4 0.573 0.697 5.446 5.828 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.800 5.809 calculate_rho_elec 110 8.6 0.077 0.081 5.799 5.809 init_scf_run 11 5.9 0.000 0.001 5.569 5.570 scf_env_initial_rho_setup 11 6.9 0.001 0.003 5.568 5.570 cp_fm_upper_to_full 70 13.8 3.237 4.604 3.237 4.604 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.494 4.507 ot_diis_step 99 11.5 0.011 0.011 4.169 4.169 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.065 4.155 apply_single 110 13.6 0.000 0.000 4.064 4.154 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.994 3.995 dbcsr_complete_redistribute 325 12.2 0.418 0.463 2.776 3.891 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.653 3.591 pw_transfer 1331 11.6 0.065 0.073 3.414 3.449 grid_integrate_task_list 110 12.3 3.277 3.439 3.277 3.439 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.379 3.434 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.347 3.349 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.308 3.347 multiply_cannon_sync_h2d 24660 15.4 3.185 3.343 3.185 3.343 cp_fm_redistribute_end 48 14.0 1.673 3.309 1.675 3.310 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.203 3.303 cp_fm_diag_elpa_base 48 14.0 1.553 3.162 1.632 3.272 make_images_data 4110 15.4 0.046 0.050 2.819 3.239 hybrid_alltoall_any 4261 16.3 0.120 0.459 2.466 3.228 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.118 3.176 density_rs2pw 110 9.6 0.004 0.004 2.926 3.127 cp_fm_cholesky_invert 11 10.9 3.047 3.055 3.047 3.055 wfi_extrapolate 11 7.9 0.001 0.001 3.052 3.052 calculate_dm_sparse 110 9.5 0.001 0.001 2.946 2.971 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.774 2.858 rs_pw_transfer 902 11.9 0.010 0.011 2.614 2.832 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.814 2.817 fft_wrap_pw1pw2_140 451 13.1 0.202 0.216 2.755 2.799 mp_alltoall_i22 605 13.7 1.642 2.784 1.642 2.784 fft3d_ps 1111 14.6 1.064 1.104 2.612 2.633 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.414 2.448 calculate_first_density_matrix 1 7.0 0.002 0.010 2.427 2.431 grid_collocate_task_list 110 9.6 2.224 2.344 2.224 2.344 potential_pw2rs 110 12.3 0.012 0.013 2.262 2.275 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.270 2.271 mp_alltoall_d11v 2046 13.8 2.013 2.263 2.013 2.263 jit_kernel_multiply 11 15.7 1.187 1.944 1.187 1.944 cp_fm_cholesky_decompose 22 10.9 1.828 1.880 1.828 1.880 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.748 1.776 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.602 1.703 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.661 1.672 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.539 1.555 acc_transpose_blocks 24660 15.4 0.106 0.108 1.502 1.528 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.845 1.479 mp_sum_l 6514 12.8 1.028 1.473 1.028 1.473 mp_irecv_dv 62702 16.1 0.746 1.402 0.746 1.402 mp_allgather_i34 2055 14.4 0.602 1.395 0.602 1.395 mp_alltoall_z22v 1111 16.6 1.331 1.393 1.331 1.393 mp_waitany 13376 13.8 1.079 1.354 1.079 1.354 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.931000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=693.181818, yerr=9.083633 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 833.003520E+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 0.018 0.037 55.180 55.181 qs_mol_dyn_low 1 2.0 0.003 0.003 54.938 54.947 qs_forces 11 3.9 0.002 0.002 54.873 54.873 qs_energies 11 4.9 0.001 0.002 51.243 51.249 scf_env_do_scf 11 5.9 0.000 0.001 43.054 43.054 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.096 35.097 velocity_verlet 10 3.0 0.003 0.004 31.001 31.004 dbcsr_multiply_generic 2055 12.4 0.105 0.107 22.926 23.144 qs_scf_new_mos 99 7.5 0.001 0.001 20.907 20.965 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.906 20.964 ot_scf_mini 99 9.5 0.002 0.002 19.651 19.698 multiply_cannon 2055 13.4 0.249 0.264 17.385 18.662 multiply_cannon_loop 2055 14.4 0.320 0.332 16.018 16.437 rebuild_ks_matrix 110 8.3 0.000 0.000 11.517 11.560 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.517 11.559 ot_mini 99 10.5 0.001 0.001 10.557 10.598 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.255 10.292 multiply_cannon_multrec 8220 15.4 3.171 4.955 7.382 8.730 init_scf_loop 11 6.9 0.000 0.000 7.909 7.910 mp_waitall_1 103326 16.6 6.267 7.876 6.267 7.876 qs_ot_get_derivative 99 11.5 0.001 0.001 6.695 6.741 prepare_preconditioner 11 7.9 0.000 0.000 6.280 6.284 make_preconditioner 11 8.9 0.000 0.000 6.280 6.284 sum_up_and_integrate 110 10.3 0.080 0.082 6.238 6.251 integrate_v_rspace 110 11.3 0.003 0.003 6.158 6.170 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.846 5.921 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.658 5.678 calculate_rho_elec 110 8.6 0.115 0.116 5.657 5.678 qs_ot_get_p 110 10.4 0.001 0.001 5.147 5.204 init_scf_run 11 5.9 0.000 0.001 5.073 5.074 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.073 5.073 dbcsr_mm_accdrv_process 17442 15.9 2.825 3.939 4.082 4.987 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.203 4.615 make_m2s 4110 13.4 0.040 0.041 4.231 4.460 make_images 4110 14.4 0.633 0.688 4.099 4.327 ot_diis_step 99 11.5 0.012 0.012 3.837 3.838 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.741 3.804 apply_single 110 13.6 0.000 0.000 3.740 3.803 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.736 3.740 grid_integrate_task_list 110 12.3 3.358 3.470 3.358 3.470 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.429 3.430 pw_transfer 1331 11.6 0.065 0.070 3.320 3.335 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.212 3.232 cp_fm_cholesky_invert 11 10.9 3.083 3.087 3.083 3.087 multiply_cannon_sync_h2d 8220 15.4 2.887 3.048 2.887 3.048 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.892 2.893 cp_fm_redistribute_end 48 14.0 0.738 2.864 0.743 2.864 cp_fm_diag_elpa_base 48 14.0 1.935 2.660 2.112 2.824 density_rs2pw 110 9.6 0.004 0.004 2.699 2.811 make_images_data 4110 15.4 0.038 0.043 2.413 2.810 fft_wrap_pw1pw2_140 451 13.1 0.211 0.213 2.730 2.754 wfi_extrapolate 11 7.9 0.001 0.001 2.732 2.732 hybrid_alltoall_any 4261 16.3 0.199 0.859 2.338 2.690 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.661 2.661 calculate_dm_sparse 110 9.5 0.001 0.001 2.486 2.529 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.522 2.524 fft3d_ps 1111 14.6 1.118 1.148 2.473 2.483 grid_collocate_task_list 110 9.6 2.327 2.472 2.327 2.472 rs_pw_transfer 902 11.9 0.010 0.011 2.215 2.354 calculate_first_density_matrix 1 7.0 0.001 0.001 2.251 2.252 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.144 2.160 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.776 1.994 potential_pw2rs 110 12.3 0.015 0.015 1.977 1.984 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.910 1.939 mp_alltoall_d11v 2046 13.8 1.717 1.878 1.717 