=== 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: edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.12.0, 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: 28.08.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.12.0 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/01 job id: 48671393 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/02 job id: 48671394 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/03 job id: 48671395 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/04 job id: 48671396 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/05 job id: 48671397 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/06 job id: 48671398 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/07 job id: 48671399 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/08 job id: 48671400 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/09 job id: 48671401 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/10 job id: 48671402 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/11 job id: 48671403 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/12 job id: 48671404 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/13 job id: 48671405 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/14 job id: 48671406 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/15 job id: 48671407 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/16 job id: 48671408 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/17 job id: 48671409 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/18 job id: 48671410 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/19 job id: 48671411 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/20 job id: 48671412 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/21 job id: 48671413 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/22 job id: 48671415 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/23 job id: 48671416 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/24 job id: 48671417 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/25 job id: 48671418 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/26 job id: 48671419 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/27 job id: 48671420 --- 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 0.020 0.033 138.024 138.025 farming_run 1 2.0 134.814 134.817 137.985 137.989 create_force_env 2 3.0 3.075 3.075 3.075 3.075 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.488957E+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 0.025 115.651 115.651 qs_energies 1 2.0 0.000 0.000 115.428 115.432 mp2_main 1 3.0 0.000 0.000 113.073 113.079 mp2_gpw_main 1 4.0 0.028 0.034 111.904 111.909 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 92.675 93.063 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.679 55.069 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.165 41.083 46.200 get_2c_integrals 1 6.0 0.008 0.009 36.962 37.823 integrate_v_rspace 273 8.0 0.436 0.451 24.656 29.497 pw_transfer 6555 10.6 0.380 0.403 26.862 27.622 fft_wrap_pw1pw2 5465 11.4 0.044 0.048 25.470 26.250 grid_integrate_task_list 273 9.0 20.562 25.885 20.562 25.885 fft_wrap_pw1pw2_100 2178 12.4 1.147 1.433 23.017 23.764 compute_2c_integrals 1 7.0 0.003 0.005 19.460 19.461 rpa_ri_compute_en 1 5.0 0.023 0.025 19.113 19.344 compute_2c_integrals_loop_lm 1 8.0 0.013 0.025 18.676 19.176 mp2_eri_2c_integrate_gpw 1 9.0 2.357 2.421 18.662 19.174 cp_fm_cholesky_decompose 12 8.2 17.603 18.464 17.603 18.464 cholesky_decomp 1 7.0 0.000 0.000 16.352 17.221 fft3d_s 5443 13.4 16.143 16.509 16.166 16.532 ao_to_mo_and_store_B_mult_1 272 7.0 10.768 15.313 10.768 15.313 calculate_wavefunction 272 8.0 5.425 5.569 12.340 13.009 rpa_num_int 1 6.0 0.001 0.018 10.850 10.853 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.790 10.808 calc_mat_Q 8 8.0 0.000 0.000 9.506 9.605 contract_S_to_Q 8 9.0 0.000 0.000 8.928 9.023 calc_potential_gpw 544 9.5 0.005 0.006 8.211 8.752 parallel_gemm_fm 14 9.1 0.000 0.000 8.519 8.646 parallel_gemm_fm_cosma 14 10.1 8.519 8.646 8.519 8.646 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.120 8.443 potential_pw2rs 545 10.0 0.107 0.110 7.549 8.214 create_integ_mat 1 6.0 0.022 0.027 7.776 7.777 collocate_single_gaussian 272 10.0 0.040 0.042 7.380 7.653 array2fm 1 7.0 0.000 0.000 6.661 7.113 pw_scatter_s 2720 13.7 4.329 4.441 4.329 4.441 pw_gather_s 2722 13.2 3.440 3.817 3.440 3.817 array2fm_buffer_send 1 8.0 2.948 3.164 2.948 3.164 pw_poisson_solve 545 10.5 1.126 1.185 2.223 2.457 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.908971, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2807.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 0.040 0.095 423.815 423.817 farming_run 1 2.0 422.594 422.635 423.744 423.769 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.244000E+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 28665681. 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 0.031 218.514 218.515 qs_energies 1 2.0 0.002 0.014 218.208 218.219 scf_env_do_scf 1 3.0 0.000 0.000 115.885 115.885 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.946 114.953 rebuild_ks_matrix 4 6.0 0.000 0.000 114.945 114.952 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.064 114.945 114.952 hfx_ks_matrix 4 8.0 0.001 0.001 114.578 114.582 integrate_four_center 4 9.0 0.153 0.471 114.577 114.582 integrate_four_center_main 4 10.0 0.158 0.629 102.230 106.445 integrate_four_center_bin 268 11.0 102.072 106.138 102.072 106.138 mp2_main 1 3.0 0.000 0.003 102.023 102.035 mp2_gpw_main 1 4.0 0.052 0.078 101.090 101.106 init_scf_loop 1 4.0 0.000 0.000 97.166 97.166 mp2_ri_gpw_compute_in 1 5.0 0.070 0.081 74.031 75.144 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.794 54.902 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.166 41.574 46.573 integrate_v_rspace 95 8.0 0.395 0.563 28.051 32.869 pw_transfer 2240 10.6 0.143 0.165 29.553 29.988 mp2_ri_gpw_compute_en 1 5.0 0.062 0.074 26.888 29.031 fft_wrap_pw1pw2 1868 11.4 0.017 0.020 28.564 28.935 ao_to_mo_and_store_B_mult_1 91 7.0 10.546 28.638 10.546 28.638 grid_integrate_task_list 95 9.0 23.454 28.464 23.454 28.464 fft_wrap_pw1pw2_100 730 12.4 1.290 1.487 26.307 26.644 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.833 1.891 25.117 25.126 get_2c_integrals 1 6.0 0.000 0.000 20.131 20.176 compute_2c_integrals 1 7.0 0.007 0.016 19.112 19.126 compute_2c_integrals_loop_lm 1 8.0 0.005 0.012 18.661 18.981 mp2_eri_2c_integrate_gpw 1 9.0 1.728 1.824 18.656 18.980 fft3d_s 1823 13.4 18.501 18.868 18.515 18.882 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.716 18.716 calculate_wavefunction 91 8.0 2.013 2.045 9.627 9.829 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.580 8.775 9.222 potential_pw2rs 186 10.0 0.033 0.034 8.451 8.954 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.527 8.125 8.677 local_gemm 172 8.0 8.220 8.659 8.220 8.659 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.182 8.568 calc_potential_gpw 182 9.5 0.002 0.002 7.913 8.233 collocate_single_gaussian 91 10.0 0.017 0.022 7.757 8.018 mp_sync 37 10.5 4.270 7.189 4.270 7.189 integrate_four_center_load 4 10.0 0.000 0.000 6.800 6.805 hfx_load_balance 1 11.0 0.000 0.000 6.800 6.804 mp_sendrecv_dm3 2068 8.0 6.160 6.757 6.160 6.757 mp2_ri_gpw_compute_en_ener 172 7.0 6.339 6.422 6.339 6.422 pw_gather_s 912 13.2 4.498 4.962 4.498 4.962 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.084497, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1503.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 452.345856E+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 441493. 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 0.060 0.178 56.626 56.629 qs_mol_dyn_low 1 2.0 0.003 0.003 56.095 56.104 qs_forces 11 3.9 0.013 0.029 56.018 56.022 qs_energies 11 4.9 0.007 0.010 54.500 54.519 scf_env_do_scf 11 5.9 0.001 0.001 46.562 46.563 scf_env_do_scf_inner_loop 108 6.5 0.003 0.009 44.299 44.300 dbcsr_multiply_generic 2286 12.5 0.095 0.103 34.553 34.895 qs_scf_new_mos 108 7.5 0.000 0.001 33.811 34.113 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.810 34.112 ot_scf_mini 108 9.5 0.002 0.003 32.073 32.294 multiply_cannon 2286 13.5 0.186 0.191 27.009 28.416 multiply_cannon_loop 2286 14.5 1.834 1.942 26.270 27.722 velocity_verlet 10 3.0 0.001 0.002 26.094 26.095 ot_mini 108 10.5 0.001 0.001 19.308 19.553 qs_ot_get_derivative 108 11.5 0.001 0.001 16.314 16.512 mp_waitall_1 245248 16.5 8.492 14.899 8.492 14.899 multiply_cannon_metrocomm3 54864 15.5 0.074 0.082 5.979 12.815 multiply_cannon_multrec 54864 15.5 3.663 5.811 8.156 11.818 qs_ot_get_p 119 10.4 0.001 0.001 8.126 8.431 rebuild_ks_matrix 119 8.3 0.000 0.000 8.190 8.333 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 8.190 8.333 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.236 7.373 mp_sum_l 7287 12.8 5.392 6.864 5.392 6.864 init_scf_run 11 5.9 0.000 0.001 6.505 6.505 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.504 6.505 dbcsr_mm_accdrv_process 76910 16.1 1.840 2.895 4.406 6.315 multiply_cannon_sync_h2d 54864 15.5 5.148 5.890 5.148 5.890 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.421 5.875 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.641 5.747 qs_ot_p2m_diag 50 11.0 0.004 0.007 5.395 5.442 sum_up_and_integrate 119 10.3 0.001 0.002 4.681 4.687 integrate_v_rspace 119 11.3 0.002 0.003 4.670 4.677 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.603 4.603 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.367 4.542 calculate_rho_elec 119 8.7 0.011 0.017 4.366 4.542 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.365 4.365 cp_fm_redistribute_end 50 14.0 2.230 4.333 2.237 4.337 calculate_first_density_matrix 1 7.0 0.003 0.021 4.249 4.254 cp_fm_diag_elpa_base 50 14.0 2.094 4.213 2.098 4.222 calculate_dm_sparse 119 9.5 0.001 0.003 3.290 3.442 jit_kernel_multiply 13 15.8 2.502 3.397 2.502 3.397 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.957 3.157 apply_single 119 13.6 0.000 0.000 2.957 3.157 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.104 3.114 acc_transpose_blocks 54864 15.5 0.234 0.262 2.469 3.102 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.754 2.809 multiply_cannon_metrocomm1 54864 15.5 0.056 0.063 1.754 2.790 ot_diis_step 108 11.5 0.006 0.006 2.718 2.719 density_rs2pw 119 9.7 0.004 0.004 2.370 2.469 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.409 2.463 init_scf_loop 11 6.9 0.001 0.005 2.240 2.241 wfi_extrapolate 11 7.9 0.001 0.001 2.169 2.169 grid_integrate_task_list 119 12.3 2.013 2.101 2.013 2.101 potential_pw2rs 119 12.3 0.004 0.004 1.999 2.017 pw_transfer 1439 11.6 0.052 0.056 1.906 1.985 mp_sum_d 4135 12.0 1.356 1.973 1.356 1.973 make_m2s 4572 13.5 0.054 0.056 1.886 1.942 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.831 1.912 make_images 4572 14.5 0.134 0.140 1.802 1.858 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.720 1.750 mp_alltoall_d11v 2130 13.8 1.459 1.669 1.459 1.669 fft3d_ps 1201 14.6 0.371 0.478 1.547 1.622 transfer_rs2pw 487 10.6 0.005 0.006 1.523 1.621 mp_waitany 12084 13.8 1.379 1.565 1.379 1.565 fft_wrap_pw1pw2_140 487 13.2 0.139 0.156 1.409 1.489 acc_transpose_blocks_sync 164592 16.5 1.220 1.463 1.220 1.463 parallel_gemm_fm 81 9.0 0.000 0.000 1.436 1.440 parallel_gemm_fm_cosma 81 10.0 1.436 1.440 1.436 1.440 transfer_pw2rs 487 13.2 0.005 0.006 1.414 1.428 grid_collocate_task_list 119 9.7 1.353 1.422 1.353 1.422 acc_transpose_blocks_kernels 54864 16.5 0.249 0.387 0.969 1.352 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.747 1.163 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=56.629000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.468480E+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 992535. 