=== 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: 2adf4dececbac3e4208898f648fd9522801a7512 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 21.06.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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.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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/01 job id: 48154565 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/02 job id: 48154566 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/03 job id: 48154567 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/04 job id: 48154568 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/05 job id: 48154569 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/06 job id: 48154570 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/07 job id: 48154571 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/08 job id: 48154572 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/09 job id: 48154573 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/10 job id: 48154574 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/11 job id: 48154575 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/12 job id: 48154576 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/13 job id: 48154577 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/14 job id: 48154578 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/15 job id: 48154579 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/16 job id: 48154580 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/17 job id: 48154581 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/18 job id: 48154582 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/19 job id: 48154583 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/20 job id: 48154585 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/21 job id: 48154586 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/22 job id: 48154587 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/23 job id: 48154588 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/24 job id: 48154589 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/25 job id: 48154590 --- 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/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/26 job id: 48154591 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/27 job id: 48154592 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.078 135.665 135.666 farming_run 1 2.0 135.036 135.038 135.593 135.597 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.505190E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.023 115.288 115.288 qs_energies 1 2.0 0.000 0.000 115.090 115.101 mp2_main 1 3.0 0.000 0.000 113.025 113.035 mp2_gpw_main 1 4.0 0.028 0.033 111.980 111.990 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 92.641 93.027 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.004 54.846 55.232 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.168 41.218 46.297 get_2c_integrals 1 6.0 0.008 0.009 37.352 37.622 integrate_v_rspace 273 8.0 0.435 0.453 24.682 29.587 pw_transfer 6555 10.6 0.375 0.389 27.312 27.896 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 25.922 26.357 grid_integrate_task_list 273 9.0 20.512 25.872 20.512 25.872 fft_wrap_pw1pw2_100 2178 12.4 1.191 1.425 23.442 23.865 compute_2c_integrals 1 7.0 0.013 0.014 19.728 19.729 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.899 19.446 mp2_eri_2c_integrate_gpw 1 9.0 2.354 2.420 18.896 19.442 rpa_ri_compute_en 1 5.0 0.020 0.021 19.221 19.353 cp_fm_cholesky_decompose 12 8.2 17.554 17.807 17.554 17.807 cholesky_decomp 1 7.0 0.000 0.000 16.466 16.717 fft3d_s 5443 13.4 16.059 16.389 16.081 16.411 ao_to_mo_and_store_B_mult_1 272 7.0 10.786 15.325 10.786 15.325 calculate_wavefunction 272 8.0 5.381 5.545 12.443 13.055 rpa_num_int 1 6.0 0.001 0.011 10.933 10.943 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.635 10.670 calc_mat_Q 8 8.0 0.000 0.000 9.502 9.608 contract_S_to_Q 8 9.0 0.000 0.000 8.921 9.026 calc_potential_gpw 544 9.5 0.006 0.007 8.280 8.720 parallel_gemm_fm 14 9.1 0.000 0.000 8.506 8.620 parallel_gemm_fm_cosma 14 10.1 8.506 8.620 8.506 8.620 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.236 8.572 potential_pw2rs 545 10.0 0.108 0.110 7.681 8.322 create_integ_mat 1 6.0 0.013 0.028 7.808 7.818 collocate_single_gaussian 272 10.0 0.040 0.042 7.489 7.797 array2fm 1 7.0 0.000 0.000 6.811 7.259 pw_scatter_s 2720 13.7 4.399 4.520 4.399 4.520 pw_gather_s 2722 13.2 3.862 4.190 3.862 4.190 array2fm_buffer_send 1 8.0 2.944 3.182 2.944 3.182 pw_poisson_solve 545 10.5 1.104 1.133 2.202 2.437 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.990329, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2822.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.038 424.262 424.267 farming_run 1 2.0 423.532 423.545 424.222 424.225 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.235456E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 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 0.033 218.465 218.468 qs_energies 1 2.0 0.003 0.022 218.201 218.217 scf_env_do_scf 1 3.0 0.000 0.000 115.093 115.095 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.187 114.194 rebuild_ks_matrix 4 6.0 0.000 0.000 114.186 114.193 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.060 114.186 114.193 hfx_ks_matrix 4 8.0 0.001 0.001 113.805 113.811 integrate_four_center 4 9.0 0.153 0.483 113.804 113.810 integrate_four_center_main 4 10.0 0.123 0.496 102.100 105.103 integrate_four_center_bin 265 11.0 101.977 105.055 101.977 105.055 mp2_main 1 3.0 0.003 0.028 102.811 102.819 mp2_gpw_main 1 4.0 0.053 0.098 101.909 101.919 init_scf_loop 1 4.0 0.000 0.000 97.055 97.056 mp2_ri_gpw_compute_in 1 5.0 0.072 0.145 74.956 76.105 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.051 55.200 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.162 41.862 46.736 integrate_v_rspace 95 8.0 0.395 0.564 28.250 32.888 pw_transfer 2240 10.6 0.145 0.182 29.916 30.345 fft_wrap_pw1pw2 1868 11.4 0.018 0.023 28.935 29.315 ao_to_mo_and_store_B_mult_1 91 7.0 10.510 29.106 10.510 29.106 mp2_ri_gpw_compute_en 1 5.0 0.060 0.122 26.761 28.507 grid_integrate_task_list 95 9.0 23.554 28.420 23.554 28.420 fft_wrap_pw1pw2_100 730 12.4 1.264 1.477 26.652 27.074 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.838 1.888 24.850 24.860 get_2c_integrals 1 6.0 0.002 0.020 20.786 20.830 compute_2c_integrals 1 7.0 0.055 0.413 19.758 19.769 compute_2c_integrals_loop_lm 1 8.0 0.003 0.015 18.830 19.206 mp2_eri_2c_integrate_gpw 1 9.0 1.730 1.867 18.827 19.205 fft3d_s 1823 13.4 18.437 18.804 18.450 18.817 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 18.037 18.037 calculate_wavefunction 91 8.0 2.007 2.034 9.728 9.938 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.572 8.730 9.447 potential_pw2rs 186 10.0 0.034 0.036 8.611 9.172 local_gemm 172 8.0 8.175 8.878 8.175 8.878 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.236 8.588 mp2_ri_gpw_compute_en_comm 22 7.0 0.499 0.518 7.890 8.270 calc_potential_gpw 182 9.5 0.002 0.003 7.922 8.141 collocate_single_gaussian 91 10.0 0.017 0.026 7.871 8.098 mp_sync 37 10.5 3.693 6.766 3.693 6.766 integrate_four_center_load 4 10.0 0.000 0.000 6.756 6.761 hfx_load_balance 1 11.0 0.000 0.000 6.755 6.761 mp2_ri_gpw_compute_en_ener 172 7.0 6.346 6.410 6.346 6.410 mp_sendrecv_dm3 2068 8.0 5.921 6.295 5.921 6.295 pw_gather_s 912 13.2 4.910 5.374 4.910 5.374 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.907659, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1495.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.501504E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.031 51.434 51.435 qs_mol_dyn_low 1 2.0 0.003 0.003 51.215 51.224 qs_forces 11 3.9 0.002 0.003 50.678 50.680 qs_energies 11 4.9 0.010 0.013 49.221 49.236 scf_env_do_scf 11 5.9 0.000 0.001 41.672 41.673 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 39.682 39.682 dbcsr_multiply_generic 2286 12.5 0.094 0.101 31.934 32.381 qs_scf_new_mos 108 7.5 0.000 0.001 29.784 30.068 qs_scf_loop_do_ot 108 8.5 0.000 0.001 29.784 30.067 ot_scf_mini 108 9.5 0.002 0.002 28.265 28.446 multiply_cannon 2286 13.5 0.187 0.196 26.781 28.134 multiply_cannon_loop 2286 14.5 1.779 1.864 26.126 27.484 velocity_verlet 10 3.0 0.001 0.002 23.603 23.608 ot_mini 108 10.5 0.001 0.001 16.698 16.954 mp_waitall_1 245248 16.5 7.048 14.445 7.048 14.445 qs_ot_get_derivative 108 11.5 0.001 0.001 13.971 14.156 multiply_cannon_metrocomm3 54864 15.5 0.076 0.081 5.456 13.595 multiply_cannon_multrec 54864 15.5 3.969 6.125 8.397 12.205 rebuild_ks_matrix 119 8.3 0.000 0.000 7.804 7.936 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.804 7.936 qs_ot_get_p 119 10.4 0.001 0.001 7.287 7.546 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.865 6.988 dbcsr_mm_accdrv_process 76910 16.1 1.693 2.666 4.340 6.231 init_scf_run 11 5.9 0.000 0.001 6.209 6.209 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.209 6.209 multiply_cannon_sync_h2d 54864 15.5 5.505 5.978 5.505 5.978 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.892 5.343 mp_sum_l 7287 12.8 3.260 5.084 3.260 5.084 qs_ot_p2m_diag 50 11.0 0.004 0.006 4.945 4.974 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.491 4.597 sum_up_and_integrate 119 10.3 0.012 0.014 4.442 4.448 integrate_v_rspace 119 11.3 0.002 0.003 4.430 4.438 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.263 4.263 calculate_first_density_matrix 1 7.0 0.000 0.001 4.136 4.142 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.076 4.124 calculate_rho_elec 119 8.7 0.012 0.017 4.076 4.123 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.069 4.070 cp_fm_redistribute_end 50 14.0 2.078 4.047 2.084 4.049 acc_transpose_blocks 54864 15.5 0.217 0.234 3.031 3.959 cp_fm_diag_elpa_base 50 14.0 1.959 3.941 1.963 3.950 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.689 3.691 jit_kernel_multiply 13 15.8 2.586 3.551 2.586 3.551 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.472 3.517 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.867 3.064 apply_single 119 13.6 0.000 0.000 2.866 3.064 calculate_dm_sparse 119 9.5 0.000 0.001 2.914 3.053 rs_pw_transfer 974 11.9 0.012 0.013 2.655 2.756 ot_diis_step 108 11.5 0.006 0.006 2.560 2.560 acc_transpose_blocks_kernels 54864 16.5 0.233 0.353 1.650 2.343 density_rs2pw 119 9.7 0.004 0.005 2.157 2.276 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.076 2.112 grid_integrate_task_list 119 12.3 2.018 2.105 2.018 2.105 wfi_extrapolate 11 7.9 0.001 0.001 2.001 2.001 jit_kernel_transpose 5 15.5 1.418 1.997 1.418 1.997 multiply_cannon_metrocomm1 54864 15.5 0.059 0.064 0.990 1.987 init_scf_loop 11 6.9 0.000 0.000 1.974 1.974 potential_pw2rs 119 12.3 0.004 0.004 1.801 1.809 pw_transfer 1439 11.6 0.052 0.059 1.686 1.751 mp_sum_d 4135 12.0 1.024 1.685 1.024 1.685 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.611 1.678 make_m2s 4572 13.5 0.054 0.056 1.579 1.624 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.556 1.581 make_images 4572 14.5 0.133 0.140 1.497 1.543 mp_waitany 12084 13.8 1.228 1.418 1.228 1.418 grid_collocate_task_list 119 9.7 1.354 1.416 1.354 1.416 acc_transpose_blocks_sync 164592 16.5 1.121 1.336 1.121 1.336 fft_wrap_pw1pw2_140 487 13.2 0.188 0.205 1.249 1.320 fft3d_ps 1201 14.6 0.369 0.474 1.250 1.316 mp_alltoall_d11v 2130 13.8 1.111 1.278 1.111 1.278 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=51.435000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.415808E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1030290. 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 0.016 0.032 43.285 43.286 qs_mol_dyn_low 1 2.0 0.003 0.005 43.020 43.027 qs_forces 11 3.9 0.002 0.004 42.960 42.961 qs_energies 11 4.9 0.026 0.033 41.230 41.233 scf_env_do_scf 11 5.9 0.001 0.001 34.156 34.158 scf_env_do_scf_inner_loop 108 6.5 0.012 0.078 31.586 31.587 dbcsr_multiply_generic 2286 12.5 0.101 0.104 24.969 25.368 qs_scf_new_mos 108 7.5 0.001 0.001 22.070 22.301 qs_scf_loop_do_ot 108 8.5 0.001 0.001 22.070 22.301 ot_scf_mini 108 9.5 0.003 0.004 21.152 21.317 multiply_cannon 2286 13.5 0.209 0.219 19.492 20.991 multiply_cannon_loop 2286 14.5 1.190 1.246 18.299 19.776 velocity_verlet 10 3.0 0.001 0.002 18.989 18.990 ot_mini 108 10.5 0.044 0.347 13.396 13.626 mp_waitall_1 200699 16.5 6.649 13.038 6.649 13.038 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 5.056 11.606 qs_ot_get_derivative 108 11.5 0.001 0.001 10.663 10.831 multiply_cannon_multrec 27432 15.5 1.906 4.722 6.832 10.117 dbcsr_mm_accdrv_process 47894 16.0 3.779 6.618 4.844 7.667 rebuild_ks_matrix 119 8.3 0.000 0.000 7.395 7.559 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.018 7.395 7.558 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.519 6.669 init_scf_run 11 5.9 0.000 0.001 5.741 5.741 scf_env_initial_rho_setup 11 6.9 0.001 0.006 5.741 5.741 qs_ot_get_p 119 10.4 0.001 0.001 4.825 5.043 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.150 4.963 sum_up_and_integrate 119 10.3 0.025 0.030 4.272 4.279 integrate_v_rspace 119 11.3 0.002 0.002 4.247 4.255 mp_sum_l 7287 12.8 2.355 4.244 2.355 4.244 qs_rho_update_rho_low 119 7.7 0.003 0.021 4.104 4.151 calculate_rho_elec 119 8.7 0.026 0.063 4.101 4.150 calculate_first_density_matrix 1 7.0 0.000 0.003 4.088 4.089 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.289 4.055 apply_single 119 13.6 0.000 0.000 3.289 4.055 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.448 3.450 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.228 3.270 acc_transpose_blocks 27432 15.5 0.115 0.121 2.492 3.244 rs_pw_transfer 974 11.9 0.010 0.011 2.731 3.209 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.136 3.154 make_m2s 4572 13.5 0.052 0.054 2.678 3.016 jit_kernel_multiply 11 16.5 1.007 