1.878 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.833 1.850 cp_fm_cholesky_decompose 22 10.9 1.711 1.725 1.711 1.725 mp_allgather_i34 2055 14.4 0.555 1.648 0.555 1.648 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.501 1.624 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.595 1.601 dbcsr_complete_redistribute 325 12.2 0.560 0.579 1.498 1.588 jit_kernel_multiply 9 15.7 0.950 1.522 0.950 1.522 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.435 1.448 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.483 1.420 mp_irecv_dv 24056 15.7 0.459 1.377 0.459 1.377 qs_create_task_list 11 7.9 0.001 0.001 1.229 1.331 generate_qs_task_list 11 8.9 0.378 0.446 1.229 1.330 mp_waitany 9240 13.8 1.065 1.211 1.065 1.211 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.180 1.206 mp_alltoall_z22v 1111 16.6 1.122 1.141 1.122 1.141 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.181000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=777.090909, yerr=12.287869 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 1.355788E+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 0.021 0.041 86.502 86.502 qs_mol_dyn_low 1 2.0 0.003 0.003 86.190 86.199 qs_forces 11 3.9 0.002 0.002 85.856 85.857 qs_energies 11 4.9 0.002 0.002 81.812 81.814 scf_env_do_scf 11 5.9 0.001 0.001 71.810 71.810 velocity_verlet 10 3.0 0.002 0.002 55.348 55.387 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 43.381 43.382 dbcsr_multiply_generic 2055 12.4 0.120 0.122 28.835 28.913 init_scf_loop 11 6.9 0.000 0.000 28.354 28.355 qs_scf_new_mos 99 7.5 0.001 0.001 26.789 26.831 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.788 26.830 prepare_preconditioner 11 7.9 0.000 0.000 26.390 26.397 make_preconditioner 11 8.9 0.000 0.000 26.390 26.397 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.739 25.855 ot_scf_mini 99 9.5 0.002 0.002 25.001 25.034 multiply_cannon 2055 13.4 0.343 0.360 21.528 22.096 multiply_cannon_loop 2055 14.4 0.340 0.342 19.784 20.094 cp_fm_upper_to_full 70 14.2 12.533 17.884 12.533 17.884 ot_mini 99 10.5 0.001 0.001 13.762 13.791 rebuild_ks_matrix 110 8.3 0.001 0.001 13.311 13.350 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.310 13.349 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.104 12.140 dbcsr_complete_redistribute 325 12.2 1.021 1.038 7.294 10.405 multiply_cannon_multrec 8220 15.4 4.377 4.592 9.516 9.589 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.234 9.336 qs_ot_get_derivative 99 11.5 0.001 0.001 9.157 9.190 mp_waitall_1 84994 16.7 7.898 8.721 7.898 8.721 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.636 8.717 mp_alltoall_i22 605 13.7 5.249 8.334 5.249 8.334 sum_up_and_integrate 110 10.3 0.151 0.152 6.755 6.768 integrate_v_rspace 110 11.3 0.003 0.004 6.605 6.619 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.547 6.583 calculate_rho_elec 110 8.6 0.227 0.227 6.547 6.583 qs_ot_get_p 110 10.4 0.001 0.001 5.975 6.011 make_m2s 4110 13.4 0.045 0.046 5.484 5.953 make_images 4110 14.4 0.869 0.905 5.294 5.762 init_scf_run 11 5.9 0.000 0.001 5.678 5.678 scf_env_initial_rho_setup 11 6.9 0.002 0.003 5.678 5.678 cp_fm_cholesky_invert 11 10.9 5.558 5.563 5.558 5.563 dbcsr_mm_accdrv_process 11614 15.7 3.316 3.638 4.998 5.254 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.910 5.244 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.691 5.201 apply_single 110 13.6 0.000 0.000 4.691 5.201 ot_diis_step 99 11.5 0.015 0.016 4.576 4.576 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.324 4.331 multiply_cannon_sync_h2d 8220 15.4 3.951 3.955 3.951 3.955 pw_transfer 1331 11.6 0.074 0.074 3.876 3.883 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.872 3.872 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.759 3.767 qs_energies_init_hamiltonians 11 5.9 0.028 0.084 3.718 3.719 grid_integrate_task_list 110 12.3 3.657 3.708 3.657 3.708 hybrid_alltoall_any 4261 16.3 0.255 0.550 3.084 3.691 make_images_data 4110 15.4 0.042 0.044 3.056 3.670 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.927 3.381 wfi_extrapolate 11 7.9 0.001 0.001 3.279 3.279 fft_wrap_pw1pw2_140 451 13.1 0.214 0.217 3.237 3.246 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.234 3.235 cp_fm_diag_elpa_base 48 14.0 2.710 2.902 3.232 3.233 calculate_dm_sparse 110 9.5 0.001 0.001 3.152 3.177 density_rs2pw 110 9.6 0.004 0.004 3.037 3.057 fft3d_ps 1111 14.6 1.264 1.274 2.970 2.977 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.900 2.903 grid_collocate_task_list 110 9.6 2.626 2.648 2.626 2.648 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.440 2.457 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.371 2.385 calculate_first_density_matrix 1 7.0 0.032 0.035 2.284 2.285 rs_pw_transfer 902 11.9 0.010 0.011 2.218 2.270 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.174 2.240 mp_alltoall_d11v 2046 13.8 2.151 2.187 2.151 2.187 potential_pw2rs 110 12.3 0.021 0.021 2.171 2.176 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.087 2.153 cp_fm_cholesky_decompose 22 10.9 2.095 2.111 2.095 2.111 qs_create_task_list 11 7.9 0.001 0.001 1.901 1.944 generate_qs_task_list 11 8.9 0.737 0.791 1.900 1.944 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.821 1.864 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.810 1.814 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.502000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1216.272727, yerr=56.322272 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 627.617792E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955098256 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 2527665. 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 0.014 0.029 203.005 203.006 qs_mol_dyn_low 1 2.0 0.003 0.004 202.611 202.625 qs_forces 11 3.9 0.003 0.003 202.522 202.524 qs_energies 11 4.9 0.001 0.002 197.030 197.048 scf_env_do_scf 11 5.9 0.001 0.001 180.614 180.617 scf_env_do_scf_inner_loop 117 6.6 0.002 0.008 159.627 159.629 dbcsr_multiply_generic 2507 12.6 0.179 0.195 123.886 124.915 velocity_verlet 10 3.0 0.002 0.002 122.162 122.163 qs_scf_new_mos 117 7.6 0.001 0.001 120.721 120.898 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.720 120.898 ot_scf_mini 117 9.6 0.003 0.003 114.089 114.239 multiply_cannon 2507 13.6 0.236 0.244 100.527 103.578 multiply_cannon_loop 2507 14.6 2.123 2.187 98.248 101.186 ot_mini 117 10.6 0.001 0.001 65.495 65.695 multiply_cannon_multrec 60168 15.6 33.228 35.520 41.569 43.242 qs_ot_get_derivative 117 11.6 0.001 0.002 40.509 40.659 rebuild_ks_matrix 128 8.3 0.001 0.001 33.347 33.667 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.347 33.667 mp_waitall_1 267128 16.5 28.063 31.453 28.063 31.453 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.021 30.343 multiply_cannon_sync_h2d 