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 0.057 0.143 42.708 42.710 qs_mol_dyn_low 1 2.0 0.003 0.003 42.162 42.172 qs_forces 11 3.9 0.006 0.009 42.089 42.093 qs_energies 11 4.9 0.002 0.008 40.378 40.387 scf_env_do_scf 11 5.9 0.001 0.001 33.527 33.527 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 30.872 30.873 dbcsr_multiply_generic 2286 12.5 0.102 0.104 23.796 24.162 qs_scf_new_mos 108 7.5 0.001 0.001 21.734 21.962 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.733 21.961 ot_scf_mini 108 9.5 0.003 0.003 20.799 20.961 multiply_cannon 2286 13.5 0.211 0.219 18.198 19.652 velocity_verlet 10 3.0 0.002 0.002 19.201 19.202 multiply_cannon_loop 2286 14.5 1.201 1.272 16.923 18.457 ot_mini 108 10.5 0.001 0.001 12.809 13.037 mp_waitall_1 200699 16.5 6.239 11.948 6.239 11.948 qs_ot_get_derivative 108 11.5 0.001 0.001 10.395 10.562 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.465 10.358 multiply_cannon_multrec 27432 15.5 1.830 4.248 6.690 9.645 dbcsr_mm_accdrv_process 47894 16.0 4.077 6.668 4.779 7.385 rebuild_ks_matrix 119 8.3 0.000 0.000 7.067 7.216 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.066 7.215 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.217 6.352 init_scf_run 11 5.9 0.000 0.001 5.520 5.521 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.520 5.520 qs_ot_get_p 119 10.4 0.001 0.001 4.997 5.213 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.778 4.704 mp_sum_l 7287 12.8 2.253 4.285 2.253 4.285 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.104 4.103 apply_single 119 13.6 0.000 0.000 3.103 4.103 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.909 3.944 calculate_rho_elec 119 8.7 0.021 0.024 3.909 3.943 sum_up_and_integrate 119 10.3 0.001 0.001 3.933 3.938 integrate_v_rspace 119 11.3 0.002 0.003 3.919 3.923 calculate_first_density_matrix 1 7.0 0.002 0.012 3.921 3.923 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.228 3.247 make_m2s 4572 13.5 0.052 0.054 2.913 3.241 make_images 4572 14.5 0.205 0.245 2.824 3.155 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.803 2.803 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.650 2.652 init_scf_loop 11 6.9 0.000 0.003 2.633 2.633 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.474 2.559 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.432 2.469 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.429 2.430 calculate_dm_sparse 119 9.5 0.000 0.001 2.324 2.407 cp_fm_redistribute_end 50 14.0 1.229 2.393 1.236 2.397 ot_diis_step 108 11.5 0.011 0.011 2.364 2.366 cp_fm_diag_elpa_base 50 14.0 1.126 2.281 1.156 2.324 density_rs2pw 119 9.7 0.004 0.004 2.186 2.268 acc_transpose_blocks 27432 15.5 0.116 0.121 1.829 2.202 multiply_cannon_sync_h2d 27432 15.5 1.682 2.175 1.682 2.175 jit_kernel_multiply 7 16.6 0.644 2.105 0.644 2.105 pw_transfer 1439 11.6 0.065 0.069 2.014 2.049 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.922 1.960 grid_integrate_task_list 119 12.3 1.840 1.959 1.840 1.959 prepare_preconditioner 11 7.9 0.000 0.000 1.721 1.748 make_preconditioner 11 8.9 0.000 0.003 1.721 1.748 make_images_data 4572 15.5 0.048 0.055 1.344 1.731 potential_pw2rs 119 12.3 0.006 0.006 1.659 1.667 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.610 1.666 fft3d_ps 1201 14.6 0.521 0.576 1.559 1.589 hybrid_alltoall_any 4725 16.4 0.054 0.116 1.179 1.564 wfi_extrapolate 11 7.9 0.001 0.001 1.536 1.536 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.512 1.523 fft_wrap_pw1pw2_140 487 13.2 0.159 0.168 1.470 1.509 transfer_rs2pw 487 10.6 0.005 0.005 1.337 1.457 mp_alltoall_d11v 2130 13.8 1.297 1.431 1.297 1.431 grid_collocate_task_list 119 9.7 1.290 1.362 1.290 1.362 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.239 1.285 acc_transpose_blocks_kernels 27432 16.5 0.189 0.277 0.865 1.122 mp_allgather_i34 2286 14.5 0.694 1.098 0.694 1.098 mp_sum_d 4135 12.0 0.596 1.069 0.596 1.069 transfer_pw2rs 487 13.2 0.004 0.005 1.063 1.069 make_basis_sm 11 9.8 0.219 0.298 1.027 1.029 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.022 1.027 make_images_sizes 4572 15.5 0.005 0.005 0.695 0.973 mp_alltoall_i44 4572 16.5 0.691 0.969 0.691 0.969 acc_transpose_blocks_sync 82296 16.5 0.822 0.952 0.822 0.952 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.910 0.923 mp_alltoall_z22v 1201 16.6 0.791 0.868 0.791 0.868 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=42.710000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.545455, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 520.069120E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.025 36.579 36.580 qs_mol_dyn_low 1 2.0 0.003 0.003 36.217 36.224 qs_forces 11 3.9 0.002 0.003 36.150 36.152 qs_energies 11 4.9 0.002 0.003 34.556 34.559 scf_env_do_scf 11 5.9 0.000 0.001 27.719 27.719 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.993 24.993 dbcsr_multiply_generic 2286 12.5 0.097 0.098 19.589 19.803 qs_scf_new_mos 108 7.5 0.001 0.001 16.774 16.791 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.773 16.790 multiply_cannon 2286 13.5 0.195 0.200 15.712 16.657 velocity_verlet 10 3.0 0.001 0.002 16.478 16.480 ot_scf_mini 108 9.5 0.002 0.002 15.953 15.971 multiply_cannon_loop 2286 14.5 0.857 0.906 14.855 15.939 ot_mini 108 10.5 0.001 0.001 9.800 9.827 multiply_cannon_multrec 18288 15.5 1.874 2.854 8.423 8.671 qs_ot_get_derivative 108 11.5 0.001 0.001 8.187 8.203 dbcsr_mm_accdrv_process 38222 16.0 6.358 7.163 6.457 7.228 rebuild_ks_matrix 119 8.3 0.000 0.000 6.289 6.329 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.289 6.329 init_scf_run 11 5.9 0.000 0.001 5.614 5.614 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.614 5.614 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.557 5.594 mp_waitall_1 158411 16.6 3.235 4.828 3.235 4.828 calculate_first_density_matrix 1 7.0 0.000 0.000 4.238 4.240 qs_ot_get_p 119 10.4 0.001 0.001 3.789 3.824 sum_up_and_integrate 119 10.3 0.001 0.001 3.803 3.810 integrate_v_rspace 119 11.3 0.003 0.003 3.790 3.799 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.047 3.638 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.554 3.560 calculate_rho_elec 119 8.7 0.031 0.031 3.553 3.560 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.212 3.213 multiply_cannon_metrocomm3 18288 15.5 0.047 0.050 1.733 3.177 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.932 2.946 init_scf_loop 11 6.9 0.000 0.000 2.704 2.704 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.518 2.529 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.064 2.446 apply_single 119 13.6 0.000 0.000 2.064 2.446 calculate_dm_sparse 119 9.5 0.000 0.001 2.280 2.293 make_m2s 4572 13.5 0.044 0.045 2.102 2.277 density_rs2pw 119 9.7 0.004 0.004 2.102 2.212 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.193 2.194 make_images 4572 14.5 0.192 0.204 2.017 2.191 pw_transfer 1439 11.6 0.065 0.068 2.011 2.025 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.919 1.934 acc_transpose_blocks 18288 15.5 0.080 0.083 1.829 1.934 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.919 1.930 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.900 1.912 cp_fm_diag_elpa_base 50 14.0 1.875 1.888 1.898 1.909 prepare_preconditioner 11 7.9 0.000 0.000 1.892 1.895 make_preconditioner 11 8.9 0.000 0.000 1.892 1.895 grid_integrate_task_list 119 12.3 1.799 1.892 1.799 1.892 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.729 1.824 mp_sum_l 7287 12.8 1.380 1.796 1.380 1.796 potential_pw2rs 119 12.3 0.007 0.008 1.574 1.583 ot_diis_step 108 11.5 0.011 0.012 1.573 1.574 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.566 1.572 fft_wrap_pw1pw2_140 487 13.2 0.211 0.216 1.502 1.516 fft3d_ps 1201 14.6 0.528 0.544 1.488 1.505 wfi_extrapolate 11 7.9 0.001 0.001 1.320 1.320 grid_collocate_task_list 119 9.7 1.232 1.304 1.232 1.304 transfer_rs2pw 487 10.6 0.005 0.005 1.166 1.270 multiply_cannon_sync_h2d 18288 15.5 1.004 1.201 1.004 1.201 make_images_data 4572 15.5 0.048 0.052 0.906 1.096 acc_transpose_blocks_kernels 18288 16.5 0.217 0.223 1.000 1.061 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.043 1.056 transfer_pw2rs 487 13.2 0.004 0.004 0.986 0.991 hybrid_alltoall_any 4725 16.4 0.058 0.115 0.789 0.988 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 0.973 0.975 mp_alltoall_d11v 2130 13.8 0.810 0.930 0.810 0.930 mp_alltoall_z22v 1201 16.6 0.796 0.863 0.796 0.863 jit_kernel_transpose 5 15.6 0.782 0.838 0.782 0.838 cp_fm_cholesky_invert 11 10.9 0.822 0.829 0.822 0.829 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.812 0.815 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.478 0.795 acc_transpose_blocks_sync 54864 16.5 0.730 0.788 0.730 0.788 build_core_hamiltonian_matrix_ 11 4.9 0.002 0.013 0.680 0.752 dbcsr_complete_redistribute 329 12.2 0.155 0.179 0.618 0.736 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=36.580000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.090909, yerr=1.311110 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 557.969408E+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 0.055 0.140 42.456 42.458 qs_mol_dyn_low 1 2.0 0.003 0.003 42.012 42.031 qs_forces 11 3.9 0.002 0.003 41.927 41.931 qs_energies 11 4.9 0.026 0.056 40.153 40.160 scf_env_do_scf 11 5.9 0.001 0.002 32.928 32.930 scf_env_do_scf_inner_loop 108 6.5 0.010 0.024 29.058 29.059 dbcsr_multiply_generic 2286 12.5 0.103 0.111 22.579 22.705 velocity_verlet 10 3.0 0.002 0.002 20.426 20.429 qs_scf_new_mos 108 7.5 0.001 0.001 20.163 20.235 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.163 20.234 ot_scf_mini 108 9.5 0.002 0.003 19.037 19.107 multiply_cannon 2286 13.5 0.220 0.226 17.754 18.208 multiply_cannon_loop 2286 14.5 1.573 1.649 16.607 17.011 ot_mini 108 10.5 0.001 0.001 11.422 11.505 multiply_cannon_multrec 27432 15.5 2.449 3.146 9.978 10.240 qs_ot_get_derivative 108 11.5 0.001 0.001 9.410 9.483 dbcsr_mm_accdrv_process 47916 15.9 6.555 8.389 7.426 8.786 rebuild_ks_matrix 119 8.3 0.000 0.000 6.797 6.851 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.020 6.797 6.850 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.061 6.112 init_scf_run 11 5.9 0.000 0.001 5.667 5.667 scf_env_initial_rho_setup 11 6.9 0.002 0.002 5.666 5.667 qs_ot_get_p 119 10.4 0.002 0.003 4.506 4.590 calculate_first_density_matrix 1 7.0 0.000 0.002 4.070 4.073 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.312 3.850 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.806 3.848 calculate_rho_elec 119 8.7 0.040 0.046 3.806 3.847 init_scf_loop 11 6.9 0.001 0.003 3.838 3.839 sum_up_and_integrate 119 10.3 0.001 0.001 3.814 3.818 integrate_v_rspace 119 11.3 0.003 0.003 3.803 3.807 mp_waitall_1 137007 16.6 2.474 3.215 2.474 3.215 make_m2s 4572 13.5 0.055 0.059 2.827 2.995 prepare_preconditioner 11 7.9 0.000 0.000 2.912 2.920 make_preconditioner 11 8.9 0.000 0.003 2.912 2.920 make_images 4572 14.5 0.272 0.331 2.718 2.884 acc_transpose_blocks 27432 15.5 0.123 0.126 2.559 2.869 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.846 2.847 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.507 2.847 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.800 2.809 