2.957 1.007 2.957 make_images 4572 14.5 0.207 0.247 2.590 2.930 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.732 2.732 density_rs2pw 119 9.7 0.004 0.004 2.236 2.716 ot_diis_step 108 11.5 0.013 0.026 2.641 2.684 init_scf_loop 11 6.9 0.001 0.005 2.546 2.548 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.364 2.449 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.396 2.397 cp_fm_redistribute_end 50 14.0 1.212 2.363 1.217 2.366 calculate_dm_sparse 119 9.5 0.000 0.001 2.263 2.342 cp_fm_diag_elpa_base 50 14.0 1.114 2.263 1.144 2.301 multiply_cannon_sync_h2d 27432 15.5 1.695 2.270 1.695 2.270 acc_transpose_blocks_kernels 27432 16.5 0.188 0.275 1.536 2.176 pw_transfer 1439 11.6 0.065 0.070 1.996 2.034 potential_pw2rs 119 12.3 0.006 0.006 1.979 1.991 grid_integrate_task_list 119 12.3 1.837 1.944 1.837 1.944 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.904 1.944 jit_kernel_transpose 5 15.5 1.349 1.910 1.349 1.910 make_images_data 4572 15.5 0.048 0.054 1.251 1.600 prepare_preconditioner 11 7.9 0.000 0.001 1.557 1.583 make_preconditioner 11 8.9 0.000 0.001 1.557 1.583 fft_wrap_pw1pw2_140 487 13.2 0.203 0.214 1.484 1.525 wfi_extrapolate 11 7.9 0.001 0.003 1.518 1.518 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.460 1.517 mp_alltoall_d11v 2130 13.8 1.352 1.496 1.352 1.496 fft3d_ps 1201 14.6 0.520 0.577 1.454 1.485 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.115 1.480 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.456 1.465 grid_collocate_task_list 119 9.7 1.280 1.431 1.280 1.431 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.203 1.250 mp_waitany 5720 13.7 0.550 1.060 0.550 1.060 mp_sum_d 4135 12.0 0.629 1.051 0.629 1.051 qs_energies_init_hamiltonians 11 5.9 0.007 0.045 0.990 1.046 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.148 0.574 1.041 mp_allgather_i34 2286 14.5 0.613 1.009 0.613 1.009 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.935 0.949 acc_transpose_blocks_sync 82296 16.5 0.816 0.938 0.816 0.938 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=43.286000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.363636, yerr=1.067940 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_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 519.680000E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 2188913. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.054 35.102 35.104 qs_mol_dyn_low 1 2.0 0.004 0.008 34.757 34.765 qs_forces 11 3.9 0.004 0.009 34.656 34.659 qs_energies 11 4.9 0.002 0.005 33.058 33.063 scf_env_do_scf 11 5.9 0.001 0.002 26.434 26.434 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.936 23.937 dbcsr_multiply_generic 2286 12.5 0.095 0.099 18.928 19.017 multiply_cannon 2286 13.5 0.193 0.197 15.642 16.364 qs_scf_new_mos 108 7.5 0.001 0.001 15.735 15.754 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.735 15.754 multiply_cannon_loop 2286 14.5 0.857 0.900 14.867 15.621 velocity_verlet 10 3.0 0.002 0.006 15.499 15.502 ot_scf_mini 108 9.5 0.002 0.003 14.980 14.997 ot_mini 108 10.5 0.001 0.001 9.400 9.417 multiply_cannon_multrec 18288 15.5 1.888 2.901 7.548 7.966 qs_ot_get_derivative 108 11.5 0.001 0.002 7.908 7.926 dbcsr_mm_accdrv_process 38222 16.0 5.455 6.489 5.565 6.554 rebuild_ks_matrix 119 8.3 0.000 0.000 6.443 6.459 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.030 6.442 6.459 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.686 5.701 init_scf_run 11 5.9 0.000 0.001 5.291 5.291 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.291 5.291 mp_waitall_1 158411 16.6 2.851 4.050 2.851 4.050 sum_up_and_integrate 119 10.3 0.032 0.033 3.967 3.972 calculate_first_density_matrix 1 7.0 0.001 0.005 3.950 3.951 integrate_v_rspace 119 11.3 0.003 0.004 3.935 3.945 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.151 3.764 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.545 3.549 calculate_rho_elec 119 8.7 0.031 0.032 3.544 3.549 qs_ot_get_p 119 10.4 0.001 0.001 3.341 3.370 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.129 3.132 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.963 2.971 acc_transpose_blocks 18288 15.5 0.080 0.082 2.792 2.900 rs_pw_transfer 974 11.9 0.009 0.010 2.391 2.713 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.644 2.711 init_scf_loop 11 6.9 0.001 0.005 2.480 2.482 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.026 2.344 apply_single 119 13.6 0.000 0.000 2.026 2.343 density_rs2pw 119 9.7 0.004 0.004 2.017 2.313 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.246 2.252 calculate_dm_sparse 119 9.5 0.000 0.000 2.095 2.100 acc_transpose_blocks_kernels 18288 16.5 0.217 0.224 1.957 2.021 make_m2s 4572 13.5 0.044 0.046 1.829 1.984 cp_dbcsr_syevd 50 12.0 0.003 0.004 1.952 1.952 pw_transfer 1439 11.6 0.065 0.069 1.925 1.936 grid_integrate_task_list 119 12.3 1.807 1.908 1.807 1.908 make_images 4572 14.5 0.196 0.208 1.745 1.899 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.832 1.846 jit_kernel_transpose 5 15.6 1.740 1.804 1.740 1.804 potential_pw2rs 119 12.3 0.007 0.008 1.719 1.723 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.711 1.719 cp_fm_diag_elpa_base 50 14.0 1.687 1.698 1.708 1.716 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.706 1.714 prepare_preconditioner 11 7.9 0.000 0.000 1.647 1.651 make_preconditioner 11 8.9 0.000 0.001 1.647 1.651 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.502 1.590 ot_diis_step 108 11.5 0.012 0.016 1.477 1.477 fft_wrap_pw1pw2_140 487 13.2 0.256 0.264 1.453 1.466 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.435 1.440 grid_collocate_task_list 119 9.7 1.262 1.410 1.262 1.410 mp_sum_l 7287 12.8 1.077 1.394 1.077 1.394 fft3d_ps 1201 14.6 0.529 0.549 1.318 1.332 multiply_cannon_sync_h2d 18288 15.5 1.063 1.199 1.063 1.199 wfi_extrapolate 11 7.9 0.001 0.001 1.191 1.191 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.104 1.108 mp_waitany 9880 13.7 0.663 1.002 0.663 1.002 rs_pw_transfer_RS2PW_140 130 11.5 0.120 0.124 0.651 0.969 make_images_data 4572 15.5 0.047 0.051 0.785 0.955 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.934 0.952 hybrid_alltoall_any 4725 16.4 0.058 0.115 0.679 0.878 jit_kernel_multiply 4 15.0 0.053 0.843 0.053 0.843 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.829 0.830 acc_transpose_blocks_sync 54864 16.5 0.736 0.807 0.736 0.807 mp_alltoall_d11v 2130 13.8 0.641 0.799 0.641 0.799 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.676 0.746 qs_env_update_s_mstruct 11 6.9 0.006 0.011 0.638 0.703 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=35.104000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.545455, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_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 559.022080E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.223 0.281 38.410 38.413 qs_mol_dyn_low 1 2.0 0.003 0.003 37.700 37.712 qs_forces 11 3.9 0.003 0.003 37.623 37.626 qs_energies 11 4.9 0.002 0.005 35.863 35.871 scf_env_do_scf 11 5.9 0.001 0.001 30.300 30.301 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.897 26.898 dbcsr_multiply_generic 2286 12.5 0.098 0.103 20.951 21.149 velocity_verlet 10 3.0 0.022 0.022 18.580 18.582 qs_scf_new_mos 108 7.5 0.001 0.001 18.444 18.496 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.444 18.496 multiply_cannon 2286 13.5 0.217 0.225 17.334 17.860 ot_scf_mini 108 9.5 0.002 0.003 17.422 17.480 multiply_cannon_loop 2286 14.5 1.521 1.602 16.387 16.818 ot_mini 108 10.5 0.001 0.001 11.014 11.086 multiply_cannon_multrec 27432 15.5 2.791 3.706 10.056 10.428 qs_ot_get_derivative 108 11.5 0.001 0.001 9.158 9.217 dbcsr_mm_accdrv_process 47916 15.9 6.527 8.704 7.162 8.856 rebuild_ks_matrix 119 8.3 0.000 0.000 6.624 6.675 qs_ks_build_kohn_sham_matrix 119 9.3 0.017 0.019 6.623 6.675 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.871 5.915 init_scf_run 11 5.9 0.000 0.001 4.208 4.208 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.207 4.208 sum_up_and_integrate 119 10.3 0.036 0.038 3.793 3.801 integrate_v_rspace 119 11.3 0.003 0.004 3.757 3.765 qs_ot_get_p 119 10.4 0.001 0.001 3.563 3.641 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.212 3.612 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.514 3.548 calculate_rho_elec 119 8.7 0.040 0.046 3.513 3.547 init_scf_loop 11 6.9 0.001 0.003 3.383 3.383 calculate_first_density_matrix 1 7.0 0.000 0.003 2.805 2.807 calculate_dm_sparse 119 9.5 0.000 0.000 2.673 2.722 mp_waitall_1 137007 16.6 1.930 2.593 1.930 2.593 prepare_preconditioner 11 7.9 0.000 0.000 2.520 2.530 make_preconditioner 11 8.9 0.000 0.001 2.520 2.530 acc_transpose_blocks 27432 15.5 0.117 0.121 2.387 2.490 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.128 2.456 make_m2s 4572 13.5 0.054 0.056 2.236 2.364 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.314 2.345 jit_kernel_multiply 9 16.0 0.569 2.258 0.569 2.258 make_images 4572 14.5 0.275 0.338 2.129 2.255 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.097 2.215 apply_single 119 13.6 0.000 0.000 2.096 2.215 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.183 2.195 rs_pw_transfer 974 11.9 0.009 0.010 1.951 2.148 density_rs2pw 119 9.7 0.004 0.004 1.904 2.087 pw_transfer 1439 11.6 0.065 0.070 1.988 2.022 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.896 1.934 grid_integrate_task_list 119 12.3 1.828 1.905 1.828 1.905 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.827 1.828 ot_diis_step 108 11.5 0.012 0.012 1.813 1.814 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.777 1.787 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.782 1.782 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.648 1.660 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.926 1.653 fft_wrap_pw1pw2_140 487 13.2 0.289 0.300 1.566 1.605 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.546 1.555 cp_fm_diag_elpa_base 50 14.0 1.512 1.528 1.544 1.553 potential_pw2rs 119 12.3 0.008 0.010 1.539 1.541 acc_transpose_blocks_sync 82296 16.5 1.410 1.506 1.410 1.506 grid_collocate_task_list 119 9.7 1.247 1.379 1.247 1.379 fft3d_ps 1201 14.6 0.556 0.605 1.334 1.358 wfi_extrapolate 11 7.9 0.001 0.001 1.347 1.347 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.280 1.299 mp_sum_l 7287 12.8 1.013 1.294 1.013 1.294 cp_fm_upper_to_full 72 14.2 0.810 1.155 0.810 1.155 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.087 1.090 dbcsr_complete_redistribute 329 12.2 0.136 0.167 0.767 1.039 make_images_data 4572 15.5 0.047 0.051 0.826 0.950 hybrid_alltoall_any 4725 16.4 0.065 0.154 0.710 0.889 multiply_cannon_metrocomm1 27432 15.5 0.035 0.036 0.228 0.886 build_core_hamiltonian_matrix_ 11 4.9 0.004 0.026 0.802 0.876 acc_transpose_blocks_kernels 27432 16.5 0.270 0.279 0.830 0.854 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.844 0.849 mp_alltoall_d11v 2130 13.8 0.709 0.847 0.709 0.847 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.571 0.837 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=38.413000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.909091, yerr=3.824670 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_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 607.678464E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.032 30.221 30.222 qs_mol_dyn_low 1 2.0 0.003 0.003 29.802 29.810 qs_forces 11 3.9 0.002 0.002 29.735 29.743 qs_energies 11 4.9 0.001 0.001 28.011 28.021 scf_env_do_scf 11 5.9 0.000 0.001 23.164 23.164 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 20.349 20.350 velocity_verlet 10 3.0 0.002 0.002 14.889 14.892 dbcsr_multiply_generic 2286 12.5 0.091 0.096 13.694 13.782 qs_scf_new_mos 108 7.5 0.001 0.001 12.350 12.374 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.349 12.373 ot_scf_mini 108 9.5 0.002 0.002 11.643 11.665 multiply_cannon 2286 13.5 0.222 0.228 11.072 11.465 multiply_cannon_loop 2286 14.5 0.646 0.670 10.162 10.362 ot_mini 108 10.5 0.001 0.001 6.888 6.916 multiply_cannon_multrec 9144 15.5 1.660 1.887 6.318 6.570 rebuild_ks_matrix 119 8.3 0.000 0.000 6.349 6.375 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.349 6.375 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.687 5.709 qs_ot_get_derivative 108 11.5 0.001 0.001 5.597 5.620 dbcsr_mm_accdrv_process 12550 15.8 3.576 4.529 4.547 4.601 sum_up_and_integrate 119 10.3 0.038 0.041 3.657 3.662 integrate_v_rspace 119 11.3 0.003 0.003 3.619 3.624 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.547 3.554 calculate_rho_elec 119 8.7 0.060 0.061 3.547 3.553 init_scf_run 11 5.9 0.000 0.001 3.394 3.394 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.394 3.394 qs_ot_get_p 119 10.4 0.001 0.001 2.877 2.908 init_scf_loop 11 6.9 0.000 0.000 2.795 2.797 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.278 2.295 calculate_first_density_matrix 1 7.0 0.000 0.000 2.238 2.238 jit_kernel_multiply 10 15.8 0.931 2.149 0.931 2.149 pw_transfer 1439 11.6 0.066 0.068 2.029 2.042 density_rs2pw 119 9.7 0.004 0.004 1.843 1.986 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.936 1.949 make_m2s 4572 13.5 0.034 0.036 1.792 1.947 grid_integrate_task_list 119 12.3 1.844 1.926 1.844 1.926 calculate_dm_sparse 119 9.5 0.000 0.000 1.908 1.926 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.894 1.898 make_images 4572 14.5 0.272 0.303 1.705 1.857 rs_pw_transfer 974 11.9 0.008 0.009 1.693 1.817 acc_transpose_blocks 9144 15.5 0.042 0.044 1.753 1.798 mp_waitall_1 115863 16.7 1.328 1.756 1.328 1.756 prepare_preconditioner 11 7.9 0.000 0.000 1.679 1.682 make_preconditioner 11 8.9 0.000 0.000 1.679 1.682 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.652 1.653 fft_wrap_pw1pw2_140 487 13.2 0.365 0.374 1.594 1.608 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.572 1.598 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.535 1.536 grid_collocate_task_list 