60168 15.6 27.322 29.098 27.322 29.098 qs_ot_get_p 128 10.4 0.001 0.001 26.733 26.976 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.413 25.354 apply_single 128 13.6 0.001 0.001 24.413 25.353 ot_diis_step 117 11.6 0.007 0.008 24.645 24.647 init_scf_loop 11 6.9 0.000 0.001 20.904 20.906 qs_ot_p2m_diag 83 11.4 0.077 0.091 19.995 20.101 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.903 19.056 multiply_cannon_metrocomm3 60168 15.6 0.114 0.120 15.391 18.069 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.403 17.404 prepare_preconditioner 11 7.9 0.000 0.000 16.319 16.365 make_preconditioner 11 8.9 0.000 0.000 16.319 16.365 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.549 15.728 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.503 14.508 cp_fm_redistribute_end 83 14.4 11.467 14.428 11.479 14.431 make_m2s 5014 13.6 0.106 0.114 13.976 14.374 make_images 5014 14.6 0.402 0.419 13.792 14.199 cp_fm_diag_elpa_base 83 14.4 2.908 14.042 2.939 14.148 sum_up_and_integrate 128 10.3 0.089 0.107 14.086 14.101 integrate_v_rspace 128 11.3 0.003 0.004 13.997 14.014 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.779 12.865 calculate_rho_elec 128 8.7 0.045 0.063 12.778 12.864 init_scf_run 11 5.9 0.000 0.001 12.310 12.310 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.310 12.310 cp_fm_cholesky_invert 11 10.9 9.342 9.349 9.342 9.349 mp_sum_l 7870 13.0 8.121 9.225 8.121 9.225 wfi_extrapolate 11 7.9 0.001 0.001 9.062 9.062 calculate_dm_sparse 128 9.5 0.001 0.001 8.434 8.524 dbcsr_mm_accdrv_process 124484 16.2 3.299 3.441 7.904 8.432 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.856 7.996 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.809 7.919 multiply_cannon_metrocomm1 60168 15.6 0.087 0.092 5.918 7.798 make_images_data 5014 15.6 0.066 0.071 6.864 7.713 grid_integrate_task_list 128 12.3 7.064 7.473 7.064 7.473 hybrid_alltoall_any 5200 16.5 0.289 2.243 5.999 7.447 pw_transfer 1547 11.6 0.075 0.106 6.954 7.230 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.749 6.998 density_rs2pw 128 9.7 0.006 0.007 6.548 6.946 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.679 6.688 rs_pw_transfer 1046 11.9 0.016 0.019 5.742 6.210 fft_wrap_pw1pw2_140 523 13.2 0.445 0.507 5.753 5.928 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.793 5.850 fft3d_ps 1291 14.7 2.096 2.578 5.541 5.729 mp_alltoall_d11v 2415 14.1 4.318 5.639 4.318 5.639 grid_collocate_task_list 128 9.7 4.728 5.099 4.728 5.099 cp_fm_cholesky_decompose 22 10.9 4.782 4.796 4.782 4.796 potential_pw2rs 128 12.3 0.009 0.011 4.565 4.608 mp_sum_d 4455 12.2 3.594 4.446 3.594 4.446 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=203.006000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.454545, yerr=6.199173 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 827.482112E+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 4103 56957. MP_Allreduce 11297 945. MP_Sync 170 MP_Alltoall 1969 4856805. 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 0.029 0.062 189.729 189.731 qs_mol_dyn_low 1 2.0 0.004 0.006 189.268 189.281 qs_forces 11 3.9 0.003 0.004 189.178 189.180 qs_energies 11 4.9 0.002 0.009 182.466 182.477 scf_env_do_scf 11 5.9 0.001 0.002 166.233 166.243 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 133.055 133.058 velocity_verlet 10 3.0 0.001 0.003 119.503 119.506 dbcsr_multiply_generic 2507 12.6 0.186 0.192 97.311 98.567 qs_scf_new_mos 117 7.6 0.001 0.001 94.366 94.893 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.365 94.892 ot_scf_mini 117 9.6 0.004 0.005 89.561 90.158 multiply_cannon 2507 13.6 0.475 0.534 77.312 82.012 multiply_cannon_loop 2507 14.6 1.257 1.290 74.063 76.530 ot_mini 117 10.6 0.001 0.001 49.716 50.288 mp_waitall_1 214728 16.6 24.427 38.084 24.427 38.084 multiply_cannon_multrec 30084 15.6 22.061 26.736 31.656 36.839 rebuild_ks_matrix 128 8.3 0.001 0.001 32.779 33.254 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 32.778 33.254 init_scf_loop 11 6.9 0.000 0.002 33.087 33.088 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.552 30.004 multiply_cannon_metrocomm3 30084 15.6 0.094 0.099 15.589 29.185 prepare_preconditioner 11 7.9 0.000 0.000 28.717 28.783 make_preconditioner 11 8.9 0.000 0.000 28.717 28.782 qs_ot_get_derivative 117 11.6 0.001 0.002 27.829 28.417 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.406 27.947 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.949 23.070 apply_single 128 13.6 0.001 0.001 21.949 23.070 qs_ot_get_p 128 10.4 0.001 0.001 21.582 22.297 ot_diis_step 117 11.6 0.014 0.015 21.710 21.712 multiply_cannon_sync_h2d 30084 15.6 19.370 21.650 19.370 21.650 qs_ot_p2m_diag 83 11.4 0.189 0.218 16.779 16.819 cp_fm_cholesky_invert 11 10.9 16.771 16.783 16.771 16.783 make_m2s 5014 13.6 0.092 0.098 14.271 15.711 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.629 15.630 make_images 5014 14.6 1.148 1.332 14.058 15.498 sum_up_and_integrate 128 10.3 0.116 0.133 14.378 14.408 integrate_v_rspace 128 11.3 0.003 0.003 14.262 14.298 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.995 13.025 calculate_rho_elec 128 8.7 0.088 0.105 12.995 13.024 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.533 12.541 cp_fm_redistribute_end 83 14.4 7.321 12.474 7.333 12.476 cp_fm_diag_elpa_base 83 14.4 4.909 11.969 5.127 12.345 init_scf_run 11 5.9 0.000 0.001 11.446 11.448 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.446 11.448 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.955 11.397 make_images_data 5014 15.6 0.064 0.071 8.679 10.582 multiply_cannon_metrocomm4 27577 15.6 0.094 0.107 3.730 10.188 mp_irecv_dv 69486 16.3 3.539 9.816 3.539 9.816 hybrid_alltoall_any 5200 16.5 0.341 1.521 7.346 9.749 dbcsr_mm_accdrv_process 62242 16.2 4.576 5.275 9.054 9.651 wfi_extrapolate 11 7.9 0.001 0.001 8.344 8.344 pw_transfer 1547 11.6 0.085 0.102 7.788 7.854 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.564 7.625 grid_integrate_task_list 128 12.3 7.170 7.516 7.170 7.516 cp_fm_cholesky_decompose 22 10.9 7.066 7.147 7.066 7.147 density_rs2pw 128 9.7 0.006 0.006 6.760 7.124 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.186 6.916 fft_wrap_pw1pw2_140 523 13.2 0.469 0.519 6.591 6.650 calculate_dm_sparse 128 9.5 0.001 0.002 6.459 6.603 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.134 6.143 rs_pw_transfer 1046 11.9 0.014 0.016 5.631 6.023 fft3d_ps 1291 14.7 2.775 2.926 5.938 5.972 mp_sum_l 7870 13.0 3.973 5.942 3.973 5.942 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.283 5.438 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.309 5.385 grid_collocate_task_list 128 9.7 4.914 5.221 4.914 5.221 mp_allgather_i34 2507 14.6 1.828 5.015 1.828 5.015 potential_pw2rs 128 12.3 0.015 0.018 4.830 4.856 mp_alltoall_d11v 