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.338 2.753 apply_single 119 13.6 0.000 0.000 2.337 2.753 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.718 2.752 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.566 2.592 calculate_dm_sparse 119 9.5 0.000 0.000 2.520 2.574 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.340 2.342 density_rs2pw 119 9.7 0.004 0.004 2.204 2.310 pw_transfer 1439 11.6 0.065 0.069 2.232 2.272 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.140 2.184 mp_sum_l 7287 12.8 1.594 2.163 1.594 2.163 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.996 2.005 cp_fm_diag_elpa_base 50 14.0 1.958 1.974 1.994 2.002 ot_diis_step 108 11.5 0.012 0.012 1.962 1.964 jit_kernel_multiply 6 15.8 0.805 1.951 0.805 1.951 grid_integrate_task_list 119 12.3 1.846 1.920 1.846 1.920 fft_wrap_pw1pw2_140 487 13.2 0.246 0.261 1.670 1.715 fft3d_ps 1201 14.6 0.557 0.607 1.660 1.691 acc_transpose_blocks_sync 82296 16.5 1.493 1.618 1.493 1.618 potential_pw2rs 119 12.3 0.008 0.009 1.570 1.577 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.534 1.546 wfi_extrapolate 11 7.9 0.001 0.001 1.529 1.529 multiply_cannon_metrocomm3 27432 15.5 0.041 0.042 0.884 1.505 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.461 1.475 grid_collocate_task_list 119 9.7 1.249 1.328 1.249 1.328 transfer_rs2pw 487 10.6 0.004 0.005 1.206 1.325 make_images_data 4572 15.5 0.048 0.054 1.092 1.286 dbcsr_complete_redistribute 329 12.2 0.136 0.183 0.949 1.231 cp_fm_upper_to_full 72 14.2 0.854 1.216 0.854 1.216 qs_energies_init_hamiltonians 11 5.9 0.008 0.016 1.204 1.207 hybrid_alltoall_any 4725 16.4 0.065 0.154 0.940 1.174 acc_transpose_blocks_kernels 27432 16.5 0.272 0.280 0.914 1.104 mp_alltoall_d11v 2130 13.8 0.948 1.057 0.948 1.057 mp_alltoall_z22v 1201 16.6 0.984 1.009 0.984 1.009 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.685 0.960 make_basis_sm 11 9.8 0.151 0.250 0.956 0.958 cp_fm_cholesky_invert 11 10.9 0.953 0.956 0.953 0.956 multiply_cannon_metrocomm1 27432 15.5 0.035 0.036 0.358 0.943 transfer_pw2rs 487 13.2 0.004 0.004 0.924 0.928 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.002 0.801 0.885 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.851 0.857 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=42.458000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.727273, yerr=4.574453 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/07/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 117.977176E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1384136 0.0% 0.0% 100.0% average stack size 0.0 0.0 106.2 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 622.862336E+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 0.037 31.120 31.122 qs_mol_dyn_low 1 2.0 0.003 0.003 30.886 30.896 qs_forces 11 3.9 0.002 0.003 30.817 30.819 qs_energies 11 4.9 0.002 0.003 29.115 29.118 scf_env_do_scf 11 5.9 0.000 0.001 22.519 22.519 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 20.004 20.004 dbcsr_multiply_generic 2286 12.5 0.100 0.108 15.148 15.283 velocity_verlet 10 3.0 0.002 0.002 14.903 14.907 multiply_cannon 2286 13.5 0.226 0.236 12.389 12.990 qs_scf_new_mos 108 7.5 0.001 0.001 12.371 12.393 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.370 12.393 multiply_cannon_loop 2286 14.5 0.649 0.683 11.455 11.779 ot_scf_mini 108 9.5 0.002 0.002 11.659 11.682 multiply_cannon_multrec 9144 15.5 1.764 1.941 7.654 7.872 ot_mini 108 10.5 0.001 0.001 6.820 6.849 dbcsr_mm_accdrv_process 12550 15.8 4.212 5.813 5.779 5.949 rebuild_ks_matrix 119 8.3 0.000 0.000 5.790 5.819 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.789 5.818 qs_ot_get_derivative 108 11.5 0.001 0.001 5.519 5.540 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.154 5.181 init_scf_run 11 5.9 0.000 0.001 5.122 5.122 scf_env_initial_rho_setup 11 6.9 0.001 0.003 5.122 5.122 calculate_first_density_matrix 1 7.0 0.000 0.000 3.914 3.915 sum_up_and_integrate 119 10.3 0.001 0.001 3.462 3.466 integrate_v_rspace 119 11.3 0.003 0.003 3.452 3.456 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.340 3.347 calculate_rho_elec 119 8.7 0.060 0.061 3.340 3.346 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.980 2.981 qs_ot_get_p 119 10.4 0.002 0.002 2.906 2.936 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.815 2.827 jit_kernel_multiply 6 15.6 1.528 2.607 1.528 2.607 init_scf_loop 11 6.9 0.000 0.000 2.496 2.497 mp_waitall_1 115863 16.7 1.531 2.066 1.531 2.066 pw_transfer 1439 11.6 0.065 0.067 2.018 2.026 make_m2s 4572 13.5 0.034 0.035 1.863 2.022 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.927 1.980 calculate_dm_sparse 119 9.5 0.000 0.000 1.944 1.964 grid_integrate_task_list 119 12.3 1.867 1.951 1.867 1.951 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.925 1.935 make_images 4572 14.5 0.269 0.299 1.775 1.932 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.885 1.888 density_rs2pw 119 9.7 0.004 0.004 1.749 1.821 prepare_preconditioner 11 7.9 0.000 0.000 1.748 1.753 make_preconditioner 11 8.9 0.000 0.000 1.748 1.753 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.637 1.666 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.628 1.629 acc_transpose_blocks 9144 15.5 0.042 0.043 1.562 1.626 fft_wrap_pw1pw2_140 487 13.2 0.323 0.331 1.572 1.583 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.400 1.412 grid_collocate_task_list 119 9.7 1.295 1.394 1.295 1.394 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.250 1.366 apply_single 119 13.6 0.000 0.000 1.250 1.366 fft3d_ps 1201 14.6 0.568 0.579 1.338 1.352 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.337 1.345 cp_fm_diag_elpa_base 50 14.0 1.309 1.327 1.335 1.344 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.329 1.335 ot_diis_step 108 11.5 0.013 0.013 1.283 1.284 potential_pw2rs 119 12.3 0.010 0.011 1.272 1.275 qs_energies_init_hamiltonians 11 5.9 0.015 0.018 1.255 1.256 wfi_extrapolate 11 7.9 0.001 0.001 1.156 1.156 hybrid_alltoall_any 4725 16.4 0.065 0.175 0.792 1.023 make_images_data 4572 15.5 0.042 0.046 0.805 0.991 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.923 cp_fm_cholesky_invert 11 10.9 0.889 0.892 0.889 0.892 transfer_rs2pw 487 10.6 0.004 0.004 0.811 0.878 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.866 0.872 mp_alltoall_d11v 2130 13.8 0.776 0.826 0.776 0.826 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.776 0.789 acc_transpose_blocks_sync 27432 16.5 0.728 0.782 0.728 0.782 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.757 0.761 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.684 0.733 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 0.394 0.691 transfer_pw2rs 487 13.2 0.003 0.004 0.683 0.685 mp_allgather_i34 2286 14.5 0.246 0.683 0.246 0.683 mp_alltoall_z22v 1201 16.6 0.643 0.679 0.643 0.679 jit_kernel_transpose 5 15.6 0.658 0.671 0.658 0.671 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=31.122000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=590.090909, yerr=5.230236 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 772.632576E+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 0.041 0.075 44.771 44.773 qs_mol_dyn_low 1 2.0 0.003 0.003 44.436 44.443 qs_forces 11 3.9 0.018 0.045 44.369 44.371 qs_energies 11 4.9 0.005 0.006 42.329 42.347 scf_env_do_scf 11 5.9 0.001 0.001 34.909 34.909 scf_env_do_scf_inner_loop 108 6.5 0.004 0.008 27.033 27.035 velocity_verlet 10 3.0 0.002 0.002 24.300 24.306 dbcsr_multiply_generic 2286 12.5 0.103 0.107 20.725 20.912 qs_scf_new_mos 108 7.5 0.001 0.001 17.522 17.611 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.522 17.610 multiply_cannon 2286 13.5 0.301 0.305 16.657 17.553 ot_scf_mini 108 9.5 0.002 0.002 16.339 16.435 multiply_cannon_loop 2286 14.5 0.873 0.888 15.349 16.282 multiply_cannon_multrec 9144 15.5 3.318 4.738 10.059 10.202 ot_mini 108 10.5 0.001 0.001 9.931 10.044 qs_ot_get_derivative 108 11.5 0.001 0.001 7.834 7.932 dbcsr_mm_accdrv_process 12550 15.8 5.335 7.637 6.603 7.885 init_scf_loop 11 6.9 0.000 0.000 7.848 7.850 rebuild_ks_matrix 119 8.3 0.000 0.000 7.191 7.327 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.191 7.326 prepare_preconditioner 11 7.9 0.000 0.000 6.873 6.885 make_preconditioner 11 8.9 0.000 0.000 6.873 6.885 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.454 6.761 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.481 6.604 init_scf_run 11 5.9 0.000 0.001 5.324 5.324 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.324 5.324 cp_fm_upper_to_full 72 14.2 3.187 4.552 3.187 4.552 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.112 4.168 calculate_rho_elec 119 8.7 0.118 0.121 4.112 4.167 calculate_first_density_matrix 1 7.0 0.000 0.000 3.812 3.863 sum_up_and_integrate 119 10.3 0.001 0.001 3.809 3.820 integrate_v_rspace 119 11.3 0.003 0.004 3.798 3.810 qs_ot_get_p 119 10.4 0.001 0.002 3.423 3.555 mp_waitall_1 94719 16.7 2.573 3.502 2.573 3.502 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.891 3.291 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.073 3.075 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.817 2.863 dbcsr_complete_redistribute 329 12.2 0.291 0.296 1.963 2.812 pw_transfer 1439 11.6 0.068 0.069 2.659 2.666 make_m2s 4572 13.5 0.038 0.038 2.476 2.654 calculate_dm_sparse 119 9.5 0.000 0.000 2.532 2.575 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.561 2.567 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.243 2.556 apply_single 119 13.6 0.000 0.000 2.242 2.556 make_images 4572 14.5 0.352 0.387 2.355 2.533 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.662 2.504 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.470 2.351 jit_kernel_multiply 6 15.5 1.239 2.272 1.239 2.272 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.414 2.250 mp_alltoall_i22 627 13.8 1.398 2.197 1.398 2.197 fft_wrap_pw1pw2_140 487 13.2 0.575 0.579 2.144 2.151 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.084 2.132 density_rs2pw 119 9.7 0.004 0.004 2.085 2.110 grid_integrate_task_list 119 12.3 2.092 2.109 2.092 2.109 ot_diis_step 108 11.5 0.014 0.015 2.070 2.070 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.015 2.018 acc_transpose_blocks 9144 15.5 0.045 0.046 1.900 1.945 qs_energies_init_hamiltonians 11 5.9 0.004 0.005 1.793 1.811 mp_sum_l 7287 12.8 1.042 1.747 1.042 1.747 fft3d_ps 1201 14.6 0.597 0.609 1.657 1.664 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.652 1.653 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.518 1.565 grid_collocate_task_list 119 9.7 1.531 1.552 1.531 1.552 cp_fm_cholesky_invert 11 10.9 1.469 1.472 1.469 1.472 hybrid_alltoall_any 4725 16.4 0.090 0.150 1.126 1.412 wfi_extrapolate 11 7.9 0.001 0.001 1.401 1.401 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.364 1.364 cp_fm_diag_elpa_base 50 14.0 1.214 1.272 1.361 1.362 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.328 1.350 make_images_data 4572 15.5 0.047 0.050 1.092 1.322 potential_pw2rs 119 12.3 0.014 0.014 1.302 1.304 mp_alltoall_d11v 2130 13.8 1.214 1.264 1.214 1.264 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.119 1.138 acc_transpose_blocks_sync 27432 16.5 1.110 1.131 1.110 1.131 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.998 1.045 qs_create_task_list 11 7.9 0.001 0.001 0.929 0.940 generate_qs_task_list 11 8.9 0.367 0.386 0.929 0.939 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.926 0.939 mp_alltoall_z22v 1201 16.6 0.923 0.936 0.923 0.936 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.773000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=728.000000, yerr=10.427237 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 500.183040E+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 1566934. 