119 9.7 1.296 1.415 1.296 1.415 potential_pw2rs 119 12.3 0.010 0.011 1.388 1.391 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.361 1.370 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.353 1.365 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.354 1.362 cp_fm_diag_elpa_base 50 14.0 1.327 1.344 1.353 1.361 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.297 1.302 fft3d_ps 1201 14.6 0.559 0.570 1.268 1.281 ot_diis_step 108 11.5 0.012 0.012 1.275 1.276 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.241 1.249 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.214 1.248 apply_single 119 13.6 0.000 0.000 1.214 1.248 wfi_extrapolate 11 7.9 0.001 0.001 1.102 1.102 hybrid_alltoall_any 4725 16.4 0.064 0.175 0.771 0.985 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.967 0.970 make_images_data 4572 15.5 0.042 0.045 0.765 0.925 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.872 0.921 mp_alltoall_d11v 2130 13.8 0.797 0.884 0.797 0.884 jit_kernel_transpose 5 15.6 0.850 0.853 0.850 0.853 cp_fm_cholesky_invert 11 10.9 0.838 0.841 0.838 0.841 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.819 0.827 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.784 0.788 acc_transpose_blocks_sync 27432 16.5 0.728 0.771 0.728 0.771 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.688 0.735 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.346 0.677 mp_allgather_i34 2286 14.5 0.229 0.647 0.229 0.647 mp_alltoall_z22v 1201 16.6 0.584 0.614 0.584 0.614 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.222000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=575.909091, yerr=6.761583 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_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 768.827392E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.039 44.034 44.036 qs_mol_dyn_low 1 2.0 0.003 0.003 43.823 43.831 qs_forces 11 3.9 0.002 0.002 43.754 43.763 qs_energies 11 4.9 0.001 0.001 41.739 41.751 scf_env_do_scf 11 5.9 0.001 0.001 35.386 35.387 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.515 27.517 velocity_verlet 10 3.0 0.001 0.002 24.261 24.266 dbcsr_multiply_generic 2286 12.5 0.100 0.104 19.866 19.997 qs_scf_new_mos 108 7.5 0.001 0.001 17.684 17.776 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.684 17.776 multiply_cannon 2286 13.5 0.294 0.298 15.879 16.663 ot_scf_mini 108 9.5 0.002 0.002 16.526 16.625 multiply_cannon_loop 2286 14.5 0.869 0.888 14.600 15.412 ot_mini 108 10.5 0.001 0.001 10.163 10.280 multiply_cannon_multrec 9144 15.5 3.491 4.949 9.333 9.674 qs_ot_get_derivative 108 11.5 0.001 0.001 8.079 8.177 init_scf_loop 11 6.9 0.000 0.000 7.843 7.845 rebuild_ks_matrix 119 8.3 0.000 0.000 7.364 7.508 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.364 7.507 dbcsr_mm_accdrv_process 12550 15.8 4.872 6.974 5.702 7.024 prepare_preconditioner 11 7.9 0.000 0.000 6.858 6.871 make_preconditioner 11 8.9 0.000 0.000 6.858 6.871 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.637 6.766 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.416 6.751 cp_fm_upper_to_full 72 14.2 3.193 4.579 3.193 4.579 init_scf_run 11 5.9 0.000 0.001 4.301 4.301 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.301 4.301 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.278 4.284 calculate_rho_elec 119 8.7 0.118 0.121 4.278 4.284 sum_up_and_integrate 119 10.3 0.064 0.066 4.026 4.033 integrate_v_rspace 119 11.3 0.003 0.003 3.961 3.968 qs_ot_get_p 119 10.4 0.001 0.001 3.416 3.556 mp_waitall_1 94719 16.7 2.533 3.479 2.533 3.479 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.832 3.222 dbcsr_complete_redistribute 329 12.2 0.314 0.323 2.020 2.847 calculate_first_density_matrix 1 7.0 0.000 0.000 2.827 2.828 pw_transfer 1439 11.6 0.068 0.069 2.673 2.676 calculate_dm_sparse 119 9.5 0.000 0.000 2.574 2.591 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.574 2.579 make_m2s 4572 13.5 0.037 0.037 2.396 2.568 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.211 2.513 apply_single 119 13.6 0.000 0.000 2.211 2.513 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.703 2.509 make_images 4572 14.5 0.356 0.392 2.275 2.447 mp_alltoall_i22 627 13.8 1.414 2.252 1.414 2.252 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.437 2.234 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.465 2.225 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.180 2.181 density_rs2pw 119 9.7 0.004 0.004 2.162 2.180 fft_wrap_pw1pw2_140 487 13.2 0.617 0.621 2.170 2.175 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.076 2.127 grid_integrate_task_list 119 12.3 2.077 2.103 2.077 2.103 ot_diis_step 108 11.5 0.014 0.014 2.056 2.056 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.040 2.041 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.914 1.958 acc_transpose_blocks 9144 15.5 0.045 0.046 1.880 1.914 mp_sum_l 7287 12.8 1.050 1.785 1.050 1.785 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.769 1.778 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.679 1.680 fft3d_ps 1201 14.6 0.596 0.606 1.591 1.596 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.580 1.588 grid_collocate_task_list 119 9.7 1.521 1.545 1.521 1.545 rs_pw_transfer 974 11.9 0.009 0.009 1.458 1.491 cp_fm_cholesky_invert 11 10.9 1.450 1.453 1.450 1.453 wfi_extrapolate 11 7.9 0.001 0.001 1.405 1.405 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.389 1.390 potential_pw2rs 119 12.3 0.014 0.014 1.386 1.390 cp_fm_diag_elpa_base 50 14.0 1.241 1.295 1.388 1.388 jit_kernel_multiply 6 16.0 0.801 1.379 0.801 1.379 hybrid_alltoall_any 4725 16.4 0.089 0.147 1.073 1.343 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.307 1.327 make_images_data 4572 15.5 0.045 0.049 1.025 1.242 acc_transpose_blocks_sync 27432 16.5 1.105 1.134 1.105 1.134 mp_alltoall_d11v 2130 13.8 1.104 1.129 1.104 1.129 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.096 1.110 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.989 1.027 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.939 0.953 qs_create_task_list 11 7.9 0.000 0.000 0.930 0.941 generate_qs_task_list 11 8.9 0.369 0.388 0.929 0.940 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.036000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=721.363636, yerr=15.334292 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.186560E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 900443. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.031 82.814 82.816 qs_mol_dyn_low 1 2.0 0.003 0.003 82.503 82.515 qs_forces 11 3.9 0.003 0.003 82.430 82.438 qs_energies 11 4.9 0.002 0.003 79.510 79.528 scf_env_do_scf 11 5.9 0.000 0.001 69.743 69.746 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.212 64.212 dbcsr_multiply_generic 2055 12.4 0.105 0.107 51.618 51.888 qs_scf_new_mos 99 7.5 0.000 0.001 46.786 46.980 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.786 46.980 ot_scf_mini 99 9.5 0.002 0.002 44.446 44.575 multiply_cannon 2055 13.4 0.185 0.188 43.406 44.289 multiply_cannon_loop 2055 14.4 1.784 1.833 42.423 43.307 velocity_verlet 10 3.0 0.001 0.002 42.665 42.666 ot_mini 99 10.5 0.001 0.001 26.546 26.673 qs_ot_get_derivative 99 11.5 0.001 0.001 19.829 19.973 multiply_cannon_multrec 49320 15.4 11.541 12.488 18.212 18.771 rebuild_ks_matrix 110 8.3 0.000 0.000 14.679 14.867 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.030 14.679 14.867 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.837 13.001 mp_waitall_1 220248 16.4 10.411 11.616 10.411 11.616 multiply_cannon_sync_h2d 49320 15.4 9.768 10.377 9.768 10.377 qs_ot_get_p 110 10.4 0.001 0.001 9.592 9.735 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.746 8.424 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.333 7.757 apply_single 110 13.6 0.000 0.000 7.333 7.757 init_scf_run 11 5.9 0.000 0.001 7.704 7.704 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.703 7.704 multiply_cannon_metrocomm3 49320 15.4 0.085 0.090 6.301 7.474 dbcsr_mm_accdrv_process 87628 16.1 2.932 3.077 6.539 7.223 sum_up_and_integrate 110 10.3 0.036 0.042 7.160 7.177 integrate_v_rspace 110 11.3 0.003 0.003 7.124 7.148 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.651 6.761 calculate_rho_elec 110 8.6 0.021 0.026 6.651 6.760 ot_diis_step 99 11.5 0.005 0.006 6.565 6.565 qs_ot_p2m_diag 48 11.0 0.013 0.019 6.497 6.517 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.705 5.705 init_scf_loop 11 6.9 0.000 0.000 5.504 5.504 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.356 5.418 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.135 5.163 cp_fm_diag_elpa_base 48 14.0 5.122 5.149 5.133 5.162 mp_sum_l 6594 12.7 3.809 4.428 3.809 4.428 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.364 4.368 rs_pw_transfer 902 11.9 0.011 0.013 3.721 4.230 wfi_extrapolate 11 7.9 0.001 0.001 4.085 4.085 make_m2s 4110 13.4 0.061 0.065 3.960 4.073 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.002 4.072 make_images 4110 14.4 0.176 0.189 3.865 3.981 calculate_dm_sparse 110 9.5 0.001 0.001 3.821 3.941 density_rs2pw 110 9.6 0.004 0.005 3.414 3.922 jit_kernel_multiply 13 15.9 3.331 3.709 3.331 3.709 calculate_first_density_matrix 1 7.0 0.000 0.001 3.512 3.521 multiply_cannon_metrocomm1 49320 15.4 0.066 0.069 2.254 3.513 grid_integrate_task_list 110 12.3 3.292 3.441 3.292 3.441 prepare_preconditioner 11 7.9 0.000 0.000 3.315 3.332 make_preconditioner 11 8.9 0.000 0.000 3.315 3.332 pw_transfer 1331 11.6 0.055 0.066 3.180 3.256 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.091 3.169 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.087 3.130 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.005 3.049 acc_transpose_blocks 49320 15.4 0.210 0.220 2.778 2.864 fft_wrap_pw1pw2_140 451 13.1 0.458 0.502 2.638 2.731 potential_pw2rs 110 12.3 0.006 0.007 2.654 2.686 mp_waitany 14300 13.8 1.852 2.451 1.852 2.451 grid_collocate_task_list 110 9.6 2.158 2.376 2.158 2.376 mp_alltoall_d11v 2046 13.8 2.035 2.363 2.035 2.363 fft3d_ps 1111 14.6 0.784 0.871 2.276 2.337 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.973 1.997 make_images_data 4110 15.4 0.044 0.048 1.782 1.913 hybrid_alltoall_any 4261 16.3 0.085 0.492 1.553 1.840 cp_fm_cholesky_invert 11 10.9 1.789 1.792 1.789 1.792 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.756 1.784 mp_sum_d 3889 11.9 1.231 1.728 1.231 1.728 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.816000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.090909, yerr=2.193152 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_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 592.994304E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2419731. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.055 71.632 71.633 qs_mol_dyn_low 1 2.0 0.003 0.003 71.201 71.212 qs_forces 11 3.9 0.003 0.004 70.634 70.642 qs_energies 11 4.9 0.001 0.002 67.272 67.281 scf_env_do_scf 11 5.9 0.000 0.001 57.156 57.160 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.595 49.596 dbcsr_multiply_generic 2055 12.4 0.114 0.118 39.137 39.292 velocity_verlet 10 3.0 0.140 0.372 36.706 36.710 multiply_cannon 2055 13.4 0.223 0.243 32.613 33.669 qs_scf_new_mos 99 7.5 0.001 0.001 33.217 33.345 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.216 33.345 multiply_cannon_loop 2055 14.4 1.159 1.182 31.476 32.243 ot_scf_mini 99 9.5 0.003 0.003 31.554 31.675 ot_mini 99 10.5 0.001 0.001 18.425 18.559 multiply_cannon_multrec 24660 15.4 7.055 8.947 15.302 17.129 rebuild_ks_matrix 110 8.3 0.000 0.000 13.822 13.913 qs_ks_build_kohn_sham_matrix 110 9.3 0.022 0.042 13.822 13.913 qs_ot_get_derivative 99 11.5 0.001 0.001 12.650 12.778 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.145 12.222 mp_waitall_1 176588 16.5 7.527 9.989 7.527 9.989 dbcsr_mm_accdrv_process 52282 16.1 6.529 7.808 8.077 8.372 multiply_cannon_metrocomm3 24660 15.4 0.072 0.077 5.184 7.885 init_scf_run 11 5.9 0.000 0.001 7.747 7.748 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.747 7.748 init_scf_loop 11 6.9 0.000 0.000 7.526 7.527 multiply_cannon_sync_h2d 24660 15.4 6.469 7.504 6.469 7.504 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.495 7.042 apply_single 110 13.6 0.000 0.001 6.495 7.042 qs_ot_get_p 110 10.4 0.001 0.001 6.484 6.659 sum_up_and_integrate 110 10.3 0.052 0.059 6.646 6.656 integrate_v_rspace 110 11.3 0.003 0.003 6.594 6.608 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.221 6.230 calculate_rho_elec 110 8.6 0.040 0.048 6.220 6.230 ot_diis_step 99 11.5 0.010 0.010 5.729 5.730 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.767 5.524 prepare_preconditioner 11 7.9 0.000 0.000 5.497 5.519 make_preconditioner 11 8.9 0.000 0.000 5.497 5.519 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.080 5.243 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.616 4.635 make_m2s 4110 13.4 0.057 0.060 4.167 4.499 make_images 4110 14.4 0.415 0.474 4.061 4.390 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 4.256 4.257 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.152 4.152 calculate_first_density_matrix 1 7.0 0.000 0.000 4.141 4.143 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.907 3.944 density_rs2pw 110 9.6 0.004 0.005 3.293 3.862 pw_transfer 1331 11.6 0.067 0.074 3.633 3.776 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.525 3.670 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.572 3.587 cp_fm_diag_elpa_base 48 14.0 3.524 3.536 3.569 3.583 rs_pw_transfer 902 11.9 0.012 0.015 2.971 3.548 wfi_extrapolate 11 7.9 0.001 0.001 3.508 3.508 grid_integrate_task_list 110 12.3 3.158 3.337 3.158 3.337 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.278 3.337 calculate_dm_sparse 110 9.5 0.001 0.001 3.253 3.284 fft_wrap_pw1pw2_140 451 13.1 0.521 0.535 3.018 3.163 fft3d_ps 1111 14.6 1.107 1.335 2.497 2.654 make_images_data 