2415 14.1 4.151 4.563 4.151 4.563 mp_sum_d 4469 12.2 2.647 4.000 2.647 4.000 dbcsr_complete_redistribute 395 12.7 0.773 0.858 3.121 3.976 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.731000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.636364, yerr=3.391774 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 952.565760E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.033 180.546 180.546 qs_mol_dyn_low 1 2.0 0.003 0.004 180.162 180.176 qs_forces 11 3.9 0.003 0.003 180.069 180.072 qs_energies 11 4.9 0.001 0.002 173.550 173.560 scf_env_do_scf 11 5.9 0.001 0.002 158.042 158.043 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 121.188 121.189 velocity_verlet 10 3.0 0.002 0.002 115.949 115.951 qs_scf_new_mos 117 7.6 0.001 0.001 84.006 84.349 qs_scf_loop_do_ot 117 8.6 0.001 0.001 84.006 84.348 dbcsr_multiply_generic 2507 12.6 0.183 0.190 82.637 83.646 ot_scf_mini 117 9.6 0.003 0.004 79.816 80.184 multiply_cannon 2507 13.6 0.498 0.542 61.745 66.463 multiply_cannon_loop 2507 14.6 0.856 0.886 58.437 60.993 ot_mini 117 10.6 0.001 0.001 43.599 43.976 init_scf_loop 11 6.9 0.000 0.000 36.746 36.747 mp_waitall_1 170520 16.6 25.775 35.560 25.775 35.560 prepare_preconditioner 11 7.9 0.000 0.000 32.714 32.761 make_preconditioner 11 8.9 0.000 0.000 32.714 32.761 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.295 31.747 rebuild_ks_matrix 128 8.3 0.001 0.001 30.389 30.855 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 30.389 30.855 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.464 27.887 multiply_cannon_multrec 20056 15.6 13.476 16.966 22.086 25.485 multiply_cannon_metrocomm3 20056 15.6 0.060 0.065 15.264 25.269 qs_ot_get_derivative 117 11.6 0.001 0.002 23.768 24.144 qs_ot_get_p 128 10.4 0.001 0.001 21.445 21.964 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.853 20.882 apply_single 128 13.6 0.001 0.001 19.853 20.882 ot_diis_step 117 11.6 0.018 0.018 19.711 19.712 qs_ot_p2m_diag 83 11.4 0.266 0.274 16.983 17.001 make_m2s 5014 13.6 0.083 0.088 15.452 16.294 make_images 5014 14.6 1.181 1.275 15.216 16.057 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.912 15.914 multiply_cannon_sync_h2d 20056 15.6 14.073 15.802 14.073 15.802 cp_fm_cholesky_invert 11 10.9 15.438 15.450 15.438 15.450 sum_up_and_integrate 128 10.3 0.132 0.144 14.463 14.487 integrate_v_rspace 128 11.3 0.003 0.004 14.331 14.359 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.401 13.431 calculate_rho_elec 128 8.7 0.132 0.146 13.401 13.430 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.612 12.614 cp_fm_redistribute_end 83 14.4 4.758 12.506 4.776 12.509 cp_fm_diag_elpa_base 83 14.4 7.290 11.869 7.705 12.353 make_images_data 5014 15.6 0.061 0.069 9.551 10.856 init_scf_run 11 5.9 0.000 0.001 10.499 10.499 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.499 10.499 hybrid_alltoall_any 5200 16.5 0.433 1.992 8.245 9.857 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.475 9.752 multiply_cannon_metrocomm4 17549 15.6 0.061 0.071 3.440 8.875 mp_irecv_dv 50230 16.2 3.319 8.631 3.319 8.631 dbcsr_mm_accdrv_process 41502 16.2 4.379 4.868 8.081 8.211 pw_transfer 1547 11.6 0.085 0.104 8.046 8.151 cp_fm_cholesky_decompose 22 10.9 7.981 8.068 7.981 8.068 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.822 7.934 grid_integrate_task_list 128 12.3 7.311 7.860 7.311 7.860 cp_fm_upper_to_full 105 14.5 5.826 7.533 5.826 7.533 wfi_extrapolate 11 7.9 0.001 0.001 7.518 7.518 density_rs2pw 128 9.7 0.006 0.006 6.848 7.281 fft_wrap_pw1pw2_140 523 13.2 0.476 0.534 6.725 6.841 dbcsr_complete_redistribute 395 12.7 1.167 1.197 4.704 6.477 fft3d_ps 1291 14.7 2.692 2.893 6.093 6.155 calculate_dm_sparse 128 9.5 0.001 0.001 5.793 5.889 rs_pw_transfer 1046 11.9 0.014 0.015 5.473 5.872 grid_collocate_task_list 128 9.7 5.077 5.521 5.077 5.521 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.483 5.491 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.706 5.399 mp_alltoall_d11v 2415 14.1 4.415 5.366 4.415 5.366 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.427 5.192 mp_allgather_i34 2507 14.6 1.828 5.148 1.828 5.148 mp_sum_l 7870 13.0 3.622 5.034 3.622 5.034 potential_pw2rs 128 12.3 0.020 0.022 4.796 4.824 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.538 4.656 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.184 4.226 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.400 4.137 mp_alltoall_i22 716 14.1 2.004 3.919 2.004 3.919 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.734 3.735 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=180.546000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=895.454545, yerr=13.391745 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/18/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 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243750E+12 0.0% 0.0% 100.0% flops max/rank 4.387745E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806546176 0.0% 0.0% 100.0% number of processed stacks 6032640 0.0% 0.0% 100.0% average stack size 0.0 0.0 1128.3 marketing flops 145.662591E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 1.136730E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1155504 MPI messages size (bytes): total size 2.039825E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.765312E+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 323336 36533174272 131072 < size <= 4194304 723016 792311627776 4194304 < size <= 16777216 70800 669922098320 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4127 56626. MP_Allreduce 11356 1063. MP_Sync 172 MP_Alltoall 1724 12509574. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_comm_split 84 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.094 187.219 187.221 qs_mol_dyn_low 1 2.0 0.003 0.004 186.756 186.769 qs_forces 11 3.9 0.003 0.005 186.644 186.654 qs_energies 11 4.9 0.002 0.006 179.630 179.639 scf_env_do_scf 11 5.9 0.001 0.002 162.596 162.609 velocity_verlet 10 3.0 0.003 0.003 124.457 124.460 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 116.162 116.163 dbcsr_multiply_generic 2534 12.6 0.192 0.200 80.414 81.150 qs_scf_new_mos 118 7.6 0.001 0.001 80.377 80.715 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.376 80.715 ot_scf_mini 118 9.6 0.004 0.004 75.840 76.123 multiply_cannon 2534 13.6 0.562 0.603 55.655 58.083 multiply_cannon_loop 2534 14.6 1.196 1.219 51.721 53.586 init_scf_loop 11 6.9 0.001 0.002 46.302 46.302 ot_mini 118 10.6 0.001 0.001 42.734 43.014 prepare_preconditioner 11 7.9 0.000 0.000 42.257 42.281 make_preconditioner 11 8.9 0.000 0.002 42.257 42.281 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.942 40.877 multiply_cannon_multrec 30408 15.6 14.018 19.333 26.067 30.937 rebuild_ks_matrix 129 8.3 0.001 0.001 29.151 29.436 