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 0.034 0.045 83.556 83.557 qs_mol_dyn_low 1 2.0 0.003 0.003 83.170 83.180 qs_forces 11 3.9 0.003 0.004 83.090 83.094 qs_energies 11 4.9 0.003 0.003 80.214 80.232 scf_env_do_scf 11 5.9 0.000 0.001 70.014 70.017 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.504 64.504 dbcsr_multiply_generic 2055 12.4 0.112 0.121 52.479 52.817 qs_scf_new_mos 99 7.5 0.000 0.001 47.461 47.572 qs_scf_loop_do_ot 99 8.5 0.000 0.001 47.460 47.571 ot_scf_mini 99 9.5 0.002 0.002 45.061 45.164 multiply_cannon 2055 13.4 0.194 0.221 43.511 44.591 multiply_cannon_loop 2055 14.4 1.790 1.832 42.544 43.672 velocity_verlet 10 3.0 0.001 0.001 42.964 42.965 ot_mini 99 10.5 0.001 0.001 26.911 27.022 qs_ot_get_derivative 99 11.5 0.001 0.001 20.126 20.224 multiply_cannon_multrec 49320 15.4 11.437 12.087 18.260 19.068 rebuild_ks_matrix 110 8.3 0.000 0.000 14.488 14.670 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.488 14.670 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.709 12.868 mp_waitall_1 220248 16.4 10.622 11.988 10.622 11.988 multiply_cannon_sync_h2d 49320 15.4 9.664 10.373 9.664 10.373 qs_ot_get_p 110 10.4 0.001 0.001 9.661 9.780 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.498 8.311 init_scf_run 11 5.9 0.000 0.001 8.133 8.133 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.133 8.133 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.563 8.078 apply_single 110 13.6 0.000 0.000 7.563 8.078 multiply_cannon_metrocomm3 49320 15.4 0.084 0.088 6.449 8.068 dbcsr_mm_accdrv_process 87628 16.1 2.998 3.117 6.693 7.178 sum_up_and_integrate 110 10.3 0.002 0.003 6.980 6.992 integrate_v_rspace 110 11.3 0.003 0.003 6.954 6.972 ot_diis_step 99 11.5 0.006 0.006 6.595 6.595 qs_ot_p2m_diag 48 11.0 0.012 0.020 6.515 6.553 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.335 6.477 calculate_rho_elec 110 8.6 0.021 0.025 6.335 6.476 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.672 5.672 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.616 5.671 mp_sum_l 6594 12.7 4.547 5.529 4.547 5.529 init_scf_loop 11 6.9 0.000 0.000 5.481 5.481 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.098 5.116 cp_fm_diag_elpa_base 48 14.0 5.085 5.104 5.097 5.115 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.917 4.922 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.302 4.380 make_m2s 4110 13.4 0.061 0.066 3.974 4.081 wfi_extrapolate 11 7.9 0.001 0.001 4.040 4.040 calculate_first_density_matrix 1 7.0 0.000 0.000 3.992 3.997 make_images 4110 14.4 0.177 0.189 3.877 3.988 jit_kernel_multiply 13 15.9 3.416 3.894 3.416 3.894 calculate_dm_sparse 110 9.5 0.001 0.001 3.771 3.873 density_rs2pw 110 9.6 0.004 0.005 3.333 3.556 multiply_cannon_metrocomm1 49320 15.4 0.065 0.069 2.338 3.544 grid_integrate_task_list 110 12.3 3.258 3.400 3.258 3.400 prepare_preconditioner 11 7.9 0.000 0.000 3.294 3.310 make_preconditioner 11 8.9 0.000 0.000 3.294 3.310 acc_transpose_blocks 49320 15.4 0.213 0.221 2.714 3.227 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.174 3.214 pw_transfer 1331 11.6 0.055 0.065 3.090 3.166 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.080 3.133 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.001 3.079 fft_wrap_pw1pw2_140 451 13.1 0.390 0.431 2.570 2.655 potential_pw2rs 110 12.3 0.006 0.006 2.574 2.592 mp_alltoall_d11v 2046 13.8 2.096 2.589 2.096 2.589 fft3d_ps 1111 14.6 0.796 0.887 2.296 2.355 grid_collocate_task_list 110 9.6 2.157 2.290 2.157 2.290 mp_waitany 14300 13.8 1.832 2.181 1.832 2.181 transfer_rs2pw 451 10.6 0.005 0.006 1.881 2.168 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.908 1.935 make_images_data 4110 15.4 0.044 0.047 1.777 1.928 mp_sum_d 3889 11.9 1.316 1.825 1.316 1.825 hybrid_alltoall_any 4261 16.3 0.084 0.480 1.536 1.811 cp_fm_cholesky_invert 11 10.9 1.788 1.792 1.788 1.792 transfer_pw2rs 451 13.1 0.006 0.007 1.746 1.756 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.663 1.691 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.557000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.909091, yerr=2.609582 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 590.106624E+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 2092664. 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 0.055 0.057 71.492 71.494 qs_mol_dyn_low 1 2.0 0.003 0.003 70.630 70.881 qs_forces 11 3.9 0.003 0.005 70.536 70.538 qs_energies 11 4.9 0.002 0.002 67.191 67.196 scf_env_do_scf 11 5.9 0.000 0.001 58.537 58.540 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 50.602 50.603 dbcsr_multiply_generic 2055 12.4 0.116 0.120 38.607 38.799 velocity_verlet 10 3.0 0.002 0.002 37.070 37.071 qs_scf_new_mos 99 7.5 0.001 0.001 34.277 34.391 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.277 34.390 ot_scf_mini 99 9.5 0.003 0.003 32.552 32.676 multiply_cannon 2055 13.4 0.225 0.244 31.541 32.542 multiply_cannon_loop 2055 14.4 1.172 1.197 30.231 31.019 ot_mini 99 10.5 0.001 0.001 18.835 18.962 multiply_cannon_multrec 24660 15.4 6.991 8.614 14.057 15.578 rebuild_ks_matrix 110 8.3 0.000 0.000 13.796 13.897 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.796 13.897 qs_ot_get_derivative 99 11.5 0.001 0.001 12.970 13.096 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.147 12.232 mp_waitall_1 176588 16.5 7.773 10.562 7.773 10.562 init_scf_loop 11 6.9 0.000 0.000 7.893 7.893 multiply_cannon_metrocomm3 24660 15.4 0.073 0.075 5.157 7.751 multiply_cannon_sync_h2d 24660 15.4 6.402 7.515 6.402 7.515 dbcsr_mm_accdrv_process 52282 16.1 5.556 6.315 6.897 7.219 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.508 7.167 apply_single 110 13.6 0.000 0.001 6.507 7.166 qs_ot_get_p 110 10.4 0.001 0.001 6.966 7.106 sum_up_and_integrate 110 10.3 0.001 0.002 6.510 6.524 integrate_v_rspace 110 11.3 0.002 0.003 6.483 6.498 init_scf_run 11 5.9 0.000 0.001 6.248 6.249 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.248 6.249 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.163 6.177 calculate_rho_elec 110 8.6 0.039 0.047 6.162 6.177 prepare_preconditioner 11 7.9 0.000 0.000 5.845 5.870 make_preconditioner 11 8.9 0.000 0.000 5.845 5.870 ot_diis_step 99 11.5 0.010 0.010 5.809 5.809 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.406 5.560 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.813 5.543 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.958 4.979 make_m2s 4110 13.4 0.057 0.060 4.459 4.883 make_images 4110 14.4 0.407 0.463 4.350 4.771 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.461 4.461 pw_transfer 1331 11.6 0.066 0.073 3.774 3.912 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.846 3.857 cp_fm_diag_elpa_base 48 14.0 3.799 3.815 3.844 3.854 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.667 3.809 wfi_extrapolate 11 7.9 0.001 0.001 3.610 3.610 density_rs2pw 110 9.6 0.004 0.005 3.365 3.543 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.450 3.510 grid_integrate_task_list 110 12.3 3.170 3.322 3.170 3.322 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.294 3.296 fft_wrap_pw1pw2_140 451 13.1 0.461 0.476 3.095 3.237 calculate_dm_sparse 110 9.5 0.001 0.001 3.039 3.070 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.933 2.974 fft3d_ps 1111 14.6 1.111 1.341 2.762 2.917 make_images_data 4110 15.4 0.049 0.054 2.403 2.865 hybrid_alltoall_any 4261 16.3 0.107 0.458 2.106 2.846 cp_fm_cholesky_invert 11 10.9 2.750 2.761 2.750 2.761 mp_sum_l 6594 12.7 1.981 2.650 1.981 2.650 calculate_first_density_matrix 1 7.0 0.000 0.000 2.538 2.539 potential_pw2rs 110 12.3 0.008 0.009 2.375 2.389 grid_collocate_task_list 110 9.6 2.172 2.320 2.172 2.320 mp_alltoall_d11v 2046 13.8 1.873 2.063 1.873 2.063 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.042 2.061 acc_transpose_blocks 24660 15.4 0.117 0.119 1.995 2.027 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.868 1.870 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.810 1.823 transfer_rs2pw 451 10.6 0.006 0.007 1.549 1.742 jit_kernel_multiply 9 16.3 0.976 1.718 0.976 1.718 mp_allgather_i34 2055 14.4 0.616 1.682 0.616 1.682 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.561 1.570 multiply_cannon_metrocomm4 22605 15.4 0.079 0.082 0.772 1.533 mp_waitany 10164 13.8 1.269 1.494 1.269 1.494 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.386 1.488 dbcsr_complete_redistribute 325 12.2 0.237 0.293 1.207 1.480 mp_alltoall_z22v 1111 16.6 1.392 1.465 1.392 1.465 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.494000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=558.272727, yerr=6.046651 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 666.578944E+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 0.018 0.039 60.263 60.263 qs_mol_dyn_low 1 2.0 0.003 0.003 59.873 59.884 qs_forces 11 3.9 0.003 0.005 59.801 59.804 qs_energies 11 4.9 0.003 0.012 56.613 56.619 scf_env_do_scf 11 5.9 0.001 0.002 48.469 48.469 scf_env_do_scf_inner_loop 99 6.5 0.003 0.009 39.925 39.926 velocity_verlet 10 3.0 0.002 0.003 32.742 32.743 dbcsr_multiply_generic 2055 12.4 0.109 0.130 28.885 29.197 qs_scf_new_mos 99 7.5 0.001 0.001 25.279 25.375 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.279 25.374 ot_scf_mini 99 9.5 0.003 0.004 24.029 24.148 multiply_cannon 2055 13.4 0.216 0.228 22.537 23.878 multiply_cannon_loop 2055 14.4 0.817 0.841 21.355 22.246 ot_mini 99 10.5 0.001 0.001 13.761 13.877 rebuild_ks_matrix 110 8.3 0.000 0.000 12.152 12.300 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 12.152 12.300 multiply_cannon_multrec 16440 15.4 3.720 5.296 9.915 11.394 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.700 10.832 mp_waitall_1 139946 16.5 6.891 9.893 6.891 9.893 qs_ot_get_derivative 99 11.5 0.001 0.001 9.334 9.449 init_scf_loop 11 6.9 0.001 0.004 8.510 8.512 multiply_cannon_metrocomm3 16440 15.4 0.046 0.048 4.272 7.118 prepare_preconditioner 11 7.9 0.000 0.000 6.733 6.751 make_preconditioner 11 8.9 0.000 0.001 6.733 6.751 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.025 6.377 sum_up_and_integrate 110 10.3 0.001 0.002 6.268 6.282 integrate_v_rspace 110 11.3 0.003 0.003 6.242 6.257 dbcsr_mm_accdrv_process 34862 16.1 5.367 5.709 6.038 6.205 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.776 5.782 calculate_rho_elec 110 8.6 0.059 0.059 5.775 5.781 init_scf_run 11 5.9 0.000 0.001 5.686 5.686 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.686 5.686 qs_ot_get_p 110 10.4 0.001 0.002 5.401 5.552 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.882 5.401 apply_single 110 13.6 0.000 0.000 4.881 5.400 make_m2s 4110 13.4 0.049 0.051 4.289 4.675 make_images 4110 14.4 0.397 0.522 4.174 4.558 ot_diis_step 99 11.5 0.011 0.011 4.385 4.386 multiply_cannon_sync_h2d 16440 15.4 3.260 4.089 3.260 4.089 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.264 3.843 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.758 3.761 pw_transfer 1331 11.6 0.065 0.071 3.585 3.595 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.479 3.491 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.387 3.388 grid_integrate_task_list 110 12.3 3.199 3.357 3.199 3.357 density_rs2pw 110 9.6 0.004 0.005 2.996 3.187 wfi_extrapolate 11 