4110 15.4 0.049 0.054 2.253 2.603 cp_fm_cholesky_invert 11 10.9 2.569 2.578 2.569 2.578 grid_collocate_task_list 110 9.6 2.106 2.532 2.106 2.532 hybrid_alltoall_any 4261 16.3 0.107 0.459 1.926 2.524 jit_kernel_multiply 9 16.3 1.189 2.511 1.189 2.511 potential_pw2rs 110 12.3 0.008 0.009 2.421 2.433 mp_sum_l 6594 12.7 1.727 2.383 1.727 2.383 acc_transpose_blocks 24660 15.4 0.116 0.120 2.008 2.033 mp_alltoall_d11v 2046 13.8 1.713 2.013 1.713 2.013 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.968 1.992 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.864 1.871 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.834 1.845 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.826 1.836 multiply_cannon_metrocomm4 22605 15.4 0.079 0.084 0.762 1.823 mp_waitany 10164 13.8 1.224 1.767 1.224 1.767 mp_irecv_dv 57340 16.2 0.632 1.598 0.632 1.598 rs_pw_transfer_RS2PW_140 121 11.5 0.206 0.216 0.970 1.540 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.386 1.490 mp_allgather_i34 2055 14.4 0.447 1.433 0.447 1.433 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.633000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=559.272727, yerr=7.921099 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 665.501696E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.099 63.701 63.702 qs_mol_dyn_low 1 2.0 0.003 0.004 61.826 61.837 qs_forces 11 3.9 0.006 0.026 61.758 61.761 qs_energies 11 4.9 0.010 0.070 58.485 58.499 scf_env_do_scf 11 5.9 0.002 0.014 49.957 49.957 scf_env_do_scf_inner_loop 99 6.5 0.008 0.011 41.205 41.207 velocity_verlet 10 3.0 0.001 0.002 33.228 33.230 dbcsr_multiply_generic 2055 12.4 0.106 0.110 29.628 29.912 qs_scf_new_mos 99 7.5 0.001 0.001 25.915 26.025 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.915 26.025 ot_scf_mini 99 9.5 0.002 0.003 24.664 24.794 multiply_cannon 2055 13.4 0.211 0.221 23.089 24.430 multiply_cannon_loop 2055 14.4 0.818 0.869 21.854 22.989 ot_mini 99 10.5 0.001 0.001 14.089 14.226 rebuild_ks_matrix 110 8.3 0.000 0.000 12.565 12.714 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.565 12.713 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.044 11.181 multiply_cannon_multrec 16440 15.4 3.546 4.403 10.176 11.052 mp_waitall_1 139946 16.5 7.201 10.515 7.201 10.515 qs_ot_get_derivative 99 11.5 0.001 0.001 9.544 9.674 init_scf_loop 11 6.9 0.004 0.027 8.717 8.718 multiply_cannon_metrocomm3 16440 15.4 0.045 0.047 4.554 7.601 prepare_preconditioner 11 7.9 0.000 0.000 6.802 6.823 make_preconditioner 11 8.9 0.000 0.001 6.802 6.823 dbcsr_mm_accdrv_process 34862 16.1 5.682 6.150 6.473 6.678 sum_up_and_integrate 110 10.3 0.060 0.061 6.615 6.632 integrate_v_rspace 110 11.3 0.003 0.003 6.554 6.570 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.128 6.479 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.110 6.121 calculate_rho_elec 110 8.6 0.059 0.060 6.109 6.120 init_scf_run 11 5.9 0.000 0.001 6.020 6.021 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.020 6.020 qs_ot_get_p 110 10.4 0.001 0.001 5.743 5.918 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.048 5.514 apply_single 110 13.6 0.000 0.000 5.048 5.514 make_m2s 4110 13.4 0.050 0.051 4.341 4.754 make_images 4110 14.4 0.404 0.529 4.229 4.643 ot_diis_step 99 11.5 0.010 0.011 4.511 4.512 density_rs2pw 110 9.6 0.004 0.005 3.158 4.465 qs_ot_p2m_diag 48 11.0 0.063 0.211 4.050 4.060 rs_pw_transfer 902 11.9 0.010 0.011 2.747 4.059 multiply_cannon_sync_h2d 16440 15.4 3.207 3.919 3.207 3.919 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.192 3.849 pw_transfer 1331 11.6 0.066 0.072 3.693 3.700 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.649 3.671 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.586 3.595 grid_integrate_task_list 110 12.3 3.185 3.488 3.185 3.488 fft_wrap_pw1pw2_140 451 13.1 0.641 0.650 3.078 3.088 wfi_extrapolate 11 7.9 0.001 0.001 3.051 3.051 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.954 2.964 cp_fm_diag_elpa_base 48 14.0 2.885 2.918 2.952 2.962 make_images_data 4110 15.4 0.046 0.050 2.446 2.921 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.890 2.891 calculate_dm_sparse 110 9.5 0.001 0.001 2.859 2.890 calculate_first_density_matrix 1 7.0 0.000 0.002 2.873 2.875 hybrid_alltoall_any 4261 16.3 0.109 0.379 2.187 2.785 cp_fm_cholesky_invert 11 10.9 2.664 2.670 2.664 2.670 mp_waitany 17072 13.8 1.228 2.610 1.228 2.610 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.545 2.606 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.494 2.558 grid_collocate_task_list 110 9.6 2.127 2.520 2.127 2.520 multiply_cannon_metrocomm4 14385 15.4 0.047 0.051 0.913 2.484 fft3d_ps 1111 14.6 1.089 1.098 2.404 2.417 mp_sum_l 6594 12.7 1.620 2.356 1.620 2.356 mp_irecv_dv 48980 15.7 0.839 2.349 0.839 2.349 potential_pw2rs 110 12.3 0.011 0.011 2.317 2.323 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.180 0.932 2.245 mp_alltoall_d11v 2046 13.8 1.830 2.208 1.830 2.208 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.996 1.999 dbcsr_complete_redistribute 325 12.2 0.330 0.360 1.483 1.944 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.795 1.808 mp_allgather_i34 2055 14.4 0.549 1.756 0.549 1.756 cp_fm_upper_to_full 70 14.2 1.375 1.748 1.375 1.748 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.717 1.733 acc_transpose_blocks 16440 15.4 0.080 0.083 1.596 1.722 cp_fm_cholesky_decompose 22 10.9 1.605 1.626 1.605 1.626 jit_kernel_multiply 6 16.8 0.393 1.611 0.393 1.611 create_qs_kind_set 1 2.0 0.000 0.001 1.526 1.535 read_qs_kind 2 3.0 0.013 0.037 1.526 1.535 parser_read_line 2821 4.0 0.001 0.001 1.513 1.527 parser_read_line_low 5 5.0 0.047 1.488 1.512 1.527 broadcast_input_information 5 6.0 0.002 0.002 1.465 1.526 mp_bcast_i_src 20 7.0 1.458 1.515 1.458 1.515 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.372 1.466 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.009 1.455 rs_gather_matrices 110 12.3 0.232 0.259 0.975 1.306 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.296 1.306 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.702000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.272727, yerr=8.791431 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 737.947648E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.041 68.846 68.847 qs_mol_dyn_low 1 2.0 0.003 0.004 68.232 68.452 qs_forces 11 3.9 0.004 0.005 67.712 67.716 qs_energies 11 4.9 0.008 0.054 64.248 64.255 scf_env_do_scf 11 5.9 0.001 0.002 54.827 54.829 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 42.921 42.921 velocity_verlet 10 3.0 0.280 0.448 38.425 38.427 dbcsr_multiply_generic 2055 12.4 0.112 0.119 31.306 31.519 qs_scf_new_mos 99 7.5 0.001 0.001 27.709 27.828 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.708 27.828 ot_scf_mini 99 9.5 0.003 0.003 26.010 26.091 multiply_cannon 2055 13.4 0.238 0.258 24.110 25.427 multiply_cannon_loop 2055 14.4 1.403 1.492 22.645 23.351 ot_mini 99 10.5 0.001 0.001 14.936 15.042 multiply_cannon_multrec 24660 15.4 4.095 6.840 13.506 14.458 rebuild_ks_matrix 110 8.3 0.000 0.000 12.419 12.527 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.027 12.418 12.526 init_scf_loop 11 6.9 0.005 0.035 11.864 11.865 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.981 11.078 qs_ot_get_derivative 99 11.5 0.001 0.001 10.691 10.775 dbcsr_mm_accdrv_process 52304 16.0 8.120 9.512 9.252 10.178 prepare_preconditioner 11 7.9 0.000 0.000 9.977 9.993 make_preconditioner 11 8.9 0.000 0.001 9.977 9.992 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.185 9.651 sum_up_and_integrate 110 10.3 0.068 0.071 6.631 6.647 init_scf_run 11 5.9 0.000 0.001 6.586 6.587 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.586 6.586 integrate_v_rspace 110 11.3 0.003 0.003 6.563 6.576 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.235 6.247 calculate_rho_elec 110 8.6 0.078 0.082 6.235 6.246 qs_ot_get_p 110 10.4 0.001 0.001 5.913 6.041 mp_waitall_1 121746 16.5 4.250 6.036 4.250 6.036 make_m2s 4110 13.4 0.059 0.062 5.504 5.865 make_images 4110 14.4 0.586 0.711 5.366 5.724 cp_fm_upper_to_full 70 14.2 3.342 4.856 3.342 4.856 ot_diis_step 99 11.5 0.011 0.011 4.198 4.199 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.104 4.190 apply_single 110 13.6 0.000 0.000 4.103 4.189 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.022 4.040 dbcsr_complete_redistribute 325 12.2 0.466 0.514 2.724 3.878 pw_transfer 1331 11.6 0.066 0.076 3.807 3.838 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.701 3.735 density_rs2pw 110 9.6 0.004 0.005 3.111 3.630 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.531 3.575 grid_integrate_task_list 110 12.3 3.272 3.481 3.272 3.481 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.454 3.455 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.387 3.439 calculate_first_density_matrix 1 7.0 0.000 0.003 3.394 3.397 calculate_dm_sparse 110 9.5 0.001 0.001 3.330 3.378 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.192 3.333 fft_wrap_pw1pw2_140 451 13.1 0.671 0.694 3.156 3.196 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.394 3.128 hybrid_alltoall_any 4261 16.3 0.123 0.462 2.321 3.096 make_images_data 4110 15.4 0.048 0.052 2.728 3.087 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.077 3.079 wfi_extrapolate 11 7.9 0.001 0.001 3.068 3.068 rs_pw_transfer 902 11.9 0.010 0.011 2.516 3.041 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.783 2.909 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.897 2.904 cp_fm_diag_elpa_base 48 14.0 2.748 2.806 2.894 2.902 mp_alltoall_i22 605 13.7 1.668 2.839 1.668 2.839 cp_fm_cholesky_invert 11 10.9 2.732 2.742 2.732 2.742 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.662 2.695 acc_transpose_blocks 24660 15.4 0.111 0.115 2.434 2.550 multiply_cannon_sync_h2d 24660 15.4 2.381 2.537 2.381 2.537 grid_collocate_task_list 110 9.6 2.223 2.484 2.223 2.484 fft3d_ps 1111 14.6 1.084 1.115 2.460 2.476 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 2.283 2.286 potential_pw2rs 110 12.3 0.013 0.013 2.253 2.263 mp_alltoall_d11v 2046 13.8 1.771 2.040 1.771 2.040 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.875 1.893 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.821 1.860 cp_fm_cholesky_decompose 22 10.9 1.728 1.791 1.728 1.791 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.710 1.723 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.002 1.612 1.706 jit_kernel_multiply 8 15.9 0.788 1.668 0.788 1.668 mp_sum_l 6594 12.7 1.013 1.664 1.013 1.664 mp_waitany 13376 13.8 1.137 1.638 1.137 1.638 mp_allgather_i34 2055 14.4 0.546 1.618 0.546 1.618 acc_transpose_blocks_sync 73980 16.4 1.389 1.512 1.389 1.512 multiply_cannon_metrocomm4 20550 15.4 0.061 0.065 0.859 1.440 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=68.847000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=701.000000, yerr=9.486833 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_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 858.279936E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.034 57.371 57.372 qs_mol_dyn_low 1 2.0 0.003 0.003 57.050 57.065 qs_forces 11 3.9 0.021 0.035 56.888 56.889 qs_energies 11 4.9 0.001 0.001 53.144 53.164 scf_env_do_scf 11 5.9 0.000 0.001 43.871 43.871 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 36.081 36.081 velocity_verlet 10 3.0 0.020 0.042 31.700 31.707 dbcsr_multiply_generic 2055 12.4 0.104 0.109 23.889 24.004 qs_scf_new_mos 99 7.5 0.001 0.001 20.919 20.989 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.919 20.988 ot_scf_mini 99 9.5 0.002 0.002 19.678 19.740 multiply_cannon 2055 13.4 0.235 0.250 18.293 19.619 multiply_cannon_loop 2055 14.4 0.601 0.631 17.047 17.310 rebuild_ks_matrix 110 8.3 0.000 0.000 11.961 12.009 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.961 12.009 ot_mini 99 10.5 0.001 0.001 10.817 10.881 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.611 10.656 multiply_cannon_multrec 8220 15.4 3.306 4.682 8.267 9.419 mp_waitall_1 103326 16.6 5.846 7.777 5.846 7.777 init_scf_loop 11 6.9 0.000 0.000 7.744 7.745 qs_ot_get_derivative 99 11.5 0.001 0.001 7.062 7.128 sum_up_and_integrate 110 10.3 0.079 0.080 6.554 6.566 integrate_v_rspace 110 11.3 0.003 0.003 6.474 6.488 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.366 6.377 calculate_rho_elec 110 8.6 0.115 0.116 6.365 6.376 init_scf_run 11 5.9 0.000 0.001 6.118 6.118 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.118 6.118 prepare_preconditioner 11 7.9 0.000 0.000 6.080 6.085 make_preconditioner 11 8.9 0.000 0.000 6.080 6.085 dbcsr_mm_accdrv_process 17442 15.9 3.457 4.440 4.822 5.951 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.679 5.749 qs_ot_get_p 110 10.4 0.001 0.001 4.890 4.968 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 3.050 4.538 make_m2s 4110 13.4 0.038 0.040 4.230 4.482 make_images 4110 14.4 0.661 0.721 4.102 4.351 pw_transfer 1331 11.6 0.066 0.070 3.945 3.952 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.838 3.849 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.788 3.817 apply_single 110 13.6 0.000 0.000 3.788 3.817 ot_diis_step 99 11.5 0.012 0.012 3.732 3.732 grid_integrate_task_list 110 12.3 3.356 3.555 3.356 3.555 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.508 3.512 density_rs2pw 110 9.6 0.004 0.004 3.058 3.433 fft_wrap_pw1pw2_140 451 13.1 0.835 0.847 3.297 3.312 calculate_first_density_matrix 1 7.0 0.000 0.000 3.292 3.293 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.167 3.168 cp_fm_cholesky_invert 11 10.9 2.947 2.952 2.947 2.952 calculate_dm_sparse 110 9.5 0.001 0.001 2.888 2.931 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.776 2.777 wfi_extrapolate 11 7.9 0.001 0.001 2.724 2.724 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.687 2.705 hybrid_alltoall_any 4261 16.3 0.201 0.856 2.263 2.666 