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 29.150 29.436 mp_waitall_1 149452 16.7 17.745 27.618 17.745 27.618 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.371 26.625 qs_ot_get_derivative 118 11.6 0.001 0.002 22.828 23.113 make_m2s 5068 13.6 0.098 0.103 20.350 21.530 make_images 5068 14.6 1.964 2.284 20.037 21.219 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.319 19.914 apply_single 129 13.6 0.001 0.001 19.319 19.913 ot_diis_step 118 11.6 0.018 0.018 19.786 19.787 qs_ot_get_p 129 10.4 0.001 0.001 19.201 19.520 cp_fm_cholesky_invert 11 10.9 16.365 16.374 16.365 16.374 cp_fm_upper_to_full 106 14.7 10.904 16.109 10.904 16.109 multiply_cannon_metrocomm3 30408 15.6 0.047 0.049 6.848 15.463 qs_ot_p2m_diag 84 11.4 0.347 0.394 15.073 15.127 sum_up_and_integrate 129 10.3 0.141 0.154 14.140 14.170 integrate_v_rspace 129 11.3 0.003 0.003 13.999 14.034 cp_dbcsr_syevd 84 12.4 0.005 0.005 13.763 13.765 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.228 13.297 calculate_rho_elec 129 8.7 0.176 0.192 13.228 13.296 multiply_cannon_sync_h2d 30408 15.6 11.788 12.871 11.788 12.871 make_images_data 5068 15.6 0.065 0.068 10.928 12.809 dbcsr_complete_redistribute 397 12.7 1.511 1.633 8.947 12.639 dbcsr_mm_accdrv_process 62840 16.2 7.497 8.596 11.616 12.188 hybrid_alltoall_any 5255 16.5 0.531 2.232 9.877 11.682 copy_fm_to_dbcsr 210 11.7 0.002 0.002 7.540 11.238 init_scf_run 11 5.9 0.000 0.001 11.092 11.093 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.092 11.093 cp_fm_diag_elpa 84 13.4 0.000 0.001 10.618 10.620 cp_fm_redistribute_end 84 14.4 1.839 10.549 1.853 10.553 cp_fm_diag_elpa_base 84 14.4 8.088 9.928 8.668 10.420 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.292 9.869 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.510 9.724 mp_alltoall_i22 720 14.1 5.512 9.133 5.512 9.133 pw_transfer 1559 11.6 0.085 0.099 7.912 7.998 grid_integrate_task_list 129 12.3 7.623 7.905 7.623 7.905 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.685 7.776 cp_fm_cholesky_decompose 22 10.9 7.566 7.654 7.566 7.654 wfi_extrapolate 11 7.9 0.001 0.001 7.613 7.613 multiply_cannon_metrocomm4 25340 15.6 0.076 0.086 2.858 7.398 mp_irecv_dv 76866 16.2 2.712 7.121 2.712 7.121 fft_wrap_pw1pw2_140 527 13.2 0.484 0.496 6.794 6.889 calculate_dm_sparse 129 9.5 0.001 0.002 6.607 6.682 density_rs2pw 129 9.7 0.006 0.006 6.306 6.659 fft3d_ps 1301 14.7 2.821 2.901 5.902 5.970 grid_collocate_task_list 129 9.7 5.247 5.553 5.247 5.553 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.467 5.518 mp_alltoall_d11v 2429 14.1 4.835 5.163 4.835 5.163 rs_pw_transfer 1054 12.0 0.013 0.015 4.660 4.951 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.471 4.473 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.376 4.459 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.364 4.449 potential_pw2rs 129 12.3 0.023 0.023 4.285 4.300 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.224 4.271 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.221000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1067.636364, yerr=21.141397 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/19/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 5.865089E+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 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.481069E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58209. MP_Allreduce 11002 1175. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.038 175.474 175.475 qs_mol_dyn_low 1 2.0 0.003 0.005 175.057 175.070 qs_forces 11 3.9 0.003 0.004 174.947 174.951 qs_energies 11 4.9 0.009 0.009 167.634 167.641 scf_env_do_scf 11 5.9 0.001 0.002 150.573 150.586 velocity_verlet 10 3.0 0.002 0.002 114.758 114.762 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 113.612 113.613 qs_scf_new_mos 117 7.6 0.001 0.001 77.695 77.828 qs_scf_loop_do_ot 117 8.6 0.001 0.001 77.694 77.827 dbcsr_multiply_generic 2507 12.6 0.183 0.190 75.121 75.710 ot_scf_mini 117 9.6 0.003 0.004 73.169 73.257 multiply_cannon 2507 13.6 0.580 0.620 54.967 59.865 multiply_cannon_loop 2507 14.6 0.446 0.455 50.206 51.311 ot_mini 117 10.6 0.001 0.001 40.444 40.526 init_scf_loop 11 6.9 0.000 0.000 36.797 36.798 mp_waitall_1 125778 16.7 26.539 34.026 26.539 34.026 prepare_preconditioner 11 7.9 0.000 0.000 32.918 32.954 make_preconditioner 11 8.9 0.000 0.000 32.918 32.954 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.804 31.093 rebuild_ks_matrix 128 8.3 0.001 0.001 28.621 28.747 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 28.620 28.746 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.109 26.225 multiply_cannon_multrec 10028 15.6 10.482 14.139 17.728 20.484 qs_ot_get_derivative 117 11.6 0.001 0.002 20.313 20.380 ot_diis_step 117 11.6 0.020 0.020 20.056 20.057 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.592 20.002 apply_single 128 13.6 0.001 0.001 19.592 20.002 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 12.700 19.867 qs_ot_get_p 128 10.4 0.001 0.001 19.644 19.764 cp_fm_cholesky_invert 11 10.9 18.997 19.003 18.997 19.003 make_m2s 5014 13.6 0.067 0.074 16.363 18.556 make_images 5014 14.6 2.305 2.668 16.056 18.243 qs_ot_p2m_diag 83 11.4 0.496 0.501 15.676 15.694 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.469 14.471 sum_up_and_integrate 128 10.3 0.181 0.191 14.164 14.212 integrate_v_rspace 128 11.3 0.003 0.004 13.982 14.037 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.438 13.475 calculate_rho_elec 128 8.7 0.258 0.269 13.437 13.475 make_images_data 5014 15.6 0.053 0.060 9.932 12.405 hybrid_alltoall_any 5200 16.5 0.816 3.631 9.651 11.875 multiply_cannon_sync_h2d 10028 15.6 11.605 11.855 11.605 11.855 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.302 11.314 cp_fm_diag_elpa_base 83 14.4 11.045 11.131 11.294 11.308 init_scf_run 11 5.9 0.000 0.001 10.246 10.247 scf_env_initial_rho_setup 11 6.9 0.003 0.015 10.246 10.247 cp_fm_cholesky_decompose 22 10.9 8.265 8.372 8.265 8.372 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.224 8.274 grid_integrate_task_list 128 12.3 7.768 8.011 7.768 8.011 pw_transfer 1547 11.6 0.084 0.091 7.938 7.954 mp_allgather_i34 2507 14.6 2.976 7.932 2.976 7.932 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.717 7.741 dbcsr_mm_accdrv_process 20762 16.1 2.613 3.432 6.879 7.491 wfi_extrapolate 11 7.9 0.001 0.001 7.349 7.349 multiply_cannon_metrocomm1 10028 15.6 0.029 0.029 4.563 7.336 fft_wrap_pw1pw2_140 523 13.2 0.501 0.525 6.714 6.733 calculate_dm_sparse 128 9.5 0.001 0.001 6.186 6.278 density_rs2pw 128 9.7 0.005 0.006 6.141 6.267 fft3d_ps 1291 14.7 2.725 2.795 5.881 5.910 grid_collocate_task_list 128 9.7 5.489 5.857 5.489 5.857 dbcsr_complete_redistribute 395 12.7 2.121 2.181 5.280 5.690 qs_energies_init_hamiltonians 11 5.9 