7.9 0.001 0.001 3.038 3.038 fft_wrap_pw1pw2_140 451 13.1 0.579 0.588 3.000 3.014 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.892 2.893 make_images_data 4110 15.4 0.047 0.078 2.385 2.881 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.851 2.862 cp_fm_diag_elpa_base 48 14.0 2.783 2.819 2.850 2.860 hybrid_alltoall_any 4261 16.3 0.111 0.381 2.149 2.774 cp_fm_cholesky_invert 11 10.9 2.633 2.639 2.633 2.639 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.513 2.577 calculate_dm_sparse 110 9.5 0.001 0.001 2.540 2.572 calculate_first_density_matrix 1 7.0 0.000 0.002 2.552 2.554 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.425 2.487 multiply_cannon_metrocomm4 14385 15.4 0.048 0.053 0.877 2.442 fft3d_ps 1111 14.6 1.089 1.097 2.421 2.430 grid_collocate_task_list 110 9.6 2.219 2.383 2.219 2.383 mp_irecv_dv 48980 15.7 0.801 2.303 0.801 2.303 mp_sum_l 6594 12.7 1.434 2.118 1.434 2.118 potential_pw2rs 110 12.3 0.011 0.011 2.108 2.115 mp_alltoall_d11v 2046 13.8 1.855 2.040 1.855 2.040 qs_energies_init_hamiltonians 11 5.9 0.006 0.012 2.010 2.013 dbcsr_complete_redistribute 325 12.2 0.334 0.382 1.512 1.962 cp_fm_upper_to_full 70 14.2 1.367 1.735 1.367 1.735 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.649 1.666 acc_transpose_blocks 16440 15.4 0.077 0.079 1.580 1.650 cp_fm_cholesky_decompose 22 10.9 1.610 1.629 1.610 1.629 mp_allgather_i34 2055 14.4 0.489 1.516 0.489 1.516 transfer_rs2pw 451 10.6 0.005 0.006 1.324 1.505 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.057 1.495 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.475 1.489 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.372 1.482 mp_waitany 17072 13.8 1.175 1.395 1.175 1.395 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.281 1.292 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.263000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.181818, yerr=8.065845 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 739.033088E+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 0.065 0.167 70.385 70.394 qs_mol_dyn_low 1 2.0 0.003 0.003 69.215 69.225 qs_forces 11 3.9 0.043 0.059 69.091 69.113 qs_energies 11 4.9 0.006 0.018 65.624 65.635 scf_env_do_scf 11 5.9 0.001 0.004 56.192 56.195 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.157 44.158 velocity_verlet 10 3.0 0.002 0.002 38.324 38.326 dbcsr_multiply_generic 2055 12.4 0.115 0.122 32.454 32.671 qs_scf_new_mos 99 7.5 0.001 0.001 29.193 29.309 qs_scf_loop_do_ot 99 8.5 0.001 0.001 29.192 29.308 ot_scf_mini 99 9.5 0.003 0.004 27.487 27.595 multiply_cannon 2055 13.4 0.246 0.264 24.750 25.877 multiply_cannon_loop 2055 14.4 1.429 1.516 23.110 23.873 ot_mini 99 10.5 0.001 0.001 15.866 15.997 multiply_cannon_multrec 24660 15.4 4.098 6.834 13.396 14.637 rebuild_ks_matrix 110 8.3 0.000 0.000 12.193 12.317 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.018 12.193 12.317 init_scf_loop 11 6.9 0.002 0.015 11.992 11.994 qs_ot_get_derivative 99 11.5 0.001 0.001 11.571 11.681 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.794 10.905 dbcsr_mm_accdrv_process 52304 16.0 7.918 9.933 9.140 10.235 prepare_preconditioner 11 7.9 0.000 0.000 10.190 10.206 make_preconditioner 11 8.9 0.001 0.002 10.190 10.206 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.404 9.867 mp_waitall_1 121746 16.5 4.807 6.695 4.807 6.695 qs_ot_get_p 110 10.4 0.001 0.002 6.367 6.524 init_scf_run 11 5.9 0.000 0.001 6.496 6.497 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.496 6.496 sum_up_and_integrate 110 10.3 0.001 0.002 6.333 6.346 integrate_v_rspace 110 11.3 0.003 0.004 6.307 6.319 make_m2s 4110 13.4 0.060 0.062 5.632 6.100 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.020 6.033 calculate_rho_elec 110 8.6 0.078 0.081 6.019 6.032 make_images 4110 14.4 0.577 0.700 5.490 5.955 cp_fm_upper_to_full 70 14.2 3.390 4.907 3.390 4.907 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.229 4.579 apply_single 110 13.6 0.000 0.000 4.229 4.579 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.365 4.382 ot_diis_step 99 11.5 0.011 0.011 4.237 4.239 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.985 4.145 dbcsr_complete_redistribute 325 12.2 0.433 0.493 2.789 3.966 pw_transfer 1331 11.6 0.065 0.074 3.833 3.869 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.726 3.764 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.751 3.751 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.570 3.630 grid_integrate_task_list 110 12.3 3.284 3.441 3.284 3.441 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.406 3.408 multiply_cannon_metrocomm3 24660 15.4 0.039 0.040 1.633 3.381 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.216 3.375 calculate_first_density_matrix 1 7.0 0.001 0.006 3.296 3.298 hybrid_alltoall_any 4261 16.3 0.123 0.463 2.373 3.228 fft_wrap_pw1pw2_140 451 13.1 0.609 0.632 3.183 3.223 density_rs2pw 110 9.6 0.004 0.004 3.090 3.210 make_images_data 4110 15.4 0.051 0.070 2.733 3.203 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.183 3.194 cp_fm_diag_elpa_base 48 14.0 3.024 3.083 3.180 3.192 wfi_extrapolate 11 7.9 0.001 0.001 3.098 3.098 calculate_dm_sparse 110 9.5 0.001 0.001 3.016 3.051 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.968 3.006 mp_alltoall_i22 605 13.7 1.726 2.932 1.726 2.932 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.776 2.921 cp_fm_cholesky_invert 11 10.9 2.818 2.827 2.818 2.827 acc_transpose_blocks 24660 15.4 0.115 0.119 2.448 2.637 fft3d_ps 1111 14.6 1.087 1.117 2.610 2.623 multiply_cannon_sync_h2d 24660 15.4 2.386 2.580 2.386 2.580 grid_collocate_task_list 110 9.6 2.266 2.382 2.266 2.382 qs_energies_init_hamiltonians 11 5.9 0.002 0.007 2.347 2.358 potential_pw2rs 110 12.3 0.012 0.013 2.107 2.112 mp_sum_l 6594 12.7 1.442 2.094 1.442 2.094 mp_alltoall_d11v 2046 13.8 1.928 2.079 1.928 2.079 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.878 1.917 jit_kernel_multiply 8 15.8 0.877 1.916 0.877 1.916 cp_fm_cholesky_decompose 22 10.9 1.750 1.796 1.750 1.796 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.002 1.612 1.708 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.667 1.679 mp_allgather_i34 2055 14.4 0.719 1.670 0.719 1.670 acc_transpose_blocks_sync 73980 16.4 1.417 1.594 1.417 1.594 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.572 1.590 multiply_cannon_metrocomm4 20550 15.4 0.062 0.065 0.837 1.487 transfer_rs2pw 451 10.6 0.005 0.006 1.294 1.429 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=70.394000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=697.727273, yerr=11.029637 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 877.793280E+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 0.033 0.067 64.430 64.439 qs_mol_dyn_low 1 2.0 0.003 0.003 64.093 64.107 qs_forces 11 3.9 0.022 0.033 64.003 64.004 qs_energies 11 4.9 0.011 0.025 60.207 60.230 scf_env_do_scf 11 5.9 0.000 0.001 49.955 49.955 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 40.794 40.795 velocity_verlet 10 3.0 0.002 0.002 34.676 34.680 dbcsr_multiply_generic 2055 12.4 0.110 0.114 27.291 27.560 qs_scf_new_mos 99 7.5 0.001 0.001 25.002 25.086 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.001 25.085 ot_scf_mini 99 9.5 0.002 0.003 23.606 23.676 multiply_cannon 2055 13.4 0.241 0.249 20.163 21.724 multiply_cannon_loop 2055 14.4 0.606 0.634 18.643 19.212 rebuild_ks_matrix 110 8.3 0.000 0.000 12.475 12.570 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.475 12.569 ot_mini 99 10.5 0.001 0.001 12.363 12.437 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.113 11.192 multiply_cannon_multrec 8220 15.4 3.136 4.420 8.691 9.855 mp_waitall_1 103326 16.6 7.467 9.745 7.467 9.745 init_scf_loop 11 6.9 0.000 0.000 9.103 9.105 qs_ot_get_derivative 99 11.5 0.001 0.001 8.322 8.395 prepare_preconditioner 11 7.9 0.000 0.000 7.327 7.338 make_preconditioner 11 8.9 0.000 0.000 7.327 7.338 init_scf_run 11 5.9 0.000 0.001 6.964 6.964 scf_env_initial_rho_setup 11 6.9 0.007 0.010 6.964 6.964 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.853 6.950 qs_ot_get_p 110 10.4 0.001 0.001 6.827 6.925 sum_up_and_integrate 110 10.3 0.001 0.002 6.726 6.740 integrate_v_rspace 110 11.3 0.003 0.003 6.700 6.713 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.569 6.581 calculate_rho_elec 110 8.6 0.114 0.114 6.568 6.581 dbcsr_mm_accdrv_process 17442 15.9 3.164 3.939 5.413 6.338 make_m2s 4110 13.4 0.038 0.040 5.160 5.511 make_images 4110 14.4 0.643 0.701 5.030 5.380 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 3.395 5.286 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.961 4.967 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.535 4.537 pw_transfer 1331 11.6 0.066 0.070 4.433 4.451 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.326 4.343 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.984 4.068 apply_single 110 13.6 0.000 0.000 3.983 4.068 ot_diis_step 99 11.5 0.012 0.012 3.994 3.995 calculate_first_density_matrix 1 7.0 0.000 0.000 3.855 3.864 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.776 3.783 cp_fm_diag_elpa_base 48 14.0 3.711 3.739 3.772 3.780 cp_fm_cholesky_invert 11 10.9 3.770 3.777 3.770 3.777 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.696 3.698 fft_wrap_pw1pw2_140 451 13.1 0.777 0.790 3.607 3.635 density_rs2pw 110 9.6 0.004 0.004 3.429 3.555 grid_integrate_task_list 110 12.3 3.369 3.517 3.369 3.517 make_images_data 4110 15.4 0.041 0.047 2.812 3.319 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.262 3.294 hybrid_alltoall_any 4261 16.3 0.200 0.852 2.746 3.237 calculate_dm_sparse 110 9.5 0.001 0.001 3.049 3.087 fft3d_ps 1111 14.6 1.150 1.199 2.991 3.022 wfi_extrapolate 11 7.9 0.001 0.001 2.986 2.987 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.716 2.736 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.516 2.555 mp_alltoall_d11v 2046 13.8 2.280 2.491 2.280 2.491 grid_collocate_task_list 110 9.6 2.365 2.489 2.365 2.489 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.421 2.465 multiply_cannon_sync_h2d 8220 15.4 2.317 2.441 2.317 2.441 potential_pw2rs 110 12.3 0.015 0.015 2.368 2.373 jit_kernel_multiply 7 15.7 1.934 2.243 1.934 2.243 mp_allgather_i34 2055 14.4 0.687 2.090 0.687 2.090 cp_fm_cholesky_decompose 22 10.9 2.021 2.042 2.021 2.042 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 1.296 2.017 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.787 1.994 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.900 1.914 dbcsr_complete_redistribute 325 12.2 0.548 0.571 1.731 1.855 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.712 1.730 acc_transpose_blocks 8220 15.4 0.040 0.041 1.652 1.728 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.548 1.648 mp_alltoall_z22v 1111 16.6 1.604 1.627 1.604 1.627 mp_sum_l 6594 12.7 1.220 1.604 1.220 1.604 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.452 1.478 transfer_rs2pw 451 10.6 0.005 0.005 1.343 1.478 mp_waitany 9240 13.8 1.230 1.402 1.230 1.402 qs_create_task_list 11 7.9 0.001 0.004 1.239 1.338 generate_qs_task_list 11 8.9 0.376 0.445 1.238 1.334 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.471 1.320 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.280 1.310 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=64.439000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=812.363636, yerr=22.507666 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 1.375527E+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 0.016 