rs_pw_transfer 902 11.9 0.010 0.010 2.213 2.608 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.580 2.588 cp_fm_diag_elpa_base 48 14.0 2.522 2.545 2.578 2.586 grid_collocate_task_list 110 9.6 2.315 2.570 2.315 2.570 make_images_data 4110 15.4 0.041 0.047 2.244 2.556 multiply_cannon_sync_h2d 8220 15.4 2.343 2.504 2.343 2.504 fft3d_ps 1111 14.6 1.131 1.138 2.381 2.395 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.316 2.335 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.035 2.081 potential_pw2rs 110 12.3 0.016 0.016 2.077 2.081 mp_alltoall_d11v 2046 13.8 1.729 2.007 1.729 2.007 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.785 2.005 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.940 1.976 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.850 1.865 cp_fm_cholesky_decompose 22 10.9 1.706 1.726 1.706 1.726 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.683 1.686 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.505 1.648 dbcsr_complete_redistribute 325 12.2 0.659 0.687 1.529 1.616 mp_waitany 9240 13.8 1.078 1.506 1.078 1.506 mp_allgather_i34 2055 14.4 0.423 1.489 0.423 1.489 acc_transpose_blocks 8220 15.4 0.040 0.041 1.378 1.425 multiply_cannon_metrocomm4 6165 15.4 0.018 0.020 0.478 1.353 qs_create_task_list 11 7.9 0.000 0.001 1.214 1.318 generate_qs_task_list 11 8.9 0.376 0.445 1.214 1.318 mp_irecv_dv 24056 15.7 0.452 1.306 0.452 1.306 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.229 1.264 jit_kernel_multiply 6 15.7 1.053 1.249 1.053 1.249 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.164 0.819 1.224 rs_gather_matrices 110 12.3 0.324 0.366 0.952 1.203 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.372000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=810.636364, yerr=10.890135 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_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 1.405952E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.048 90.263 90.264 qs_mol_dyn_low 1 2.0 0.003 0.003 89.952 89.961 qs_forces 11 3.9 0.003 0.003 89.561 89.562 qs_energies 11 4.9 0.001 0.002 85.317 85.319 scf_env_do_scf 11 5.9 0.001 0.001 74.633 74.633 velocity_verlet 10 3.0 0.002 0.002 57.080 57.127 scf_env_do_scf_inner_loop 99 6.5 0.010 0.014 45.879 45.880 dbcsr_multiply_generic 2055 12.4 0.119 0.124 30.393 30.534 init_scf_loop 11 6.9 0.000 0.000 28.684 28.685 qs_scf_new_mos 99 7.5 0.001 0.001 27.329 27.412 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.328 27.411 prepare_preconditioner 11 7.9 0.000 0.000 26.530 26.546 make_preconditioner 11 8.9 0.000 0.000 26.530 26.546 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.656 26.005 ot_scf_mini 99 9.5 0.002 0.002 25.532 25.616 multiply_cannon 2055 13.4 0.335 0.365 23.160 23.840 multiply_cannon_loop 2055 14.4 0.839 0.853 21.333 21.752 cp_fm_upper_to_full 70 14.2 12.845 18.447 12.845 18.447 ot_mini 99 10.5 0.001 0.001 14.565 14.643 rebuild_ks_matrix 110 8.3 0.000 0.001 14.341 14.439 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.341 14.438 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.974 13.061 dbcsr_complete_redistribute 325 12.2 1.022 1.055 7.427 10.788 multiply_cannon_multrec 8220 15.4 4.052 4.218 10.170 10.478 qs_ot_get_derivative 99 11.5 0.001 0.001 9.758 9.838 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.424 9.780 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.858 9.177 mp_waitall_1 84994 16.7 7.852 8.898 7.852 8.898 mp_alltoall_i22 605 13.7 5.500 8.890 5.500 8.890 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.839 7.877 calculate_rho_elec 110 8.6 0.225 0.225 7.838 7.877 sum_up_and_integrate 110 10.3 0.150 0.151 7.442 7.455 integrate_v_rspace 110 11.3 0.004 0.004 7.292 7.306 init_scf_run 11 5.9 0.000 0.001 6.392 6.392 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.391 6.392 dbcsr_mm_accdrv_process 11614 15.7 4.137 5.213 5.970 6.385 make_m2s 4110 13.4 0.043 0.043 5.407 5.946 qs_ot_get_p 110 10.4 0.001 0.001 5.747 5.857 make_images 4110 14.4 0.896 0.955 5.221 5.759 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 5.068 5.510 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.902 5.416 apply_single 110 13.6 0.000 0.000 4.901 5.416 cp_fm_cholesky_invert 11 10.9 5.338 5.342 5.338 5.342 pw_transfer 1331 11.6 0.075 0.076 5.064 5.072 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 4.946 4.954 ot_diis_step 99 11.5 0.015 0.016 4.789 4.789 fft_wrap_pw1pw2_140 451 13.1 1.336 1.344 4.318 4.328 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.044 4.052 grid_integrate_task_list 110 12.3 3.743 3.842 3.743 3.842 density_rs2pw 110 9.6 0.004 0.004 3.763 3.823 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.277 3.748 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.688 3.689 calculate_dm_sparse 110 9.5 0.001 0.001 3.588 3.619 hybrid_alltoall_any 4261 16.3 0.263 0.563 2.843 3.597 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.528 3.529 make_images_data 4110 15.4 0.045 0.047 2.859 3.518 wfi_extrapolate 11 7.9 0.001 0.001 3.380 3.380 multiply_cannon_sync_h2d 8220 15.4 3.099 3.127 3.099 3.127 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.935 2.936 cp_fm_diag_elpa_base 48 14.0 2.384 2.583 2.933 2.933 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.896 2.899 calculate_first_density_matrix 1 7.0 0.000 0.000 2.890 2.891 fft3d_ps 1111 14.6 1.298 1.306 2.875 2.879 grid_collocate_task_list 110 9.6 2.724 2.794 2.724 2.794 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.670 2.712 potential_pw2rs 110 12.3 0.021 0.022 2.412 2.419 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.376 2.410 rs_pw_transfer 902 11.9 0.010 0.011 2.311 2.381 qs_env_update_s_mstruct 11 6.9 0.002 0.009 2.183 2.240 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.106 2.214 mp_alltoall_d11v 2046 13.8 2.006 2.158 2.006 2.158 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.088 2.103 jit_kernel_multiply 10 15.1 1.634 1.997 1.634 1.997 cp_fm_cholesky_decompose 22 10.9 1.967 1.989 1.967 1.989 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.967 1.978 qs_create_task_list 11 7.9 0.000 0.001 1.881 1.931 generate_qs_task_list 11 8.9 0.734 0.785 1.881 1.930 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=90.264000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1253.363636, yerr=57.017179 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 634.552320E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57761. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 1874318. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.032 206.798 206.799 qs_mol_dyn_low 1 2.0 0.003 0.003 206.349 206.365 qs_forces 11 3.9 0.005 0.005 206.264 206.265 qs_energies 11 4.9 0.002 0.002 200.631 200.655 scf_env_do_scf 11 5.9 0.001 0.001 183.788 183.793 scf_env_do_scf_inner_loop 117 6.6 0.003 0.007 163.438 163.440 dbcsr_multiply_generic 2507 12.6 0.178 0.181 125.119 125.515 qs_scf_new_mos 117 7.6 0.001 0.001 123.461 123.714 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.460 123.713 velocity_verlet 10 3.0 0.001 0.002 123.295 123.296 ot_scf_mini 117 9.6 0.003 0.003 116.806 117.023 multiply_cannon 2507 13.6 0.240 0.250 101.444 102.847 multiply_cannon_loop 2507 14.6 2.391 2.465 99.204 100.521 ot_mini 117 10.6 0.001 0.001 65.962 66.217 multiply_cannon_multrec 60168 15.6 31.980 34.127 41.935 43.499 qs_ot_get_derivative 117 11.6 0.001 0.001 41.359 41.589 rebuild_ks_matrix 128 8.3 0.001 0.001 33.575 33.900 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.016 33.575 33.899 mp_waitall_1 267128 16.5 28.250 30.978 28.250 30.978 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.109 30.419 qs_ot_get_p 128 10.4 0.001 0.001 28.780 29.119 multiply_cannon_sync_h2d 60168 15.6 26.440 27.953 26.440 27.953 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.207 24.812 apply_single 128 13.6 0.001 0.001 24.207 24.812 ot_diis_step 117 11.6 0.007 0.008 24.386 24.387 qs_ot_p2m_diag 83 11.4 0.079 0.092 22.286 22.351 init_scf_loop 11 6.9 0.000 0.001 20.276 20.278 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.626 19.809 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.493 19.494 multiply_cannon_metrocomm3 60168 15.6 0.118 0.126 15.743 17.497 cp_fm_diag_elpa 83 13.4 0.000 0.001 16.496 16.528 cp_fm_diag_elpa_base 83 14.4 16.430 16.458 16.492 16.525 prepare_preconditioner 11 7.9 0.000 0.000 15.716 15.769 make_preconditioner 11 8.9 0.000 0.000 15.716 15.769 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.950 15.124 make_m2s 5014 13.6 0.103 0.111 13.805 14.162 sum_up_and_integrate 128 10.3 0.089 0.108 14.133 14.149 integrate_v_rspace 128 11.3 0.004 0.004 14.043 14.061 make_images 5014 14.6 0.399 0.419 13.630 13.998 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.788 13.931 calculate_rho_elec 128 8.7 0.046 0.065 13.787 13.930 init_scf_run 11 5.9 0.000 0.001 12.678 12.679 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.678 12.678 density_rs2pw 128 9.7 0.006 0.007 7.043 10.431 dbcsr_mm_accdrv_process 124484 16.2 4.564 4.838 9.518 10.096 mp_sum_l 7950 12.9 8.600 9.991 8.600 9.991 wfi_extrapolate 11 7.9 0.001 0.001 9.114 9.114 rs_pw_transfer 1046 11.9 0.017 0.019 5.666 9.056 cp_fm_cholesky_invert 11 10.9 8.931 8.939 8.931 8.939 calculate_dm_sparse 128 9.5 0.001 0.001 8.477 8.577 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.189 8.359 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.028 8.130 pw_transfer 1547 11.6 0.075 0.089 7.721 7.891 make_images_data 5014 15.6 0.070 0.076 6.850 7.748 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 7.518 7.690 multiply_cannon_metrocomm1 60168 15.6 0.096 0.100 5.999 7.668 grid_integrate_task_list 128 12.3 7.114 7.531 7.114 7.531 hybrid_alltoall_any 5200 16.5 0.296 2.267 5.996 7.374 fft_wrap_pw1pw2_140 523 13.2 1.283 1.327 6.610 6.777 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.666 6.678 mp_waitany 16020 13.9 2.649 6.075 2.649 6.075 grid_collocate_task_list 128 9.7 4.664 5.975 4.664 5.975 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.786 5.870 fft3d_ps 1291 14.7 2.133 2.767 5.270 5.580 rs_pw_transfer_RS2PW_140 139 11.5 0.282 0.300 2.141 5.550 mp_alltoall_d11v 2415 14.1 4.173 5.348 4.173 5.348 cp_fm_cholesky_decompose 22 10.9 4.620 4.634 4.620 4.634 potential_pw2rs 128 12.3 0.009 0.011 4.515 4.539 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.799000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.818182, yerr=7.296552 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 839.155712E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 5193241. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.042 190.822 190.824 qs_mol_dyn_low 1 2.0 0.003 0.003 190.406 190.420 qs_forces 11 3.9 0.004 0.005 190.234 190.235 qs_energies 11 4.9 0.002 0.007 183.319 183.333 scf_env_do_scf 11 5.9 0.001 0.001 166.482 166.492 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 133.722 133.725 velocity_verlet 10 3.0 0.001 0.002 119.580 119.583 dbcsr_multiply_generic 2507 12.6 0.189 0.194 96.820 98.059 qs_scf_new_mos 117 7.6 0.001 0.001 94.002 94.485 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.001 94.484 ot_scf_mini 117 9.6 0.004 0.004 89.198 89.739 multiply_cannon 2507 13.6 0.501 0.554 76.759 81.755 multiply_cannon_loop 2507 14.6 1.565 1.633 73.845 76.372 ot_mini 117 10.6 0.001 0.001 50.070 50.717 mp_waitall_1 214728 16.6 24.026 37.536 24.026 37.536 multiply_cannon_multrec 30084 15.6 20.910 25.105 31.436 35.909 rebuild_ks_matrix 128 8.3 0.001 0.001 32.954 33.581 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 32.953 33.581 init_scf_loop 11 6.9 0.001 0.003 32.670 32.672 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.541 30.112 multiply_cannon_metrocomm3 30084 15.6 0.096 0.101 15.763 29.076 qs_ot_get_derivative 117 11.6 0.001 0.002 28.037 28.576 prepare_preconditioner 11 7.9 0.000 0.000 28.285 28.356 make_preconditioner 11 8.9 0.000 0.002 28.285 28.356 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.979 27.508 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.119 23.316 apply_single 128 13.6 0.001 0.001 22.118 23.315 qs_ot_get_p 128 10.4 0.001 0.001 21.314 21.981 ot_diis_step 117 11.6 0.014 0.015 21.858 21.860 multiply_cannon_sync_h2d 30084 15.6 17.901 20.084 17.901 20.084 cp_fm_cholesky_invert 11 10.9 16.553 16.566 16.553 16.566 qs_ot_p2m_diag 83 11.4 0.188 0.216 16.412 16.447 make_m2s 5014 13.6 0.092 0.098 14.130 15.336 cp_dbcsr_syevd 83 12.4 0.006 0.006 15.205 15.205 make_images 5014 14.6 1.199 1.402 13.921 15.125 sum_up_and_integrate 128 10.3 0.116 0.133 14.799 14.839 integrate_v_rspace 128 11.3 0.004 0.004 14.682 14.726 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.152 14.184 calculate_rho_elec 128 8.7 0.089 0.106 14.151 14.184 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.053 12.088 cp_fm_diag_elpa_base 83 14.4 11.797 11.900 12.046 12.076 init_scf_run 11 5.9 0.000 0.001 11.996 11.998 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.996 11.997 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.093 11.510 multiply_cannon_metrocomm4 27577 15.6 0.108 0.120 3.899 10.809 dbcsr_mm_accdrv_process 62242 16.2 5.303 6.104 9.977 10.522 mp_irecv_dv 69486 16.3 3.691 10.395 3.691 10.395 density_rs2pw 128 9.7 0.007 0.007 7.453 10.056 make_images_data 5014 15.6 0.069 0.080 8.279 9.787 hybrid_alltoall_any 5200 16.5 0.355 1.569 7.003 8.994 pw_transfer 1547 11.6 0.087 0.099 8.843 8.918 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 8.617 8.695 wfi_extrapolate 11 7.9 0.001 0.001 8.343 8.343 rs_pw_transfer 1046 11.9 0.015 0.017 5.724 8.339 fft_wrap_pw1pw2_140 523 13.2 1.337 1.369 7.629 7.728 grid_integrate_task_list 128 12.3 7.189 7.570 7.189 7.570 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.259 7.001 cp_fm_cholesky_decompose 22 10.9 6.878 6.961 6.878 6.961 calculate_dm_sparse 128 9.5 0.001 0.001 6.466 6.612 mp_sum_l 7950 12.9 4.139 6.438 4.139 6.438 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.097 6.107 fft3d_ps 1291 14.7 2.808 3.002 5.921 5.991 grid_collocate_task_list 128 9.7 4.780 5.987 4.780 5.987 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.383 5.456 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.203 5.388 mp_waitany 11748 13.9 2.509 5.189 2.509 5.189 potential_pw2rs 128 12.3 0.015 0.018 5.021 5.055 mp_alltoall_d11v 2415 14.1 4.122 4.743 4.122 4.743 rs_pw_transfer_RS2PW_140 139 11.5 0.351 0.386 2.114 4.711 mp_allgather_i34 2507 14.6 1.433 4.388 1.433 4.388 mp_sum_d 4465 12.1 2.682 4.100 2.682 4.100 dbcsr_complete_redistribute 395 12.7 0.778 0.855 3.103 3.929 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.805 3.862 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.824000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=800.727273, yerr=0.445362 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/17/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 2.928533E+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 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 950.263808E+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 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 58094. MP_Allreduce 11102 999. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.051 0.146 177.822 177.823 qs_mol_dyn_low 1 2.0 0.003 0.003 176.876 176.934 qs_forces 11 3.9 0.006 0.011 176.747 176.757 qs_energies 11 4.9 0.060 0.239 170.045 170.059 scf_env_do_scf 11 5.9 0.001 0.003 153.870 153.871 scf_env_do_scf_inner_loop 117 6.6 0.042 0.319 119.007 119.008 velocity_verlet 10 3.0 0.002 0.002 113.284 113.287 dbcsr_multiply_generic 2507 12.6 0.179 0.184 82.261 83.507 qs_scf_new_mos 117 7.6 0.001 0.001 80.130 80.450 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.129 80.449 ot_scf_mini 117 9.6 0.003 0.004 75.935 76.334 multiply_cannon 2507 13.6 0.505 0.528 62.492 67.847 multiply_cannon_loop 2507 14.6 1.124 1.184 59.588 62.517 ot_mini 117 10.6 0.001 0.001 42.709 43.114 init_scf_loop 11 6.9 0.001 0.003 34.765 34.766 mp_waitall_1 170520 16.6 24.663 34.705 24.663 34.705 rebuild_ks_matrix 128 8.3 0.001 0.001 30.949 31.423 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 30.948 31.422 prepare_preconditioner 11 7.9 0.000 0.001 30.721 30.772 make_preconditioner 11 8.9 0.000 0.001 30.721 30.772 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.315 29.772 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.820 28.249 multiply_cannon_multrec 20056 15.6 13.195 16.116 22.845 25.812 multiply_cannon_metrocomm3 20056 15.6 0.061 0.065 15.212 24.981 qs_ot_get_derivative 117 11.6 0.001 0.002 23.167 23.565 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.623 20.664 apply_single 128 13.6 0.001 0.001 19.623 20.664 ot_diis_step 117 11.6 0.018 0.018 19.433 19.433 qs_ot_get_p 128 10.4 0.001 0.001 18.643 19.105 make_m2s 5014 13.6 0.081 0.089 14.635 15.838 make_images 5014 14.6 1.167 1.263 14.403 15.605 multiply_cannon_sync_h2d 20056 15.6 13.580 15.235 13.580 15.235 sum_up_and_integrate 128 10.3 0.134 0.146 14.672 14.700 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.555 14.579 calculate_rho_elec 128 8.7 0.132 0.147 14.554 14.579 integrate_v_rspace 128 11.3 0.004 0.005 14.538 14.563 qs_ot_p2m_diag 83 11.4 0.265 0.272 14.272 14.281 cp_fm_cholesky_invert 11 10.9 14.202 14.211 14.202 14.211 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.229 13.230 init_scf_run 11 5.9 0.000 0.001 10.882 10.882 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.882 10.882 make_images_data 5014 15.6 0.063 0.073 8.769 10.310 cp_fm_diag_elpa 83 13.4 0.001 0.001 10.073 10.092 cp_fm_diag_elpa_base 83 14.4 9.656 9.810 10.069 10.087 hybrid_alltoall_any 5200 16.5 0.451 2.057 7.531 9.909 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.151 9.452 multiply_cannon_metrocomm4 17549 15.6 0.065 0.075 3.457 9.291 dbcsr_mm_accdrv_process 41502 16.2 5.967 6.283 9.112 9.248 density_rs2pw 128 9.7 0.006 0.007 7.192 9.211 mp_irecv_dv 50230 16.2 3.330 9.031 3.330 9.031 pw_transfer 1547 11.6 0.086 0.103 8.895 8.995 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 8.671 8.778 fft_wrap_pw1pw2_140 523 13.2 1.416 1.463 7.703 7.823 grid_integrate_task_list 128 12.3 7.266 7.767 7.266 7.767 wfi_extrapolate 11 7.9 0.001 0.001 7.475 7.475 cp_fm_upper_to_full 105 14.8 5.775 7.370 5.775 7.370 cp_fm_cholesky_decompose 22 10.9 7.289 7.314 7.289 7.314 rs_pw_transfer 1046 11.9 0.013 0.014 5.152 7.174 dbcsr_complete_redistribute 395 12.7 1.177 1.206 4.615 6.443 calculate_dm_sparse 128 9.5 0.001 0.001 6.168 6.293 fft3d_ps 1291 14.7 2.744 2.979 5.786 5.873 grid_collocate_task_list 128 9.7 4.989 5.777 4.989 5.777 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.429 5.434 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.619 5.338 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.465 5.286 mp_alltoall_d11v 2415 14.1 4.473 5.212 4.473 5.212 mp_sum_l 7950 12.9 3.356 4.966 3.356 4.966 potential_pw2rs 128 12.3 0.021 0.023 4.702 4.715 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.539 4.678 mp_waitany 11748 13.9 2.343 4.419 2.343 4.419 mp_allgather_i34 2507 14.6 1.400 4.339 1.400 4.339 transfer_fm_to_dbcsr 11 9.9 0.018 0.023 2.386 4.163 mp_alltoall_i22 716 14.1 2.042 4.044 2.042 4.044 rs_pw_transfer_RS2PW_140 139 11.5 0.332 0.354 1.924 3.922 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.881 3.907 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.776 3.785 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.626 3.676 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.823000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=901.636364, yerr=10.915150 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 4.387043E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6024576 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.8 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.166684E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1152312 MPI messages size (bytes): total size 2.039354E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.769794E+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 321600 36333158400 131072 < size <= 4194304 721560 792041881600 4194304 < size <= 16777216 70800 669921260160 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57894. MP_Allreduce 11137 1080. MP_Sync 87 MP_Alltoall 1724 12509439. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 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 0.024 0.082 192.435 192.437 qs_mol_dyn_low 1 2.0 0.003 0.003 191.860 191.874 qs_forces 11 3.9 0.044 0.062 191.405 191.416 qs_energies 11 4.9 0.002 0.004 184.089 184.141 scf_env_do_scf 11 5.9 0.001 0.001 166.694 166.701 velocity_verlet 10 3.0 0.005 0.031 127.029 127.045 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 120.558 120.563 dbcsr_multiply_generic 2527 12.6 0.187 0.193 82.343 83.123 qs_scf_new_mos 118 7.6 0.001 0.001 82.049 82.371 qs_scf_loop_do_ot 118 8.6 0.001 0.001 82.048 82.370 ot_scf_mini 118 9.6 0.003 0.004 77.345 77.694 multiply_cannon 2527 13.6 0.561 0.596 57.267 60.265 multiply_cannon_loop 2527 14.6 1.839 1.946 53.085 55.061 init_scf_loop 11 6.9 0.001 0.006 46.012 46.022 ot_mini 118 10.6 0.001 0.001 43.909 44.252 prepare_preconditioner 11 7.9 0.000 0.000 41.869 41.900 make_preconditioner 11 8.9 0.000 0.001 41.869 41.900 make_full_inverse_cholesky 11 9.9 0.011 0.020 35.530 40.448 multiply_cannon_multrec 30324 15.6 13.607 19.171 27.211 32.507 rebuild_ks_matrix 129 8.3 0.001 0.001 30.332 30.656 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 30.331 30.656 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.305 27.593 mp_waitall_1 149060 16.7 17.450 26.656 17.450 26.656 qs_ot_get_derivative 118 11.6 0.002 0.002 23.887 24.241 make_m2s 5054 13.6 0.097 0.103 20.671 21.813 make_images 5054 14.6 1.993 2.315 20.365 21.509 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.404 20.013 apply_single 129 13.6 0.001 0.001 19.404 20.012 ot_diis_step 118 11.6 0.018 0.019 19.896 19.897 qs_ot_get_p 129 10.4 0.001 0.001 19.338 19.687 cp_fm_upper_to_full 105 14.8 11.321 16.631 11.321 16.631 cp_fm_cholesky_invert 11 10.9 15.987 15.996 15.987 15.996 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.267 15.302 calculate_rho_elec 129 8.7 0.177 0.194 15.267 15.302 multiply_cannon_metrocomm3 30324 15.6 0.050 0.053 6.510 15.155 qs_ot_p2m_diag 83 11.4 0.342 0.390 15.036 15.090 sum_up_and_integrate 129 10.3 0.141 0.153 14.891 14.914 integrate_v_rspace 129 11.3 0.004 0.004 14.750 14.778 dbcsr_mm_accdrv_process 62756 16.2 8.982 9.933 13.166 13.690 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.630 13.631 make_images_data 5054 15.6 0.067 0.073 11.143 13.358 dbcsr_complete_redistribute 395 12.7 1.506 1.620 9.072 12.778 hybrid_alltoall_any 5240 16.5 0.535 2.219 9.909 12.104 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.699 11.434 multiply_cannon_sync_h2d 30324 15.6 10.579 11.325 10.579 11.325 init_scf_run 11 5.9 0.000 0.001 11.294 11.295 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.294 11.295 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.416 10.430 cp_fm_diag_elpa_base 83 14.4 9.447 9.755 10.409 10.422 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.953 10.211 transfer_fm_to_dbcsr 11 9.9 0.001 0.004 6.317 9.976 pw_transfer 1559 11.6 0.087 0.101 9.217 9.297 mp_alltoall_i22 718 14.1 5.548 9.220 5.548 9.220 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.989 9.077 density_rs2pw 129 9.7 0.006 0.006 7.066 8.544 fft_wrap_pw1pw2_140 527 13.2 1.570 1.604 7.997 8.099 grid_integrate_task_list 129 12.3 7.574 7.919 7.574 7.919 wfi_extrapolate 11 7.9 0.001 0.001 7.816 7.816 cp_fm_cholesky_decompose 22 10.9 7.438 7.535 7.438 7.535 multiply_cannon_metrocomm4 25270 15.6 0.084 0.095 2.896 7.145 mp_irecv_dv 76685 16.2 2.738 6.869 2.738 6.869 calculate_dm_sparse 129 9.5 0.001 0.001 6.784 6.850 rs_pw_transfer 1054 12.0 0.013 0.015 4.910 6.475 mp_alltoall_d11v 2423 14.1 5.636 6.338 5.636 6.338 fft3d_ps 1301 14.7 2.861 2.967 5.890 5.945 grid_collocate_task_list 129 9.7 5.189 5.929 5.189 5.929 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.514 5.559 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.596 4.697 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 4.625 4.671 potential_pw2rs 129 12.3 0.024 0.024 4.596 4.606 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.413 4.486 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.362 4.427 mp_allgather_i34 2527 14.6 2.246 4.114 2.246 4.114 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=192.437000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1085.636364, yerr=19.606459 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 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 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 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 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 5.910120E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.525314E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526746E+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 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592214928 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57624. MP_Allreduce 11184 1163. MP_Sync 88 MP_Alltoall 1724 18848066. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.044 173.899 173.900 qs_mol_dyn_low 1 2.0 0.003 0.003 173.349 173.487 qs_forces 11 3.9 0.004 0.004 173.228 173.242 qs_energies 11 4.9 0.001 0.002 165.627 165.637 scf_env_do_scf 11 5.9 0.001 0.001 147.651 147.655 velocity_verlet 10 3.0 0.002 0.002 113.765 113.768 scf_env_do_scf_inner_loop 118 6.6 0.012 0.025 112.019 112.020 dbcsr_multiply_generic 2529 12.6 0.181 0.193 73.331 73.745 qs_scf_new_mos 118 7.6 0.001 0.001 72.894 73.006 qs_scf_loop_do_ot 118 8.6 0.001 0.001 72.893 73.005 ot_scf_mini 118 9.6 0.003 0.004 68.413 68.519 multiply_cannon 2529 13.6 0.566 0.588 54.630 57.737 multiply_cannon_loop 2529 14.6 0.820 0.846 51.038 52.022 ot_mini 118 10.6 0.001 0.001 38.111 38.211 init_scf_loop 11 6.9 0.000 0.000 35.487 35.489 prepare_preconditioner 11 7.9 0.000 0.000 31.606 31.636 make_preconditioner 11 8.9 0.000 0.000 31.605 31.636 mp_waitall_1 126876 16.7 24.971 31.349 24.971 31.349 rebuild_ks_matrix 129 8.3 0.001 0.001 30.171 30.313 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 30.170 30.312 make_full_inverse_cholesky 11 9.9 0.014 0.028 29.518 29.784 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.394 27.525 multiply_cannon_multrec 10116 15.6 10.459 16.261 18.145 22.521 qs_ot_get_derivative 118 11.6 0.002 0.002 20.692 20.789 multiply_cannon_metrocomm3 10116 15.6 0.025 0.027 13.020 20.549 cp_fm_cholesky_invert 11 10.9 18.127 18.133 18.127 18.133 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.444 17.625 apply_single 129 13.6 0.001 0.001 17.444 17.625 qs_ot_get_p 129 10.4 0.001 0.001 17.446 17.547 ot_diis_step 118 11.6 0.020 0.020 17.347 17.349 make_m2s 5058 13.6 0.066 0.070 14.808 15.798 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.624 15.643 calculate_rho_elec 129 8.7 0.261 0.272 15.624 15.643 make_images 5058 14.6 2.231 2.664 14.503 15.488 sum_up_and_integrate 129 10.3 0.181 0.191 15.156 15.200 integrate_v_rspace 129 11.3 0.004 0.004 14.975 15.027 qs_ot_p2m_diag 84 11.4 0.502 0.507 13.672 13.688 cp_dbcsr_syevd 84 12.4 0.005 0.006 12.493 12.495 multiply_cannon_sync_h2d 10116 15.6 10.806 11.262 10.806 11.262 init_scf_run 11 5.9 0.000 0.001 11.155 11.156 scf_env_initial_rho_setup 11 6.9 0.018 0.021 11.155 11.155 make_images_data 5058 15.6 0.057 0.066 8.437 9.880 pw_transfer 1559 11.6 0.087 0.096 9.625 9.644 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.574 9.587 cp_fm_diag_elpa_base 84 14.4 9.333 9.415 9.571 9.583 hybrid_alltoall_any 5245 16.5 0.851 3.816 8.279 9.473 