0.011 0.014 5.238 5.242 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.138 5.151 mp_alltoall_d11v 2415 14.1 4.833 5.104 4.833 5.104 rs_pw_transfer 1046 11.9 0.012 0.013 4.284 4.397 potential_pw2rs 128 12.3 0.026 0.027 4.211 4.222 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.049 4.101 multiply_cannon_metrocomm4 7521 15.6 0.024 0.026 1.859 3.963 mp_irecv_dv 28860 15.9 1.824 3.890 1.824 3.890 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.830 3.854 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.557 3.851 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.436 3.769 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.597 3.677 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.521 3.540 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=175.475000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1395.818182, yerr=38.173809 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.062473E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3943 59996. MP_Allreduce 10855 1518. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.049 285.701 285.702 qs_mol_dyn_low 1 2.0 0.003 0.003 285.143 285.157 qs_forces 11 3.9 0.003 0.003 285.044 285.047 qs_energies 11 4.9 0.001 0.001 276.488 276.499 scf_env_do_scf 11 5.9 0.001 0.005 254.797 254.810 velocity_verlet 10 3.0 0.002 0.002 206.863 206.872 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 130.195 130.197 init_scf_loop 11 6.9 0.000 0.000 124.338 124.341 prepare_preconditioner 11 7.9 0.000 0.000 119.655 119.682 make_preconditioner 11 8.9 0.000 0.000 119.655 119.682 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.092 116.826 qs_scf_new_mos 116 7.6 0.001 0.001 88.838 88.955 qs_scf_loop_do_ot 116 8.6 0.001 0.001 88.837 88.955 ot_scf_mini 116 9.6 0.004 0.004 84.122 84.160 dbcsr_multiply_generic 2485 12.5 0.211 0.218 81.966 82.523 cp_fm_upper_to_full 104 14.8 52.827 76.443 52.827 76.443 multiply_cannon 2485 13.5 0.704 0.759 57.835 58.570 multiply_cannon_loop 2485 14.5 0.468 0.481 54.211 55.518 ot_mini 116 10.6 0.001 0.001 43.752 43.791 dbcsr_complete_redistribute 393 12.7 3.981 4.019 30.239 43.574 copy_fm_to_dbcsr 208 11.6 0.002 0.002 26.852 40.201 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.518 37.759 mp_alltoall_i22 712 14.1 22.345 35.622 22.345 35.622 cp_fm_cholesky_invert 11 10.9 33.162 33.168 33.162 33.168 rebuild_ks_matrix 127 8.3 0.001 0.001 33.068 33.108 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 33.068 33.108 mp_waitall_1 102768 16.8 27.660 31.589 27.660 31.589 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.817 30.860 qs_ot_get_p 127 10.4 0.001 0.001 25.145 25.158 qs_ot_get_derivative 116 11.6 0.001 0.002 23.854 23.893 qs_ot_p2m_diag 82 11.4 0.868 0.872 21.181 21.209 make_m2s 4970 13.5 0.078 0.081 19.645 20.629 make_images 4970 14.5 3.716 3.888 19.168 20.155 ot_diis_step 116 11.6 0.021 0.022 19.870 19.870 cp_dbcsr_syevd 82 12.4 0.005 0.006 19.483 19.486 multiply_cannon_metrocomm3 9940 15.5 0.024 0.024 18.407 19.458 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.122 19.315 apply_single 127 13.6 0.001 0.001 19.121 19.315 multiply_cannon_multrec 9940 15.5 10.439 12.143 17.759 17.857 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.343 16.345 cp_fm_diag_elpa_base 82 14.4 11.982 13.617 16.339 16.339 multiply_cannon_sync_h2d 9940 15.5 15.540 15.578 15.540 15.578 sum_up_and_integrate 127 10.3 0.319 0.322 15.454 15.551 integrate_v_rspace 127 11.3 0.004 0.004 15.135 15.232 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.768 14.787 calculate_rho_elec 127 8.7 0.479 0.479 14.767 14.786 hybrid_alltoall_any 5155 16.4 1.292 3.018 10.716 12.666 make_images_data 4970 15.5 0.060 0.064 10.510 12.293 init_scf_run 11 5.9 0.000 0.001 11.721 11.722 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.721 11.721 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.251 9.291 cp_fm_cholesky_decompose 22 10.9 8.845 8.876 8.845 8.876 dbcsr_mm_accdrv_process 20590 16.0 3.840 5.624 7.085 8.850 wfi_extrapolate 11 7.9 0.001 0.001 8.698 8.698 grid_integrate_task_list 127 12.3 8.478 8.651 8.478 8.651 pw_transfer 1535 11.6 0.091 0.092 8.259 8.266 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 8.027 8.030 qs_energies_init_hamiltonians 11 5.9 0.003 0.005 7.973 7.973 fft_wrap_pw1pw2_140 519 13.2 0.535 0.539 7.117 7.138 mp_alltoall_d11v 2401 14.1 6.778 6.913 6.778 6.913 calculate_dm_sparse 127 9.5 0.001 0.001 6.558 6.654 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.317 6.371 grid_collocate_task_list 127 9.7 6.288 6.326 6.288 6.326 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.060 6.124 fft3d_ps 1281 14.7 2.738 2.778 6.089 6.103 density_rs2pw 127 9.7 0.005 0.005 5.919 5.952 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=285.702000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2729.181818, yerr=170.699000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 1.261814E+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 23161601. 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 0.023 0.038 84.745 84.747 qs_energies 1 2.0 0.000 0.000 84.288 84.294 ls_scf 1 3.0 0.000 0.001 83.382 83.388 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.544 72.689 multiply_cannon 111 7.7 0.017 0.020 55.851 56.949 multiply_cannon_loop 111 8.7 0.211 0.224 52.468 53.829 ls_scf_main 1 4.0 0.000 0.000 52.166 52.166 density_matrix_trs4 2 5.0 0.002 0.003 46.684 46.796 ls_scf_init_scf 1 4.0 0.000 0.001 28.170 28.171 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.087 27.142 mp_waitall_1 11031 10.9 22.428 25.211 22.428 25.211 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.989 25.003 multiply_cannon_multrec 2664 9.7 8.160 9.036 15.467 17.389 multiply_cannon_sync_h2d 2664 9.7 13.639 15.263 13.639 15.263 make_m2s 222 7.7 0.009 0.011 12.986 13.430 make_images 222 8.7 0.099 0.109 12.964 13.409 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.732 12.838 make_images_data 222 9.7 0.004 0.005 7.564 8.122 dbcsr_mm_accdrv_process 4760 10.4 0.507 0.616 6.925 7.896 hybrid_alltoall_any 227 10.6 0.215 1.828 6.552 7.722 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.461 7.676 dbcsr_mm_accdrv_process_sort 4760 11.4 6.215 7.115 6.215 7.115 calculate_norms 4752 9.8 5.510 6.116 5.510 6.116 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.046 5.266 mp_sum_l 807 5.4 3.178 4.778 3.178 4.778 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.061 3.787 mp_irecv_dv 6231 10.9 2.044 3.757 2.044 3.757 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.369 3.623 make_images_sizes 222 9.7 0.000 0.000 0.695 3.580 mp_alltoall_i44 222 10.7 0.695 3.579 0.695 3.579 arnoldi_extremal 4 6.8 0.000 0.000 3.207 3.234 arnoldi_normal_ev 4 7.8 0.001 0.004 3.207 3.234 build_subspace 16 8.4 0.009 0.012 3.108 3.111 