0.037 90.049 90.051 qs_mol_dyn_low 1 2.0 0.003 0.003 89.401 89.411 qs_forces 11 3.9 0.003 0.004 89.249 89.251 qs_energies 11 4.9 0.047 0.048 85.127 85.130 scf_env_do_scf 11 5.9 0.001 0.001 74.548 74.548 velocity_verlet 10 3.0 0.002 0.003 56.639 56.654 scf_env_do_scf_inner_loop 99 6.5 0.004 0.007 45.940 45.942 dbcsr_multiply_generic 2055 12.4 0.121 0.126 30.341 30.444 init_scf_loop 11 6.9 0.000 0.000 28.533 28.537 qs_scf_new_mos 99 7.5 0.001 0.001 28.061 28.091 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.060 28.090 prepare_preconditioner 11 7.9 0.000 0.000 26.488 26.496 make_preconditioner 11 8.9 0.000 0.000 26.488 26.496 ot_scf_mini 99 9.5 0.002 0.002 26.229 26.253 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.727 25.957 multiply_cannon 2055 13.4 0.361 0.387 22.774 23.527 multiply_cannon_loop 2055 14.4 0.833 0.844 20.903 21.188 cp_fm_upper_to_full 70 14.2 12.548 18.059 12.548 18.059 ot_mini 99 10.5 0.001 0.001 14.572 14.600 rebuild_ks_matrix 110 8.3 0.000 0.001 13.948 13.983 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.948 13.982 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.662 12.692 dbcsr_complete_redistribute 325 12.2 1.017 1.050 7.378 10.585 multiply_cannon_multrec 8220 15.4 4.044 4.186 9.918 10.026 qs_ot_get_derivative 99 11.5 0.001 0.001 9.915 9.937 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.347 9.560 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.746 8.927 mp_alltoall_i22 605 13.7 5.404 8.632 5.404 8.632 mp_waitall_1 84994 16.7 7.796 8.630 7.796 8.630 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.411 7.447 calculate_rho_elec 110 8.6 0.224 0.224 7.411 7.446 sum_up_and_integrate 110 10.3 0.002 0.002 6.989 7.006 integrate_v_rspace 110 11.3 0.004 0.004 6.961 6.978 qs_ot_get_p 110 10.4 0.002 0.003 6.303 6.338 make_m2s 4110 13.4 0.044 0.044 5.654 6.147 init_scf_run 11 5.9 0.000 0.001 6.119 6.119 scf_env_initial_rho_setup 11 6.9 0.003 0.004 6.119 6.119 make_images 4110 14.4 0.884 0.938 5.465 5.956 dbcsr_mm_accdrv_process 11614 15.7 3.819 4.093 5.726 5.936 cp_fm_cholesky_invert 11 10.9 5.542 5.547 5.542 5.547 pw_transfer 1331 11.6 0.074 0.075 5.255 5.263 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.758 5.231 apply_single 110 13.6 0.000 0.000 4.758 5.231 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.877 5.152 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 5.138 5.146 ot_diis_step 99 11.5 0.015 0.017 4.636 4.636 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.586 4.593 fft_wrap_pw1pw2_140 451 13.1 1.276 1.280 4.406 4.411 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.039 4.040 density_rs2pw 110 9.6 0.004 0.004 3.801 3.830 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.769 3.771 grid_integrate_task_list 110 12.3 3.688 3.745 3.688 3.745 hybrid_alltoall_any 4261 16.3 0.264 0.564 2.951 3.701 make_images_data 4110 15.4 0.046 0.049 2.968 3.661 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.155 3.610 wfi_extrapolate 11 7.9 0.001 0.001 3.410 3.410 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.397 3.397 cp_fm_diag_elpa_base 48 14.0 2.825 3.061 3.394 3.395 calculate_dm_sparse 110 9.5 0.001 0.001 3.217 3.247 fft3d_ps 1111 14.6 1.301 1.328 3.192 3.200 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.198 3.199 multiply_cannon_sync_h2d 8220 15.4 3.118 3.135 3.118 3.135 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.826 2.841 grid_collocate_task_list 110 9.6 2.677 2.691 2.677 2.691 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.661 2.672 calculate_first_density_matrix 1 7.0 0.000 0.000 2.590 2.592 potential_pw2rs 110 12.3 0.021 0.021 2.420 2.423 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.294 2.350 mp_alltoall_d11v 2046 13.8 2.264 2.311 2.264 2.311 cp_fm_cholesky_decompose 22 10.9 2.165 2.181 2.165 2.181 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.091 2.167 qs_create_task_list 11 7.9 0.062 0.067 1.961 2.007 jit_kernel_multiply 10 15.2 1.711 1.970 1.711 1.970 generate_qs_task_list 11 8.9 0.730 0.782 1.900 1.950 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.898 1.903 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.793 1.832 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=90.051000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1237.000000, yerr=52.653240 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 625.946624E+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 3427603. 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 0.029 0.080 207.621 207.624 qs_mol_dyn_low 1 2.0 0.003 0.005 206.875 206.893 qs_forces 11 3.9 0.009 0.042 206.531 206.535 qs_energies 11 4.9 0.006 0.020 200.898 200.913 scf_env_do_scf 11 5.9 0.001 0.005 183.576 183.581 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 162.816 162.818 dbcsr_multiply_generic 2507 12.6 0.181 0.188 125.779 126.378 velocity_verlet 10 3.0 0.004 0.014 124.530 124.535 qs_scf_new_mos 117 7.6 0.001 0.001 123.496 123.849 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.495 123.848 ot_scf_mini 117 9.6 0.003 0.004 116.834 117.224 multiply_cannon 2507 13.6 0.242 0.250 101.938 104.386 multiply_cannon_loop 2507 14.6 2.373 2.465 99.763 101.655 ot_mini 117 10.6 0.001 0.002 65.823 66.213 multiply_cannon_multrec 60168 15.6 32.062 34.502 41.940 44.001 qs_ot_get_derivative 117 11.6 0.001 0.002 40.837 41.199 rebuild_ks_matrix 128 8.3 0.001 0.001 33.622 34.053 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.021 33.622 34.053 mp_waitall_1 267128 16.5 28.709 32.272 28.709 32.272 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.182 30.570 qs_ot_get_p 128 10.4 0.001 0.002 28.826 29.281 multiply_cannon_sync_h2d 60168 15.6 26.669 28.915 26.669 28.915 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.456 25.311 apply_single 128 13.6 0.001 0.001 24.456 25.311 ot_diis_step 117 11.6 0.009 0.019 24.727 24.734 qs_ot_p2m_diag 83 11.4 0.079 0.092 22.229 22.326 init_scf_loop 11 6.9 0.001 0.005 20.682 20.683 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.242 19.517 cp_dbcsr_syevd 83 12.4 0.005 0.006 19.387 19.388 multiply_cannon_metrocomm3 60168 15.6 0.120 0.125 15.832 18.268 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.418 16.449 cp_fm_diag_elpa_base 83 14.4 16.350 16.380 16.414 16.446 prepare_preconditioner 11 7.9 0.000 0.001 16.028 16.082 make_preconditioner 11 8.9 0.000 0.002 16.028 16.082 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.238 15.396 make_m2s 5014 13.6 0.106 0.113 13.976 14.331 make_images 5014 14.6 0.396 0.418 13.791 14.155 sum_up_and_integrate 128 10.3 0.009 0.056 13.950 14.000 integrate_v_rspace 128 11.3 0.004 0.005 13.884 13.903 qs_rho_update_rho_low 128 7.7 0.001 0.003 13.090 13.220 calculate_rho_elec 128 8.7 0.045 0.064 13.089 13.217 init_scf_run 11 5.9 0.000 0.002 13.008 13.009 scf_env_initial_rho_setup 11 6.9 0.001 0.003 13.008 13.009 dbcsr_mm_accdrv_process 124484 16.2 4.585 4.835 9.437 10.208 mp_sum_l 7950 12.9 8.618 9.973 8.618 9.973 wfi_extrapolate 11 7.9 0.002 0.013 9.179 9.179 cp_fm_cholesky_invert 11 10.9 9.169 9.176 9.169 9.176 calculate_dm_sparse 128 9.5 0.001 0.001 8.525 8.640 multiply_cannon_metrocomm1 60168 15.6 0.094 0.100 6.315 8.480 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.181 8.321 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.979 8.068 make_images_data 5014 15.6 0.068 0.073 6.882 7.834 pw_transfer 1547 11.6 0.074 0.092 7.570 7.741 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.368 7.542 grid_integrate_task_list 128 12.3 7.054 7.453 7.054 7.453 density_rs2pw 128 9.7 0.006 0.007 6.743 7.281 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.188 7.201 hybrid_alltoall_any 5200 16.5 0.294 2.257 6.015 7.109 fft_wrap_pw1pw2_140 523 13.2 1.129 1.184 6.489 6.653 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.276 6.353 fft3d_ps 1291 14.7 2.188 2.838 5.372 5.699 mp_alltoall_d11v 2415 14.1 4.329 5.507 4.329 5.507 grid_collocate_task_list 128 9.7 4.844 5.297 4.844 5.297 cp_fm_cholesky_decompose 22 10.9 4.655 4.671 4.655 4.671 potential_pw2rs 128 12.3 0.009 0.010 4.447 4.467 mp_sum_d 4470 12.1 3.484 4.187 3.484 4.187 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.624000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.636364, yerr=5.014030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 845.398016E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57903. MP_Allreduce 11138 958. MP_Sync 87 MP_Alltoall 1983 5926009. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.136 191.921 191.930 qs_mol_dyn_low 1 2.0 0.003 0.004 191.345 191.372 qs_forces 11 3.9 0.008 0.018 191.251 191.253 qs_energies 11 4.9 0.006 0.032 184.495 184.512 scf_env_do_scf 11 5.9 0.001 0.002 167.232 167.242 scf_env_do_scf_inner_loop 118 6.6 0.004 0.008 134.412 134.415 velocity_verlet 10 3.0 0.002 0.002 120.409 120.410 dbcsr_multiply_generic 2527 12.6 0.191 0.197 99.124 100.411 qs_scf_new_mos 118 7.6 0.001 0.001 95.312 95.819 qs_scf_loop_do_ot 118 8.6 0.001 0.001 95.311 95.818 ot_scf_mini 118 9.6 0.004 0.005 90.457 91.083 multiply_cannon 2527 13.6 0.508 0.565 78.559 83.649 multiply_cannon_loop 2527 14.6 1.589 1.654 75.097 77.607 ot_mini 118 10.6 0.001 0.001 50.864 51.411 mp_waitall_1 216438 16.6 24.452 39.657 24.452 39.657 multiply_cannon_multrec 30324 15.6 21.146 25.849 32.253 37.401 rebuild_ks_matrix 129 8.3 0.001 0.001 32.663 33.405 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 32.662 33.405 init_scf_loop 11 6.9 0.001 0.005 32.728 32.729 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.407 30.077 qs_ot_get_derivative 118 11.6 0.001 0.002 28.656 29.267 multiply_cannon_metrocomm3 30324 15.6 0.098 0.103 15.918 29.254 prepare_preconditioner 11 7.9 0.000 0.000 28.411 28.479 make_preconditioner 11 8.9 0.000 0.002 28.411 28.479 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.058 27.577 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.294 23.508 apply_single 129 13.6 0.001 0.001 22.293 23.508 qs_ot_get_p 129 10.4 0.001 0.002 21.443 22.102 ot_diis_step 118 11.6 0.014 0.015 22.027 22.029 multiply_cannon_sync_h2d 30324 15.6 18.147 20.147 18.147 20.147 cp_fm_cholesky_invert 11 10.9 16.481 16.493 16.481 16.493 qs_ot_p2m_diag 83 11.4 0.189 0.217 16.392 16.424 make_m2s 5054 13.6 0.090 0.095 14.423 16.127 make_images 5054 14.6 1.177 1.379 14.211 15.916 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.179 15.180 sum_up_and_integrate 129 10.3 0.002 0.004 14.146 14.178 integrate_v_rspace 129 11.3 0.003 0.004 14.086 14.119 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.557 13.584 calculate_rho_elec 129 8.7 0.088 0.105 13.556 13.584 init_scf_run 11 5.9 0.000 0.001 12.368 12.369 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.367 12.369 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.961 11.995 cp_fm_diag_elpa_base 83 14.4 11.702 11.788 11.955 11.985 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.331 11.814 dbcsr_mm_accdrv_process 62734 16.2 5.508 6.441 10.550 11.336 multiply_cannon_metrocomm4 27797 15.6 0.107 0.121 3.892 10.840 make_images_data 5054 15.6 0.068 0.078 8.469 10.509 mp_irecv_dv 70031 16.3 3.682 10.423 3.682 10.423 hybrid_alltoall_any 5240 16.5 0.353 1.559 7.226 9.822 pw_transfer 1559 11.6 0.085 0.106 8.719 8.781 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.494 8.560 wfi_extrapolate 11 7.9 0.001 0.001 8.424 8.424 density_rs2pw 129 9.7 0.006 0.007 7.211 7.712 fft_wrap_pw1pw2_140 527 13.2 1.224 1.246 7.511 7.592 grid_integrate_task_list 129 12.3 7.233 7.555 7.233 7.555 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.430 7.267 cp_fm_cholesky_decompose 22 10.9 6.970 7.067 6.970 7.067 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.861 6.869 calculate_dm_sparse 129 9.5 0.001 0.001 6.518 6.695 mp_sum_l 8010 12.9 4.415 6.395 4.415 6.395 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.927 6.141 fft3d_ps 1301 14.7 2.866 3.033 6.033 6.084 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.430 5.501 grid_collocate_task_list 129 9.7 5.042 5.365 5.042 5.365 mp_alltoall_d11v 2423 14.1 4.225 5.205 4.225 5.205 mp_allgather_i34 2527 14.6 1.969 5.059 1.969 5.059 potential_pw2rs 129 12.3 0.015 0.017 4.557 4.568 dbcsr_complete_redistribute 395 12.7 0.780 0.874 3.244 4.083 mp_sum_d 4499 12.1 2.710 4.044 2.710 4.044 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.930000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=804.363636, yerr=6.064393 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 956.661760E+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 4020 57952. MP_Allreduce 11127 998. 