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 9.399 9.422 fft_wrap_pw1pw2_140 527 13.2 1.936 1.990 8.329 8.355 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 8.157 8.234 grid_integrate_task_list 129 12.3 7.799 8.228 7.799 8.228 dbcsr_mm_accdrv_process 20934 16.1 3.332 4.406 7.336 8.071 cp_fm_cholesky_decompose 22 10.9 7.862 7.983 7.862 7.983 density_rs2pw 129 9.7 0.006 0.006 7.022 7.954 wfi_extrapolate 11 7.9 0.001 0.001 7.471 7.471 multiply_cannon_metrocomm1 10116 15.6 0.030 0.031 4.286 7.378 calculate_dm_sparse 129 9.5 0.001 0.001 6.161 6.249 grid_collocate_task_list 129 9.7 5.517 6.210 5.517 6.210 mp_alltoall_d11v 2429 14.1 4.995 5.990 4.995 5.990 fft3d_ps 1301 14.7 2.791 2.866 5.836 5.877 multiply_cannon_metrocomm4 7587 15.6 0.026 0.029 1.877 5.854 mp_irecv_dv 29102 15.9 1.838 5.758 1.838 5.758 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.695 5.744 dbcsr_complete_redistribute 397 12.7 2.146 2.210 5.207 5.560 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.298 5.305 rs_pw_transfer 1054 12.0 0.013 0.013 4.207 5.151 mp_allgather_i34 2529 14.6 1.333 4.608 1.333 4.608 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.562 4.608 potential_pw2rs 129 12.3 0.027 0.028 4.475 4.488 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.571 3.854 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.693 3.736 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.402 3.720 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.612 3.707 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.685 3.697 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.409 3.500 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=173.900000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1445.181818, yerr=14.868014 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430451062784 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986258751488 0.0% 0.0% 100.0% flops 32 x 32 x 32 1986422374400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1991997444096 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 11613050155008 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239170400256 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239170400256 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 94.256601E+12 0.0% 0.0% 100.0% flops max/rank 11.789007E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806737984 0.0% 0.0% 100.0% number of processed stacks 1984336 0.0% 0.0% 100.0% average stack size 0.0 0.0 3430.2 marketing flops 145.674553E+12 ------------------------------------------------------------------------------- # multiplications 2540 max memory usage/rank 3.166892E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101600 MPI messages size (bytes): total size 1.145338E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.273011E+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 46088 35802578944 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592076688 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4035 58670. MP_Allreduce 11164 1500. MP_Sync 88 MP_Alltoall 1724 36993736. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.054 296.835 296.838 qs_mol_dyn_low 1 2.0 0.003 0.003 296.264 296.301 qs_forces 11 3.9 0.005 0.005 296.151 296.164 qs_energies 11 4.9 0.002 0.002 286.828 286.848 scf_env_do_scf 11 5.9 0.001 0.001 263.782 263.794 velocity_verlet 10 3.0 0.005 0.005 214.077 214.086 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 140.190 140.195 init_scf_loop 11 6.9 0.000 0.000 123.313 123.322 prepare_preconditioner 11 7.9 0.000 0.000 118.078 118.099 make_preconditioner 11 8.9 0.000 0.000 118.078 118.099 make_full_inverse_cholesky 11 9.9 0.038 0.039 94.411 115.220 qs_scf_new_mos 118 7.6 0.001 0.001 89.986 90.065 qs_scf_loop_do_ot 118 8.6 0.001 0.001 89.986 90.064 ot_scf_mini 118 9.6 0.004 0.004 85.088 85.111 dbcsr_multiply_generic 2540 12.6 0.212 0.223 82.688 83.113 cp_fm_upper_to_full 106 14.8 52.367 75.000 52.367 75.000 multiply_cannon 2540 13.6 0.688 0.740 59.000 60.280 multiply_cannon_loop 2540 14.6 1.042 1.079 54.930 56.273 ot_mini 118 10.6 0.001 0.001 44.955 44.990 dbcsr_complete_redistribute 397 12.7 4.033 4.078 29.662 42.795 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.137 39.237 rebuild_ks_matrix 129 8.3 0.001 0.001 37.838 37.876 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 37.837 37.876 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 23.624 36.482 qs_ks_update_qs_env 129 7.6 0.001 0.001 34.775 34.819 mp_alltoall_i22 720 14.1 21.377 34.587 21.377 34.587 cp_fm_cholesky_invert 11 10.9 33.086 33.092 33.086 33.092 mp_waitall_1 105020 16.8 28.177 32.381 28.177 32.381 qs_ot_get_p 129 10.4 0.001 0.001 25.326 25.400 qs_ot_get_derivative 118 11.6 0.002 0.002 24.812 24.839 qs_ot_p2m_diag 84 11.4 0.891 0.896 21.197 21.226 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.581 20.590 calculate_rho_elec 129 8.7 0.487 0.488 20.580 20.589 make_m2s 5080 13.6 0.075 0.077 18.984 20.471 ot_diis_step 118 11.6 0.022 0.022 20.110 20.111 multiply_cannon_metrocomm3 10160 15.6 0.025 0.028 18.657 20.003 make_images 5080 14.6 3.117 3.315 18.504 19.993 sum_up_and_integrate 129 10.3 0.324 0.327 19.616 19.700 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.544 19.682 apply_single 129 13.6 0.001 0.001 19.543 19.681 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.393 19.394 integrate_v_rspace 129 11.3 0.004 0.004 19.291 19.373 multiply_cannon_multrec 10160 15.6 10.416 12.262 18.245 18.294 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.211 16.211 cp_fm_diag_elpa_base 84 14.4 11.780 13.429 16.207 16.208 multiply_cannon_sync_h2d 10160 15.6 14.444 14.483 14.444 14.483 hybrid_alltoall_any 5267 16.5 1.321 3.085 11.028 13.178 make_images_data 5080 15.6 0.065 0.070 10.803 12.979 init_scf_run 11 5.9 0.000 0.001 12.334 12.335 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.334 12.335 pw_transfer 1559 11.6 0.093 0.094 11.529 11.540 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 11.293 11.304 fft_wrap_pw1pw2_140 527 13.2 3.142 3.213 10.073 10.091 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.733 9.754 dbcsr_mm_accdrv_process 20978 16.1 4.314 6.214 7.578 9.472 wfi_extrapolate 11 7.9 0.001 0.001 9.193 9.193 grid_integrate_task_list 129 12.3 8.643 8.835 8.643 8.835 cp_fm_cholesky_decompose 22 10.9 8.820 8.834 8.820 8.834 mp_alltoall_d11v 2429 14.1 7.962 8.659 7.962 8.659 qs_energies_init_hamiltonians 11 5.9 0.277 0.280 8.479 8.490 density_rs2pw 129 9.7 0.006 0.006 7.994 8.158 calculate_dm_sparse 129 9.5 0.001 0.001 6.757 6.828 grid_collocate_task_list 129 9.7 6.478 6.548 6.478 6.548 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.331 6.396 fft3d_ps 1301 14.7 2.872 2.882 6.288 6.366 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.201 6.306 rs_scatter_matrices 140 9.7 3.982 4.869 6.086 6.302 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=296.838000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2878.545455, yerr=154.323370 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_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 1.259663E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 43298903. 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 85.497 85.498 qs_energies 1 2.0 0.000 0.000 84.967 84.989 ls_scf 1 3.0 0.000 0.000 84.069 84.091 dbcsr_multiply_generic 111 6.7 0.016 0.017 72.895 73.074 multiply_cannon 111 7.7 0.017 0.020 56.250 57.395 multiply_cannon_loop 111 8.7 0.216 0.239 52.771 53.910 ls_scf_main 1 4.0 0.000 0.000 52.494 52.495 density_matrix_trs4 2 5.0 0.002 0.003 46.895 46.992 ls_scf_init_scf 1 4.0 0.000 0.000 28.530 28.531 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.376 27.438 mp_waitall_1 11031 10.9 22.421 25.525 22.421 25.525 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.245 25.271 multiply_cannon_multrec 2664 9.7 8.200 8.994 15.699 17.635 multiply_cannon_sync_h2d 2664 9.7 13.687 15.674 13.687 15.674 make_m2s 222 7.7 0.010 0.012 13.063 13.508 make_images 222 8.7 0.101 0.112 13.041 13.489 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.582 12.900 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.575 9.306 dbcsr_mm_accdrv_process 4760 10.4 0.578 0.688 7.115 8.239 make_images_data 222 9.7 0.005 0.006 7.596 8.179 hybrid_alltoall_any 227 10.6 0.220 1.864 6.535 7.935 dbcsr_mm_accdrv_process_sort 4760 11.4 6.313 7.219 6.313 7.219 calculate_norms 4752 9.8 5.534 6.209 5.534 6.209 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.004 5.180 mp_sum_l 887 5.1 3.035 4.430 3.035 4.430 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.064 3.834 mp_irecv_dv 6231 10.9 2.047 3.801 2.047 3.801 make_images_sizes 222 9.7 0.000 0.000 0.745 3.597 mp_alltoall_i44 222 10.7 0.744 3.597 0.744 3.597 arnoldi_extremal 4 6.8 0.000 0.000 3.274 3.303 arnoldi_normal_ev 4 7.8 0.001 0.003 3.274 3.303 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.253 3.259 build_subspace 16 8.4 0.009 0.012 3.176 3.179 ls_scf_post 1 4.0 0.000 0.000 3.045 3.066 ls_scf_store_result 1 5.0 0.000 0.000 2.847 2.896 dbcsr_special_finalize 555 9.7 0.006 0.007 2.357 2.872 dbcsr_merge_single_wm 555 10.7 0.476 0.590 2.348 2.861 make_images_pack 222 9.7 2.214 2.636 2.216 2.638 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.313 2.552 dbcsr_sort_data 658 11.4 2.129 2.543 2.129 2.543 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.377 2.500 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.459 2.069 2.460 buffer_matrices_ensure_size 222 8.7 1.813 2.158 1.813 2.158 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.796 1.797 rebuild_ks_matrix 3 7.3 0.000 0.000 1.786 1.788 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.786 1.788 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.498000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_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 2.145206E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.075 90.417 90.418 qs_energies 1 2.0 0.000 0.000 89.887 89.891 ls_scf 1 3.0 0.000 0.000 88.323 88.326 dbcsr_multiply_generic 111 6.7 0.016 0.017 74.250 74.653 multiply_cannon 111 7.7 0.027 0.040 52.550 57.537 ls_scf_main 1 4.0 0.000 0.000 54.644 54.650 multiply_cannon_loop 111 8.7 0.135 0.146 49.708 53.594 density_matrix_trs4 2 5.0 0.002 0.003 48.919 49.125 ls_scf_init_scf 1 4.0 0.000 0.000 29.973 29.974 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.735 28.844 mp_waitall_1 9105 10.9 20.207 28.780 20.207 28.780 multiply_cannon_multrec 1332 9.7 13.484 16.835 22.884 27.357 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.299 26.310 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 10.953 20.422 make_m2s 222 7.7 0.007 0.008 15.093 15.656 make_images 222 8.7 1.386 1.716 15.062 15.626 dbcsr_mm_accdrv_process 4041 10.4 0.314 0.531 8.994 10.542 dbcsr_mm_accdrv_process_sort 4041 11.4 8.511 10.011 8.511 10.011 hybrid_alltoall_any 227 10.6 0.546 2.549 8.053 9.692 make_images_data 222 9.7 0.005 0.006 8.718 9.680 mp_sum_l 887 5.1 5.180 8.511 5.180 8.511 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.166 7.817 mp_irecv_dv 3311 11.0 3.145 7.755 3.145 7.755 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.909 6.734 calculate_norms 2376 9.8 6.081 6.721 6.081 6.721 multiply_cannon_sync_h2d 1332 9.7 4.934 6.250 4.934 6.250 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.922 5.186 arnoldi_extremal 4 6.8 0.000 0.000 4.662 4.689 arnoldi_normal_ev 4 7.8 0.001 0.005 4.662 4.689 build_subspace 16 8.4 0.014 0.021 4.413 4.418 ls_scf_post 1 4.0 0.000 0.000 3.706 3.711 ls_scf_store_result 1 5.0 0.000 0.000 3.387 3.529 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.157 3.385 dbcsr_matrix_vector_mult_local 304 10.0 2.754 3.241 2.756 3.243 mp_allgather_i34 111 8.7 0.823 3.069 0.823 3.069 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.190 2.866 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.653 2.741 dbcsr_data_new 4174 10.1 2.116 2.406 2.116 2.406 make_images_pack 222 9.7 1.824 2.146 1.827 2.148 dbcsr_sort_data 436 11.2 1.804 2.049 1.804 2.049 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.899 1.901 rebuild_ks_matrix 3 7.3 0.000 0.000 1.886 1.888 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.886 1.888 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.418000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1757.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_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 2.841592E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.053 0.076 92.727 92.728 qs_energies 1 2.0 0.000 0.000 92.002 92.006 ls_scf 1 3.0 0.000 0.000 90.415 90.418 dbcsr_multiply_generic 111 6.7 0.016 0.017 74.608 74.880 multiply_cannon 111 7.7 0.033 0.140 51.501 56.152 ls_scf_main 1 4.0 0.000 0.000 56.090 56.095 multiply_cannon_loop 111 8.7 0.116 0.130 48.778 52.603 density_matrix_trs4 2 5.0 0.002 0.003 50.162 50.285 mp_waitall_1 7281 11.0 22.849 33.015 22.849 33.015 ls_scf_init_scf 1 4.0 0.000 0.000 30.633 30.636 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.433 29.506 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.601 26.613 multiply_cannon_multrec 888 9.7 12.837 15.280 21.499 24.596 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 10.354 22.257 make_m2s 222 7.7 0.007 0.008 16.423 17.127 make_images 222 8.7 1.602 1.877 16.386 17.087 make_images_data 222 9.7 0.004 0.005 9.554 10.423 hybrid_alltoall_any 227 10.6 0.642 2.957 9.053 10.153 dbcsr_mm_accdrv_process 3754 10.4 0.311 0.493 8.198 9.481 dbcsr_mm_accdrv_process_sort 3754 11.4 7.763 8.988 7.763 8.988 mp_sum_l 887 5.1 4.962 8.479 4.962 8.479 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.472 7.234 mp_irecv_dv 2335 11.1 2.456 7.168 2.456 7.168 multiply_cannon_sync_h2d 888 9.7 5.998 7.091 5.998 7.091 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.792 6.678 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.590 6.176 arnoldi_extremal 4 6.8 0.000 0.000 5.130 5.153 arnoldi_normal_ev 4 7.8 0.001 0.005 5.130 5.153 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.731 5.019 calculate_norms 1584 9.8 4.400 4.879 4.400 4.879 build_subspace 16 8.4 0.014 0.020 4.822 4.830 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.450 3.792 ls_scf_post 1 4.0 0.000 0.000 3.692 3.695 mp_allgather_i34 111 8.7 0.835 3.653 0.835 3.653 dbcsr_matrix_vector_mult_local 304 10.0 3.032 3.609 3.034 3.611 ls_scf_store_result 1 5.0 0.000 0.000 3.446 3.517 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.886 2.984 dbcsr_data_new 4116 9.9 2.106 2.449 2.106 2.449 make_images_sizes 222 9.7 0.000 0.000 1.019 2.348 mp_alltoall_i44 222 10.7 1.018 2.348 1.018 2.348 dbcsr_sort_data 325 11.1 1.884 2.210 1.884 2.210 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.915 1.917 rebuild_ks_matrix 3 7.3 0.000 0.000 1.897 1.899 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.897 1.899 make_images_pack 222 9.7 1.628 1.893 1.631 1.896 dbcsr_finalize 304 7.8 0.027 0.032 1.625 1.893 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.728000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2280.