ls_scf_post 1 4.0 0.000 0.001 3.046 3.052 ls_scf_store_result 1 5.0 0.000 0.000 2.861 2.903 dbcsr_special_finalize 555 9.7 0.005 0.006 2.376 2.785 dbcsr_merge_single_wm 555 10.7 0.450 0.590 2.368 2.777 make_images_pack 222 9.7 2.205 2.626 2.207 2.628 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.319 2.557 dbcsr_sort_data 658 11.4 2.176 2.520 2.176 2.520 dbcsr_matrix_vector_mult_local 304 10.0 2.070 2.458 2.072 2.460 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.221 2.345 buffer_matrices_ensure_size 222 8.7 1.753 2.089 1.753 2.089 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.755 1.759 rebuild_ks_matrix 3 7.3 0.000 0.000 1.746 1.749 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.746 1.749 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.747000, 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/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 2.081579E+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 0.034 0.051 90.461 90.463 qs_energies 1 2.0 0.000 0.001 89.953 89.957 ls_scf 1 3.0 0.000 0.002 88.643 88.647 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.810 75.145 multiply_cannon 111 7.7 0.028 0.040 53.115 56.558 ls_scf_main 1 4.0 0.000 0.004 54.560 54.564 multiply_cannon_loop 111 8.7 0.117 0.123 49.854 52.872 density_matrix_trs4 2 5.0 0.002 0.003 48.904 49.103 ls_scf_init_scf 1 4.0 0.000 0.000 30.561 30.562 mp_waitall_1 9105 10.9 20.975 30.358 20.975 30.358 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.431 29.514 multiply_cannon_multrec 1332 9.7 13.227 17.156 22.443 27.556 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.037 27.051 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.615 21.037 make_m2s 222 7.7 0.006 0.007 15.284 15.910 make_images 222 8.7 1.572 1.962 15.253 15.881 dbcsr_mm_accdrv_process 4041 10.4 0.306 0.450 8.809 10.334 dbcsr_mm_accdrv_process_sort 4041 11.4 8.400 9.885 8.400 9.885 make_images_data 222 9.7 0.004 0.004 8.788 9.661 hybrid_alltoall_any 227 10.6 0.522 2.487 8.269 9.186 mp_sum_l 807 5.4 5.105 8.155 5.105 8.155 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.204 7.735 mp_irecv_dv 3311 11.0 3.185 7.690 3.185 7.690 calculate_norms 2376 9.8 6.003 6.838 6.003 6.838 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.967 6.501 multiply_cannon_sync_h2d 1332 9.7 4.811 5.854 4.811 5.854 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.007 5.268 arnoldi_extremal 4 6.8 0.000 0.000 4.639 4.660 arnoldi_normal_ev 4 7.8 0.001 0.004 4.639 4.660 build_subspace 16 8.4 0.014 0.021 4.382 4.386 ls_scf_post 1 4.0 0.000 0.001 3.522 3.526 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.131 3.388 ls_scf_store_result 1 5.0 0.000 0.000 3.232 3.341 dbcsr_matrix_vector_mult_local 304 10.0 2.736 3.228 2.738 3.230 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.204 2.764 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.540 2.665 make_images_pack 222 9.7 2.024 2.421 2.026 2.423 mp_allgather_i34 111 8.7 0.959 2.284 0.959 2.284 dbcsr_sort_data 436 11.2 1.838 2.037 1.838 2.037 dbcsr_data_new 4174 10.1 1.616 1.853 1.616 1.853 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.812 1.814 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.463000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1703.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 2.712359E+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 0.033 0.060 93.860 93.861 qs_energies 1 2.0 0.000 0.000 93.255 93.258 ls_scf 1 3.0 0.000 0.001 91.845 91.849 dbcsr_multiply_generic 111 6.7 0.016 0.017 76.652 76.942 ls_scf_main 1 4.0 0.000 0.000 57.605 57.611 multiply_cannon 111 7.7 0.037 0.061 52.701 56.501 multiply_cannon_loop 111 8.7 0.100 0.108 49.158 52.925 density_matrix_trs4 2 5.0 0.002 0.003 51.657 51.836 mp_waitall_1 7281 11.0 23.953 33.886 23.953 33.886 ls_scf_init_scf 1 4.0 0.000 0.001 30.708 30.711 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.577 29.646 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.137 27.150 multiply_cannon_multrec 888 9.7 12.485 15.247 21.010 24.374 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.166 22.908 make_m2s 222 7.7 0.006 0.007 17.261 18.499 make_images 222 8.7 1.968 2.315 17.223 18.459 hybrid_alltoall_any 227 10.6 0.621 2.881 9.502 11.030 make_images_data 222 9.7 0.003 0.004 9.879 10.841 dbcsr_mm_accdrv_process 3754 10.4 0.245 0.415 8.042 9.254 dbcsr_mm_accdrv_process_sort 3754 11.4 7.673 8.839 7.673 8.839 mp_sum_l 807 5.4 5.121 8.702 5.121 8.702 multiply_cannon_sync_h2d 888 9.7 6.069 7.400 6.069 7.400 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.496 7.058 mp_irecv_dv 2335 11.1 2.482 7.011 2.482 7.011 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.995 6.975 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.529 6.776 arnoldi_extremal 4 6.8 0.000 0.000 5.134 5.148 arnoldi_normal_ev 4 7.8 0.001 0.005 5.134 5.148 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.948 5.144 build_subspace 16 8.4 0.014 0.020 4.831 4.839 calculate_norms 1584 9.8 4.272 4.615 4.272 4.615 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.459 3.804 mp_allgather_i34 111 8.7 1.386 3.801 1.386 3.801 dbcsr_matrix_vector_mult_local 304 10.0 3.014 3.582 3.016 3.584 ls_scf_post 1 4.0 0.000 0.000 3.532 3.536 ls_scf_store_result 1 5.0 0.000 0.000 3.286 3.356 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.781 2.901 make_images_sizes 222 9.7 0.000 0.000 1.046 2.191 mp_alltoall_i44 222 10.7 1.046 2.191 1.046 2.191 dbcsr_sort_data 325 11.1 1.914 2.144 1.914 2.144 make_images_pack 222 9.7 1.815 2.138 1.818 2.141 dbcsr_data_release 9322 10.9 1.313 1.886 1.313 1.886 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.861000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2183.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 3.343684E+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 0.048 97.324 97.325 qs_energies 1 2.0 0.000 0.000 96.740 96.745 ls_scf 1 3.0 0.000 0.000 95.111 95.118 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.980 79.243 ls_scf_main 1 4.0 0.000 0.000 59.166 59.167 multiply_cannon 111 7.7 0.060 0.153 51.696 56.391 density_matrix_trs4 2 5.0 0.002 0.003 53.086 53.208 multiply_cannon_loop 111 8.7 0.115 0.126 46.617 50.000 ls_scf_init_scf 1 4.0 0.000 0.000 32.714 32.716 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.579 31.643 mp_waitall_1 6369 11.0 23.000 30.034 23.000 30.034 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.109 29.122 multiply_cannon_multrec 1332 9.7 14.173 17.180 22.006 24.769 make_m2s 222 7.7 0.007 0.008 21.318 22.689 make_images 222 8.7 3.130 3.599 21.268 22.640 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.302 17.808 make_images_data 222 9.7 0.004 0.004 11.934 13.713 hybrid_alltoall_any 227 10.6 0.797 3.782 11.236 13.054 dbcsr_mm_accdrv_process 3641 10.4 0.227 0.408 7.479 8.992 dbcsr_mm_accdrv_process_sort 3641 11.4 7.107 8.577 7.107 8.577 mp_sum_l 807 5.4 4.380 7.754 4.380 7.754 multiply_cannon_sync_h2d 1332 9.7 5.507 6.257 5.507 6.257 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.439 6.120 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.087 6.049 mp_irecv_dv 3229 10.9 2.063 5.999 2.063 5.999 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.588 5.688 arnoldi_extremal 4 6.8 0.000 0.000 5.261 5.275 arnoldi_normal_ev 4 7.8 0.001 0.004 5.261 5.275 build_subspace 16 8.4 0.014 0.021 4.929 4.937 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.564 4.715 mp_allgather_i34 111 8.7 2.201 4.560 2.201 4.560 calculate_norms 2376 9.8 4.188 4.532 4.188 4.532 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.605 3.871 dbcsr_matrix_vector_mult_local 304 10.0 3.184 3.676 3.186 3.677 dbcsr_sort_data 658 11.4 3.048 3.334 3.048 3.334 ls_scf_post 1 4.0 0.000 0.000 3.231 3.237 dbcsr_special_finalize 555 9.7 0.006 0.007 2.786 3.221 dbcsr_merge_single_wm 555 10.7 0.532 0.671 2.778 3.213 ls_scf_store_result 1 5.0 0.000 0.000 2.967 3.036 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.928 2.967 dbcsr_data_release 10477 10.7 1.579 2.444 1.579 2.444 dbcsr_finalize 304 7.8 0.049 0.061 1.801 1.979 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.325000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2691.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 4.622770E+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 0.041 0.070 99.400 99.400 qs_energies 1 2.0 0.000 0.000 98.601 98.605 ls_scf 1 3.0 0.000 0.000 96.672 96.677 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.348 78.557 ls_scf_main 1 4.0 0.000 0.000 62.239 62.240 multiply_cannon 111 7.7 0.088 0.155 55.355 60.836 density_matrix_trs4 2 5.0 0.002 0.003 55.263 55.348 multiply_cannon_loop 111 8.7 0.070 0.075 50.939 52.614 mp_waitall_1 5436 11.0 26.953 32.806 26.953 32.806 ls_scf_init_scf 1 4.0 0.000 0.000 30.778 30.783 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.582 29.634 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.402 27.417 multiply_cannon_multrec 444 9.7 13.972 16.761 20.990 23.306 make_m2s 222 7.7 0.005 0.005 18.052 20.452 make_images 222 8.7 3.715 4.412 17.990 20.392 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.214 16.137 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.298 15.199 make_images_data 222 9.7 0.003 0.004 10.243 12.627 hybrid_alltoall_any 227 10.6 0.791 3.789 9.976 12.482 dbcsr_mm_accdrv_process 3003 10.4 0.187 0.342 6.723 7.916 multiply_cannon_sync_h2d 444 9.7 6.518 7.629 6.518 7.629 dbcsr_mm_accdrv_process_sort 3003 11.4 6.409 7.567 6.409 7.567 mp_allgather_i34 111 8.7 2.655 7.073 2.655 7.073 arnoldi_extremal 4 6.8 0.000 0.000 5.865 5.882 arnoldi_normal_ev 4 7.8 0.002 0.005 5.865 5.882 build_subspace 16 8.4 0.015 0.020 5.464 5.479 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.740 4.870 mp_irecv_dv 1241 11.2 1.719 4.839 1.719 4.839 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.583 4.690 mp_sum_l 807 5.4 2.988 4.544 2.988 4.544 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.218 4.388 dbcsr_matrix_vector_mult_local 304 10.0 3.695 4.164 3.697 4.167 make_images_sizes 222 9.7 0.000 0.000 1.177 3.683 mp_alltoall_i44 222 10.7 1.177 3.682 1.177 3.682 ls_scf_post 1 4.0 0.000 0.000 3.655 3.659 calculate_norms 792 9.8 3.535 3.647 3.535 3.647 ls_scf_store_result 1 5.0 0.000 0.000 3.422 3.478 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.388 3.468 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.047 3.259 dbcsr_finalize 304 7.8 0.062 0.078 2.193 2.246 dbcsr_merge_all 275 8.9 0.473 0.533 2.042 2.085 dbcsr_data_release 10123 10.8 1.329 2.003 1.329 2.003 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.400000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3615.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/3169bf56b7b69a13599bda2029f1ce159e56e4d8_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 8.759669E+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 0.078 0.102 107.767 107.767 qs_energies 1 2.0 0.000 0.000 106.371 106.380 ls_scf 1 3.0 0.000 0.000 103.503 103.512 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.679 77.813 ls_scf_main 1 4.0 0.000 0.000 65.842 65.842 density_matrix_trs4 2 5.0 0.002 0.003 56.985 57.036 multiply_cannon 111 7.7 0.164 0.203 49.952 51.481 multiply_cannon_loop 111 8.7 0.067 0.069 46.592 47.632 ls_scf_init_scf 1 4.0 0.000 0.000 33.972 33.972 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.638 32.660 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.828 29.841 mp_waitall_1 4527 11.1 22.459 26.062 22.459 26.062 make_m2s 222 7.7 0.005 0.005 23.995 24.896 make_images 222 8.7 4.593 5.002 23.889 24.789 multiply_cannon_multrec 444 9.7 17.892 18.774 22.484 23.149 hybrid_alltoall_any 227 10.6 1.665 3.640 13.014 15.935 make_images_data 222 9.7 0.003 0.003 13.227 15.585 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.655 11.452 multiply_cannon_sync_h2d 444 9.7 8.850 8.902 8.850 8.902 arnoldi_extremal 4 6.8 0.000 0.000 7.388 7.407 arnoldi_normal_ev 4 7.8 0.003 0.009 7.388 7.407 build_subspace 16 8.4 0.026 0.037 6.847 6.861 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.477 5.636 dbcsr_matrix_vector_mult_local 304 10.0 5.035 5.344 5.037 5.346 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.026 5.270 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.805 4.899 dbcsr_mm_accdrv_process 1814 10.4 0.207 0.318 4.421 4.539 dbcsr_mm_accdrv_process_sort 1814 11.4 4.123 4.246 4.123 4.246 ls_scf_post 1 4.0 0.000 0.000 3.690 3.698 make_images_sizes 222 9.7 0.000 0.000 1.425 3.503 mp_alltoall_i44 222 10.7 1.425 3.502 1.425 3.502 ls_scf_store_result 1 5.0 0.000 0.000 3.433 3.443 calculate_norms 792 9.8 3.240 3.268 3.240 3.268 dbcsr_finalize 304 7.8 0.082 0.089 3.090 3.150 mp_allgather_i34 111 8.7 1.005 3.073 1.005 3.073 dbcsr_merge_all 275 8.9 0.890 0.920 2.876 2.930 dbcsr_complete_redistribute 5 7.6 1.432 1.477 2.777 2.923 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.837 2.837 dbcsr_data_release 12724 10.6 2.334 2.831 2.334 2.831 matrix_ls_to_qs 2 6.0 0.000 0.000 2.414 2.564 dbcsr_sort_data 325 11.1 2.447 2.518 2.447 2.518 dbcsr_new_transposed 4 7.5 0.244 0.252 2.296 2.320 dbcsr_frobenius_norm 74 6.6 2.054 2.133 2.197 2.241 dbcsr_add_d 103 6.2 0.000 0.000 2.131 2.205 dbcsr_add_anytype 103 7.2 0.859 0.891 2.131 2.204 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.191 2.193 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.767000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6911.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 3169bf56b7b69a13599bda2029f1ce159e56e4d8 Summary: empty Status: OK