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_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 3.320033E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.063 98.530 98.531 qs_energies 1 2.0 0.000 0.000 97.804 97.814 ls_scf 1 3.0 0.001 0.007 95.708 95.719 dbcsr_multiply_generic 111 6.7 0.018 0.022 79.340 79.630 ls_scf_main 1 4.0 0.003 0.033 59.437 59.450 multiply_cannon 111 7.7 0.041 0.069 52.042 56.616 density_matrix_trs4 2 5.0 0.003 0.013 53.277 53.389 multiply_cannon_loop 111 8.7 0.152 0.167 46.990 50.060 ls_scf_init_scf 1 4.0 0.000 0.002 32.983 32.992 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.727 31.794 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.107 29.121 mp_waitall_1 6369 11.0 22.864 28.497 22.864 28.497 multiply_cannon_multrec 1332 9.7 14.312 17.286 22.334 25.821 make_m2s 222 7.7 0.008 0.010 21.340 22.728 make_images 222 8.7 3.173 3.631 21.288 22.679 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.192 17.547 make_images_data 222 9.7 0.005 0.010 11.846 13.434 hybrid_alltoall_any 227 10.6 0.804 3.779 11.218 12.942 dbcsr_mm_accdrv_process 3641 10.4 0.295 0.486 7.663 9.197 dbcsr_mm_accdrv_process_sort 3641 11.4 7.217 8.701 7.217 8.701 mp_sum_l 887 5.1 4.352 8.128 4.352 8.128 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.298 6.298 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.075 6.037 mp_irecv_dv 3229 10.9 2.048 5.946 2.048 5.946 multiply_cannon_sync_h2d 1332 9.7 5.427 5.910 5.427 5.910 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.668 5.552 arnoldi_extremal 4 6.8 0.000 0.000 5.250 5.278 arnoldi_normal_ev 4 7.8 0.001 0.005 5.250 5.278 build_subspace 16 8.4 0.014 0.020 4.902 4.912 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.659 4.835 calculate_norms 2376 9.8 4.193 4.537 4.193 4.537 mp_allgather_i34 111 8.7 2.163 4.406 2.163 4.406 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.576 3.881 dbcsr_matrix_vector_mult_local 304 10.0 3.202 3.700 3.204 3.702 dbcsr_sort_data 658 11.4 3.103 3.547 3.103 3.547 ls_scf_post 1 4.0 0.001 0.008 3.287 3.296 dbcsr_special_finalize 555 9.7 0.006 0.007 2.839 3.274 dbcsr_merge_single_wm 555 10.7 0.536 0.658 2.830 3.266 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.042 3.106 ls_scf_store_result 1 5.0 0.000 0.000 3.006 3.091 dbcsr_data_release 10477 10.7 1.599 2.402 1.599 2.402 qs_energies_init_hamiltonians 1 3.0 0.001 0.003 2.087 2.087 dbcsr_finalize 304 7.8 0.050 0.061 1.825 2.030 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.964 1.973 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.531000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2673.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_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 4.764238E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.076 0.119 92.401 92.408 qs_energies 1 2.0 0.000 0.000 91.317 91.318 ls_scf 1 3.0 0.000 0.000 89.112 89.114 dbcsr_multiply_generic 111 6.7 0.018 0.019 70.477 70.704 ls_scf_main 1 4.0 0.000 0.000 56.275 56.275 multiply_cannon 111 7.7 0.086 0.153 52.303 55.483 multiply_cannon_loop 111 8.7 0.088 0.094 49.735 51.126 density_matrix_trs4 2 5.0 0.002 0.003 49.216 49.311 ls_scf_init_scf 1 4.0 0.000 0.000 29.413 29.414 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.126 28.168 mp_waitall_1 5436 11.0 23.880 27.785 23.880 27.785 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.981 25.992 multiply_cannon_multrec 444 9.7 13.786 16.678 20.935 22.256 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.383 15.117 make_m2s 222 7.7 0.005 0.006 13.430 14.358 make_images 222 8.7 2.059 2.508 13.362 14.290 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.177 13.899 make_images_data 222 9.7 0.003 0.004 8.137 9.605 hybrid_alltoall_any 227 10.6 0.801 3.818 8.091 9.462 multiply_cannon_sync_h2d 444 9.7 6.726 8.047 6.726 8.047 dbcsr_mm_accdrv_process 3003 10.4 0.349 0.403 6.842 7.960 dbcsr_mm_accdrv_process_sort 3003 11.4 6.482 7.566 6.482 7.566 arnoldi_extremal 4 6.8 0.000 0.000 5.828 5.839 arnoldi_normal_ev 4 7.8 0.001 0.004 5.828 5.839 build_subspace 16 8.4 0.015 0.020 5.433 5.443 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.442 4.684 mp_sum_l 887 5.1 2.770 4.429 2.770 4.429 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.170 4.374 dbcsr_matrix_vector_mult_local 304 10.0 3.703 4.186 3.705 4.188 calculate_norms 792 9.8 3.624 3.713 3.624 3.713 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.554 3.678 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.595 3.663 mp_irecv_dv 1241 11.2 1.541 3.646 1.541 3.646 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.941 3.550 mp_allgather_i34 111 8.7 1.132 3.513 1.132 3.513 ls_scf_post 1 4.0 0.000 0.000 3.424 3.426 make_images_sizes 222 9.7 0.000 0.000 0.873 3.275 mp_alltoall_i44 222 10.7 0.873 3.274 0.873 3.274 ls_scf_store_result 1 5.0 0.000 0.000 3.206 3.244 dbcsr_finalize 304 7.8 0.062 0.077 2.199 2.318 dbcsr_data_new 4608 9.7 1.796 2.289 1.796 2.289 qs_energies_init_hamiltonians 1 3.0 0.199 0.217 2.187 2.187 dbcsr_merge_all 275 8.9 0.480 0.537 2.056 2.154 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.077 2.078 rebuild_ks_matrix 3 7.3 0.000 0.000 2.044 2.046 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.044 2.045 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.408000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3736.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_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 8.841535E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.137 0.167 107.011 107.012 qs_energies 1 2.0 0.000 0.000 105.217 105.220 ls_scf 1 3.0 0.000 0.000 102.267 102.270 dbcsr_multiply_generic 111 6.7 0.024 0.028 75.807 75.954 ls_scf_main 1 4.0 0.000 0.000 63.334 63.334 density_matrix_trs4 2 5.0 0.002 0.003 54.318 54.382 multiply_cannon 111 7.7 0.139 0.174 49.675 51.706 multiply_cannon_loop 111 8.7 0.099 0.103 46.607 47.204 ls_scf_init_scf 1 4.0 0.000 0.000 35.117 35.118 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.538 33.561 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.811 30.818 mp_waitall_1 4527 11.1 21.431 25.090 21.431 25.090 make_m2s 222 7.7 0.006 0.006 22.406 23.533 make_images 222 8.7 3.598 3.894 22.296 23.421 multiply_cannon_multrec 444 9.7 17.836 18.445 22.532 23.176 hybrid_alltoall_any 227 10.6 1.655 3.616 12.601 15.432 make_images_data 222 9.7 0.004 0.005 12.774 14.938 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.078 11.548 multiply_cannon_sync_h2d 444 9.7 8.792 8.836 8.792 8.836 arnoldi_extremal 4 6.8 0.000 0.000 7.378 7.386 arnoldi_normal_ev 4 7.8 0.003 0.009 7.378 7.386 build_subspace 16 8.4 0.026 0.036 6.823 6.834 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.454 5.612 dbcsr_matrix_vector_mult_local 304 10.0 5.045 5.338 5.047 5.340 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.166 5.255 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.846 5.101 dbcsr_mm_accdrv_process 1814 10.4 0.242 0.356 4.508 4.639 dbcsr_mm_accdrv_process_sort 1814 11.4 4.174 4.310 4.174 4.310 ls_scf_post 1 4.0 0.000 0.000 3.816 3.819 mp_allgather_i34 111 8.7 1.153 3.614 1.153 3.614 make_images_sizes 222 9.7 0.000 0.000 1.456 3.584 mp_alltoall_i44 222 10.7 1.456 3.584 1.456 3.584 ls_scf_store_result 1 5.0 0.000 0.000 3.504 3.539 calculate_norms 792 9.8 3.230 3.270 3.230 3.270 dbcsr_finalize 304 7.8 0.082 0.090 3.084 3.169 dbcsr_merge_all 275 8.9 0.886 0.920 2.868 2.946 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.920 2.920 dbcsr_complete_redistribute 5 7.6 1.429 1.472 2.737 2.851 matrix_ls_to_qs 2 6.0 0.000 0.000 2.406 2.536 dbcsr_sort_data 325 11.1 2.437 2.496 2.437 2.496 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.461 2.462 rebuild_ks_matrix 3 7.3 0.000 0.000 2.394 2.396 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.394 2.396 dbcsr_data_new 6591 9.6 1.866 2.322 1.866 2.322 dbcsr_new_transposed 4 7.5 0.241 0.253 2.275 2.296 dbcsr_frobenius_norm 74 6.6 2.060 2.131 2.205 2.242 dbcsr_add_d 103 6.2 0.000 0.000 2.130 2.208 dbcsr_add_anytype 103 7.2 0.859 0.890 2.130 2.207 compute_matrix_preconditioner 1 6.0 0.002 0.002 2.197 2.205 dbcsr_data_release 12724 10.6 1.986 2.204 1.986 2.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.012000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7030.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2adf4dececbac3e4208898f648fd9522801a7512_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 592.990208E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 751238291. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.243 0.844 230.184 230.189 qs_mol_dyn_low 1 2.0 0.004 0.014 228.155 228.179 qs_forces 5 3.8 0.004 0.024 227.795 227.804 qs_energies 5 4.8 0.002 0.028 224.803 224.840 scf_env_do_scf 5 5.8 0.000 0.011 210.179 210.183 scf_env_do_scf_inner_loop 105 6.6 0.008 0.329 183.626 183.943 qs_scf_new_mos 105 7.6 0.001 0.001 143.044 143.232 qs_scf_loop_do_ot 105 8.6 0.001 0.001 143.044 143.231 dbcsr_multiply_generic 1445 12.2 0.127 0.135 133.815 134.206 ot_scf_mini 105 9.6 0.003 0.005 133.115 133.242 multiply_cannon 1445 13.2 0.273 0.284 114.773 117.281 multiply_cannon_loop 1445 14.2 2.773 2.996 113.091 114.254 velocity_verlet 4 3.0 0.007 0.259 106.616 106.617 ot_mini 105 10.6 0.001 0.003 60.355 60.471 multiply_cannon_multrec 69360 15.2 30.027 34.903 40.006 45.197 qs_ot_get_p 112 10.4 0.001 0.001 42.746 43.020 mp_waitall_1 488190 16.1 34.694 41.689 34.694 41.689 qs_ot_get_derivative 55 11.6 0.001 0.002 38.545 38.697 multiply_cannon_metrocomm3 69360 15.2 0.207 0.222 25.851 33.869 multiply_cannon_sync_h2d 69360 15.2 29.543 33.623 29.543 33.623 qs_ot_p2m_diag 40 11.0 0.020 0.030 31.582 31.650 rebuild_ks_matrix 110 8.4 0.000 0.000 29.061 29.228 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.017 29.061 29.228 cp_dbcsr_syevd 40 12.0 0.002 0.004 28.334 28.335 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.705 26.865 init_scf_loop 7 6.6 0.000 0.009 26.523 26.529 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.102 23.318 apply_single 62 13.6 0.000 0.000 23.102 23.318 cp_fm_syevd 40 13.0 0.000 0.004 23.061 23.253 prepare_preconditioner 7 7.6 0.000 0.000 21.264 21.295 make_preconditioner 7 8.6 0.000 0.002 21.264 21.295 ot_new_cg_direction 55 11.6 0.001 0.003 21.090 21.090 cp_fm_redistribute_end 40 14.0 9.189 18.337 9.194 18.338 cp_fm_syevd_base 40 14.0 9.135 18.283 9.135 18.283 qs_rho_update_rho_low 110 7.6 0.001 0.002 17.243 17.567 calculate_rho_elec 110 8.6 0.030 0.033 17.243 17.567 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.853 15.051 make_full_inverse_cholesky 7 9.6 0.000 0.001 14.183 14.243 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.017 14.115 mp_sum_l 4764 12.2 12.529 13.310 12.529 13.310 rs_pw_transfer 690 11.5 0.011 0.013 11.482 12.845 density_rs2pw 110 9.6 0.006 0.007 11.352 12.738 pw_transfer 1645 12.4 0.084 0.106 11.994 12.240 fft_wrap_pw1pw2 1425 13.5 0.013 0.017 11.851 12.104 calculate_dm_sparse 110 9.5 0.000 0.001 11.911 12.072 init_scf_run 5 5.8 0.000 0.001 12.015 12.016 scf_env_initial_rho_setup 5 6.8 0.002 0.004 12.015 12.016 dbcsr_mm_accdrv_process 154766 15.8 5.834 6.347 9.845 11.036 qs_vxc_create 110 10.4 0.002 0.007 10.432 10.482 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.236 10.316 fft_wrap_pw1pw2_240 915 15.0 1.192 1.296 10.088 10.313 cp_fm_cholesky_invert 7 10.6 9.903 9.911 9.903 9.911 check_diag 80 13.5 8.599 8.926 9.524 9.673 sum_up_and_integrate 60 10.3 0.028 0.031 8.614 8.627 integrate_v_rspace 60 11.3 0.002 0.004 8.586 8.599 fft3d_pb 915 16.0 2.390 2.670 8.155 8.417 acc_transpose_blocks 69360 15.2 0.341 0.374 7.181 7.797 xc_rho_set_and_dset_create 110 12.4 0.077 0.099 7.329 7.592 multiply_cannon_metrocomm1 69360 15.2 0.096 0.105 4.625 7.245 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 7.227 7.243 calculate_first_density_matrix 1 7.0 0.000 0.003 7.226 7.242 xc_vxc_pw_create 60 11.3 0.039 0.052 7.000 7.049 make_full_single_inverse 7 9.6 0.001 0.004 6.815 6.846 make_m2s 2890 13.2 0.082 0.092 6.205 6.806 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.747 6.795 make_images 2890 14.2 0.239 0.259 6.098 6.699 xc_pw_derive 510 13.4 0.005 0.007 6.185 6.291 mp_waitany 7680 13.5 4.370 5.672 4.370 5.672 mp_alltoall_z22v 2340 17.7 5.216 5.507 5.216 5.507 acc_transpose_blocks_kernels 69360 16.2 0.807 0.876 4.574 4.943 potential_pw2rs 60 12.3 0.003 0.003 4.840 4.864 multiply_cannon_metrocomm4 67915 15.2 0.185 0.215 2.011 4.668 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=230.189000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 2adf4dececbac3e4208898f648fd9522801a7512 Summary: empty Status: OK