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 0.023 0.073 178.583 178.585 qs_mol_dyn_low 1 2.0 0.004 0.010 178.041 178.060 qs_forces 11 3.9 0.008 0.029 177.291 177.296 qs_energies 11 4.9 0.004 0.020 170.693 170.709 scf_env_do_scf 11 5.9 0.001 0.001 153.918 153.919 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 118.632 118.632 velocity_verlet 10 3.0 0.004 0.013 113.928 113.950 dbcsr_multiply_generic 2507 12.6 0.181 0.187 82.994 84.239 qs_scf_new_mos 117 7.6 0.001 0.001 81.390 81.722 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.389 81.721 ot_scf_mini 117 9.6 0.003 0.004 77.154 77.551 multiply_cannon 2507 13.6 0.511 0.528 63.287 67.101 multiply_cannon_loop 2507 14.6 1.132 1.205 60.001 63.364 ot_mini 117 10.6 0.001 0.002 42.724 43.144 init_scf_loop 11 6.9 0.000 0.000 35.186 35.187 mp_waitall_1 170520 16.6 24.527 34.048 24.527 34.048 prepare_preconditioner 11 7.9 0.000 0.000 31.182 31.237 make_preconditioner 11 8.9 0.000 0.000 31.182 31.237 rebuild_ks_matrix 128 8.3 0.001 0.001 30.328 30.965 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 30.327 30.965 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.879 30.239 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.309 27.887 multiply_cannon_multrec 20056 15.6 13.098 16.154 23.018 26.052 multiply_cannon_metrocomm3 20056 15.6 0.063 0.066 15.201 25.081 qs_ot_get_derivative 117 11.6 0.002 0.003 23.172 23.574 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.611 20.640 apply_single 128 13.6 0.001 0.001 19.611 20.639 qs_ot_get_p 128 10.4 0.001 0.002 19.775 20.292 ot_diis_step 117 11.6 0.020 0.032 19.441 19.450 make_m2s 5014 13.6 0.079 0.084 14.560 16.056 make_images 5014 14.6 1.148 1.248 14.327 15.820 qs_ot_p2m_diag 83 11.4 0.266 0.273 15.334 15.340 multiply_cannon_sync_h2d 20056 15.6 13.698 15.190 13.698 15.190 cp_fm_cholesky_invert 11 10.9 14.677 14.685 14.677 14.685 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.274 14.274 sum_up_and_integrate 128 10.3 0.002 0.003 14.025 14.049 integrate_v_rspace 128 11.3 0.003 0.004 13.965 13.992 qs_rho_update_rho_low 128 7.7 0.001 0.002 13.649 13.690 calculate_rho_elec 128 8.7 0.130 0.145 13.649 13.689 init_scf_run 11 5.9 0.000 0.001 11.644 11.644 scf_env_initial_rho_setup 11 6.9 0.000 0.003 11.644 11.644 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.035 11.056 cp_fm_diag_elpa_base 83 14.4 10.622 10.786 11.031 11.052 make_images_data 5014 15.6 0.063 0.073 8.595 10.544 hybrid_alltoall_any 5200 16.5 0.448 2.025 7.442 9.877 dbcsr_mm_accdrv_process 41502 16.2 6.227 6.543 9.380 9.506 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.195 9.501 multiply_cannon_metrocomm4 17549 15.6 0.067 0.078 3.451 9.197 mp_irecv_dv 50230 16.2 3.321 8.940 3.321 8.940 pw_transfer 1547 11.6 0.085 0.103 8.812 8.919 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.588 8.704 fft_wrap_pw1pw2_140 523 13.2 1.302 1.330 7.586 7.718 grid_integrate_task_list 128 12.3 7.339 7.717 7.339 7.717 cp_fm_cholesky_decompose 22 10.9 7.552 7.613 7.552 7.613 wfi_extrapolate 11 7.9 0.001 0.001 7.534 7.534 density_rs2pw 128 9.7 0.006 0.006 7.006 7.362 cp_fm_upper_to_full 105 14.8 5.642 7.119 5.642 7.119 dbcsr_complete_redistribute 395 12.7 1.188 1.222 4.579 6.306 calculate_dm_sparse 128 9.5 0.001 0.001 6.009 6.132 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.102 6.109 fft3d_ps 1291 14.7 2.770 2.998 5.962 6.043 grid_collocate_task_list 128 9.7 5.159 5.604 5.159 5.604 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.164 5.345 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.706 5.327 mp_sum_l 7950 12.9 3.334 5.178 3.334 5.178 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.404 5.130 mp_alltoall_d11v 2415 14.1 4.293 4.943 4.293 4.943 mp_allgather_i34 2507 14.6 1.772 4.759 1.772 4.759 potential_pw2rs 128 12.3 0.020 0.022 4.450 4.461 transfer_fm_to_dbcsr 11 9.9 0.019 0.024 2.283 3.992 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.927 3.960 calculate_first_density_matrix 1 7.0 0.000 0.000 3.919 3.922 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.812 3.819 mp_alltoall_i22 716 14.1 1.928 3.817 1.928 3.817 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.518 3.577 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.585000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=891.181818, yerr=12.013766 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.144349E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+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 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922227920 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57626. MP_Allreduce 11184 1079. MP_Sync 88 MP_Alltoall 1724 12509604. 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 0.012 0.035 202.284 202.285 qs_mol_dyn_low 1 2.0 0.003 0.003 201.794 201.809 qs_forces 11 3.9 0.005 0.007 201.643 201.650 qs_energies 11 4.9 0.002 0.004 194.424 194.433 scf_env_do_scf 11 5.9 0.001 0.001 175.565 175.578 velocity_verlet 10 3.0 0.015 0.024 133.248 133.251 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 125.637 125.638 qs_scf_new_mos 118 7.6 0.001 0.001 87.470 87.816 qs_scf_loop_do_ot 118 8.6 0.001 0.001 87.469 87.815 dbcsr_multiply_generic 2529 12.6 0.193 0.198 85.526 86.359 ot_scf_mini 118 9.6 0.003 0.004 82.697 83.031 multiply_cannon 2529 13.6 0.553 0.590 58.736 62.292 multiply_cannon_loop 2529 14.6 1.885 1.991 54.745 56.666 init_scf_loop 11 6.9 0.000 0.000 49.798 49.799 ot_mini 118 10.6 0.001 0.001 45.430 45.780 prepare_preconditioner 11 7.9 0.000 0.000 45.583 45.611 make_preconditioner 11 8.9 0.000 0.000 45.583 45.611 make_full_inverse_cholesky 11 9.9 0.011 0.022 39.178 44.154 multiply_cannon_multrec 30348 15.6 13.316 18.770 27.634 32.861 rebuild_ks_matrix 129 8.3 0.001 0.001 30.522 30.836 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 30.522 30.835 mp_waitall_1 149172 16.7 19.383 30.204 19.383 30.204 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.558 27.849 qs_ot_get_derivative 118 11.6 0.002 0.002 25.324 25.669 qs_ot_get_p 129 10.4 0.002 0.003 22.860 23.212 make_m2s 5058 13.6 0.098 0.101 21.674 22.703 make_images 5058 14.6 1.960 2.310 21.361 22.390 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.370 19.974 apply_single 129 13.6 0.001 0.001 19.370 19.974 ot_diis_step 118 11.6 0.018 0.020 19.949 19.951 cp_fm_cholesky_invert 11 10.9 18.289 18.300 18.289 18.300 qs_ot_p2m_diag 84 11.4 0.347 0.395 18.237 18.296 cp_fm_upper_to_full 106 14.8 11.614 17.036 11.614 17.036 multiply_cannon_metrocomm3 30348 15.6 0.051 0.054 7.355 16.981 cp_dbcsr_syevd 84 12.4 0.005 0.006 16.689 16.692 sum_up_and_integrate 129 10.3 0.002 0.003 14.713 14.742 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.629 14.684 calculate_rho_elec 129 8.7 0.174 0.190 14.628 14.684 integrate_v_rspace 129 11.3 0.003 0.004 14.653 14.682 dbcsr_mm_accdrv_process 62780 16.2 8.523 9.383 13.889 14.403 make_images_data 5058 15.6 0.068 0.075 11.738 13.596 dbcsr_complete_redistribute 397 12.7 1.501 1.623 9.446 13.225 cp_fm_diag_elpa 84 13.4 0.000 0.001 13.187 13.202 cp_fm_diag_elpa_base 84 14.4 12.156 12.516 13.178 13.192 init_scf_run 11 5.9 0.000 0.001 12.761 12.763 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.761 12.762 hybrid_alltoall_any 5245 16.5 0.533 2.235 10.550 12.757 multiply_cannon_sync_h2d 30348 15.6 10.446 11.829 10.446 11.829 copy_fm_to_dbcsr 210 11.7 0.001 0.002 7.901 11.686 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 10.814 11.055 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.383 10.093 pw_transfer 1559 11.6 0.086 0.103 9.755 9.846 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.528 9.622 mp_alltoall_i22 720 14.1 5.619 9.451 5.619 9.451 cp_fm_cholesky_decompose 22 10.9 8.591 8.677 8.591 8.677 fft_wrap_pw1pw2_140 527 13.2 1.452 1.483 8.296 8.399 wfi_extrapolate 11 7.9 0.001 0.001 8.160 8.160 grid_integrate_task_list 129 12.3 7.560 7.952 7.560 7.952 multiply_cannon_metrocomm4 25290 15.6 0.085 0.098 2.949 7.788 density_rs2pw 129 9.7 0.006 0.006 7.393 7.621 mp_irecv_dv 76751 16.2 2.791 7.498 2.791 7.498 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.039 7.097 fft3d_ps 1301 14.7 2.874 2.955 6.691 6.755 calculate_dm_sparse 129 9.5 0.001 0.001 6.488 6.572 mp_alltoall_d11v 2429 14.1 5.818 6.305 5.818 6.305 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.852 5.957 grid_collocate_task_list 129 9.7 5.324 5.735 5.324 5.735 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.813 4.901 potential_pw2rs 129 12.3 0.024 0.024 4.720 4.728 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.551 4.617 mp_sum_l 8016 12.9 3.173 4.475 3.173 4.475 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.431 4.435 calculate_first_density_matrix 1 7.0 0.000 0.000 4.398 4.401 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=202.285000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1076.545455, yerr=17.536267 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.549148E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108580288 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4001 58216. MP_Allreduce 11080 1167. MP_Sync 86 MP_Alltoall 1700 18828153. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.062 172.731 172.736 qs_mol_dyn_low 1 2.0 0.003 0.003 172.244 172.264 qs_forces 11 3.9 0.004 0.004 172.063 172.068 qs_energies 11 4.9 0.006 0.007 164.636 164.643 scf_env_do_scf 11 5.9 0.001 0.001 145.918 145.931 velocity_verlet 10 3.0 0.005 0.006 112.392 112.401 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 109.255 109.256 qs_scf_new_mos 116 7.6 0.001 0.001 72.553 72.659 qs_scf_loop_do_ot 116 8.6 0.001 0.001 72.553 72.658 dbcsr_multiply_generic 2485 12.5 0.180 0.193 72.324 72.584 ot_scf_mini 116 9.6 0.003 0.004 68.120 68.203 multiply_cannon 2485 13.5 0.558 0.591 53.487 56.889 multiply_cannon_loop 2485 14.5 0.808 0.839 50.471 51.073 ot_mini 116 10.6 0.001 0.001 37.234 37.319 init_scf_loop 11 6.9 0.000 0.000 36.505 36.507 prepare_preconditioner 11 7.9 0.000 0.000 32.627 32.668 make_preconditioner 11 8.9 0.000 0.000 32.627 32.668 mp_waitall_1 124680 16.7 24.589 31.054 24.589 31.054 make_full_inverse_cholesky 11 9.9 0.014 0.025 30.446 30.722 rebuild_ks_matrix 127 8.3 0.001 0.001 29.021 29.100 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 29.021 29.099 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.423 26.489 multiply_cannon_multrec 9940 15.5 10.295 14.809 18.342 21.254 qs_ot_get_derivative 116 11.6 0.001 0.002 20.497 20.584 multiply_cannon_metrocomm3 9940 15.5 0.026 0.027 12.355 19.296 cp_fm_cholesky_invert 11 10.9 18.673 18.679 18.673 18.679 qs_ot_get_p 127 10.4 0.002 0.002 18.230 18.332 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 16.792 16.933 apply_single 127 13.6 0.001 0.001 16.792 16.933 ot_diis_step 116 11.6 0.019 0.020 16.665 16.667 make_m2s 4970 13.5 0.066 0.070 14.860 15.730 make_images 4970 14.5 2.144 2.544 14.555 15.430 qs_ot_p2m_diag 82 11.4 0.490 0.496 14.404 14.421 sum_up_and_integrate 127 10.3 0.002 0.002 14.110 14.161 integrate_v_rspace 127 11.3 0.004 0.004 14.051 14.104 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.056 14.085 calculate_rho_elec 127 8.7 0.254 0.265 14.056 14.084 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.233 13.234 init_scf_run 11 5.9 0.000 0.001 11.823 11.824 scf_env_initial_rho_setup 11 6.9 0.002 0.003 11.823 11.823 multiply_cannon_sync_h2d 9940 15.5 10.797 11.265 10.797 11.265 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.280 10.291 cp_fm_diag_elpa_base 82 14.4 10.031 10.114 10.276 10.287 make_images_data 4970 15.5 0.056 0.065 8.571 9.951 hybrid_alltoall_any 5155 16.4 0.834 3.743 8.373 9.909 pw_transfer 1535 11.6 0.084 0.092 9.419 9.445 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 9.198 9.229 dbcsr_mm_accdrv_process 20590 16.1 3.226 4.282 7.697 8.401 cp_fm_cholesky_decompose 22 10.9 8.241 8.358 8.241 8.358 fft_wrap_pw1pw2_140 519 13.2 1.793 1.842 8.115 8.149 grid_integrate_task_list 127 12.3 7.713 8.116 7.713 8.116 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 8.032 8.093 wfi_extrapolate 11 7.9 0.001 0.001 7.566 7.566 multiply_cannon_metrocomm1 9940 15.5 0.030 0.030 4.375 7.433 density_rs2pw 127 9.7 0.005 0.006 6.785 6.982 calculate_dm_sparse 127 9.5 0.001 0.001 6.218 6.283 fft3d_ps 1281 14.7 2.747 2.823 5.949 5.996 mp_alltoall_d11v 2401 14.1 4.856 5.876 4.856 5.876 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.788 5.822 dbcsr_complete_redistribute 393 12.7 2.099 2.185 5.438 5.798 grid_collocate_task_list 127 9.7 5.521 5.777 5.521 5.777 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.280 5.281 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.630 4.690 mp_allgather_i34 2485 14.5 1.209 4.687 1.209 4.687 potential_pw2rs 127 12.3 0.026 0.026 4.296 4.305 calculate_first_density_matrix 1 7.0 0.000 0.000 4.014 4.016 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.535 3.869 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.599 3.868 multiply_cannon_metrocomm4 7455 15.5 0.026 0.028 1.716 3.760 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.636 3.699 mp_irecv_dv 28618 15.9 1.677 3.687 1.677 3.687 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.512 3.610 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.558 3.581 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.505 3.517 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=172.736000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1471.909091, yerr=19.261146 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 11.786061E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 3.244597E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58558. MP_Allreduce 11184 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.064 305.333 305.348 qs_mol_dyn_low 1 2.0 0.003 0.003 304.670 304.688 qs_forces 11 3.9 0.005 0.006 303.920 303.923 qs_energies 11 4.9 0.003 0.003 294.935 294.945 scf_env_do_scf 11 5.9 0.001 0.001 271.040 271.050 velocity_verlet 10 3.0 0.002 0.002 219.932 220.020 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 142.417 142.418 init_scf_loop 11 6.9 0.000 0.000 128.336 128.338 prepare_preconditioner 11 7.9 0.000 0.000 123.408 123.429 make_preconditioner 11 8.9 0.000 0.000 123.408 123.429 make_full_inverse_cholesky 11 9.9 0.038 0.039 98.958 120.513 qs_scf_new_mos 118 7.6 0.001 0.001 97.206 97.269 qs_scf_loop_do_ot 118 8.6 0.001 0.001 97.205 97.268 ot_scf_mini 118 9.6 0.004 0.004 92.125 92.174 dbcsr_multiply_generic 2529 12.6 0.215 0.229 87.313 88.016 cp_fm_upper_to_full 106 14.8 53.408 76.888 53.408 76.888 multiply_cannon 2529 13.6 0.691 0.758 62.092 63.416 multiply_cannon_loop 2529 14.6 1.075 1.083 57.769 59.641 ot_mini 118 10.6 0.001 0.001 47.319 47.382 dbcsr_complete_redistribute 397 12.7 4.009 4.064 30.662 44.348 copy_fm_to_dbcsr 210 11.7 0.002 0.002 27.042 40.712 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.408 37.856 mp_alltoall_i22 720 14.1 22.190 35.956 22.190 35.956 mp_waitall_1 104580 16.8 31.204 35.729 31.204 35.729 cp_fm_cholesky_invert 11 10.9 35.683 35.690 35.683 35.690 rebuild_ks_matrix 129 8.3 0.001 0.001 35.184 35.280 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 35.183 35.280 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.659 32.760 qs_ot_get_p 129 10.4 0.002 0.002 29.160 29.256 qs_ot_get_derivative 118 11.6 0.002 0.002 26.523 26.582 qs_ot_p2m_diag 84 11.4 0.891 0.898 24.568 24.609 multiply_cannon_metrocomm3 10116 15.6 0.026 0.028 20.718 22.725 cp_dbcsr_syevd 84 12.4 0.006 0.006 22.641 22.643 make_m2s 5058 13.6 0.078 0.080 20.073 21.732 make_images 5058 14.6 3.090 3.268 19.586 21.247 ot_diis_step 118 11.6 0.022 0.022 20.748 20.750 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.018 20.225 apply_single 129 13.6 0.001 0.001 20.018 20.225 cp_fm_diag_elpa 84 13.4 0.000 0.000 19.219 19.220 cp_fm_diag_elpa_base 84 14.4 14.649 16.366 19.214 19.215 multiply_cannon_multrec 10116 15.6 10.337 12.162 18.853 19.007 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.312 17.343 calculate_rho_elec 129 8.7 0.484 0.485 17.312 17.343 sum_up_and_integrate 129 10.3 0.002 0.002 16.616 16.707 integrate_v_rspace 129 11.3 0.004 0.005 16.553 16.644 multiply_cannon_sync_h2d 10116 15.6 14.401 14.413 14.401 14.413 make_images_data 5058 15.6 0.065 0.071 11.606 14.137 hybrid_alltoall_any 5245 16.5 1.323 3.101 11.635 14.065 init_scf_run 11 5.9 0.000 0.001 13.511 13.512 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.510 13.512 pw_transfer 1559 11.6 0.095 0.096 12.027 12.036 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 11.787 11.795 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 10.689 10.740 fft_wrap_pw1pw2_140 527 13.2 3.070 3.119 10.409 10.424 dbcsr_mm_accdrv_process 20934 16.1 4.213 6.047 8.261 10.156 cp_fm_cholesky_decompose 22 10.9 9.800 9.820 9.800 9.820 wfi_extrapolate 11 7.9 0.001 0.001 9.336 9.336 grid_integrate_task_list 129 12.3 8.625 8.803 8.625 8.803 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 8.134 8.137 density_rs2pw 129 9.7 0.006 0.006 8.068 8.123 mp_alltoall_d11v 2429 14.1 7.596 7.761 7.596 7.761 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.430 7.489 fft3d_ps 1301 14.7 2.887 2.916 6.991 7.060 calculate_dm_sparse 129 9.5 0.001 0.001 6.979 7.014 grid_collocate_task_list 129 9.7 6.466 6.535 6.466 6.535 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.399 6.511 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=305.348000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2916.363636, yerr=172.998304 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 1.259680E+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 1489051. 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 0.029 0.049 86.244 86.246 qs_energies 1 2.0 0.000 0.000 85.050 85.066 ls_scf 1 3.0 0.000 0.000 84.129 84.144 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.870 73.015 multiply_cannon 111 7.7 0.018 0.020 55.794 56.941 multiply_cannon_loop 111 8.7 0.228 0.241 52.368 53.802 ls_scf_main 1 4.0 0.000 0.000 52.742 52.742 density_matrix_trs4 2 5.0 0.002 0.003 47.146 47.246 ls_scf_init_scf 1 4.0 0.000 0.004 28.336 28.337 ls_scf_init_matrix_S 1 5.0 0.000 0.002 27.157 27.214 mp_waitall_1 11031 10.9 22.699 26.078 22.699 26.078 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.042 25.064 multiply_cannon_multrec 2664 9.7 8.145 8.770 15.539 17.148 multiply_cannon_sync_h2d 2664 9.7 13.512 15.169 13.512 15.169 make_m2s 222 7.7 0.009 0.011 13.409 13.913 make_images 222 8.7 0.098 0.108 13.387 13.892 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.627 12.407 make_images_data 222 9.7 0.004 0.005 7.975 8.571 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.429 8.550 hybrid_alltoall_any 227 10.6 0.216 1.839 6.828 8.154 dbcsr_mm_accdrv_process 4760 10.4 0.588 0.704 7.015 7.970 dbcsr_mm_accdrv_process_sort 4760 11.4 6.227 7.119 6.227 7.119 calculate_norms 4752 9.8 5.531 6.179 5.531 6.179 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.057 5.185 mp_sum_l 887 5.1 3.162 4.894 3.162 4.894 make_images_sizes 222 9.7 0.000 0.000 0.787 3.688 mp_alltoall_i44 222 10.7 0.787 3.688 0.787 3.688 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.325 3.575 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.066 3.548 mp_irecv_dv 6231 10.9 2.050 3.520 2.050 3.520 arnoldi_extremal 4 6.8 0.000 0.000 3.411 3.438 arnoldi_normal_ev 4 7.8 0.001 0.004 3.411 3.438 build_subspace 16 8.4 0.009 0.012 3.318 3.320 ls_scf_post 1 4.0 0.000 0.000 3.050 3.066 ls_scf_store_result 1 5.0 0.000 0.000 2.867 2.904 dbcsr_special_finalize 555 9.7 0.005 0.006 2.297 2.780 dbcsr_merge_single_wm 555 10.7 0.456 0.578 2.289 2.771 make_images_pack 222 9.7 2.203 2.620 2.205 2.622 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.378 2.611 dbcsr_sort_data 658 11.4 2.088 2.515 2.088 2.515 dbcsr_matrix_vector_mult_local 304 10.0 2.061 2.452 2.063 2.454 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.345 2.432 buffer_matrices_ensure_size 222 8.7 1.755 2.079 1.755 2.079 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.814 1.815 rebuild_ks_matrix 3 7.3 0.000 0.000 1.805 1.806 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.011 1.805 1.806 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.246000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 2.145587E+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 0.026 0.040 89.950 89.950 qs_energies 1 2.0 0.000 0.000 89.330 89.334 ls_scf 1 3.0 0.000 0.000 87.999 88.003 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.097 74.412 multiply_cannon 111 7.7 0.028 0.042 52.719 56.854 ls_scf_main 1 4.0 0.000 0.000 54.296 54.300 multiply_cannon_loop 111 8.7 0.136 0.146 49.892 52.825 density_matrix_trs4 2 5.0 0.002 0.003 48.654 48.831 ls_scf_init_scf 1 4.0 0.000 0.000 30.006 30.007 mp_waitall_1 9105 10.9 20.545 29.358 20.545 29.358 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.775 28.859 multiply_cannon_multrec 1332 9.7 13.347 17.455 22.646 27.836 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.401 26.412 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.426 19.774 make_m2s 222 7.7 0.006 0.007 14.981 15.590 make_images 222 8.7 1.370 1.693 14.950 15.560 dbcsr_mm_accdrv_process 4041 10.4 0.346 0.537 8.893 10.499 dbcsr_mm_accdrv_process_sort 4041 11.4 8.412 9.962 8.412 9.962 make_images_data 222 9.7 0.004 0.005 8.625 9.513 hybrid_alltoall_any 227 10.6 0.541 2.527 8.025 9.404 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.213 7.976 mp_irecv_dv 3311 11.0 3.193 7.915 3.193 7.915 mp_sum_l 887 5.1 4.911 7.639 4.911 7.639 calculate_norms 2376 9.8 6.055 6.713 6.055 6.713 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.791 6.131 multiply_cannon_sync_h2d 1332 9.7 4.902 6.123 4.902 6.123 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.867 5.105 arnoldi_extremal 4 6.8 0.000 0.000 4.660 4.674 arnoldi_normal_ev 4 7.8 0.001 0.005 4.660 4.674 build_subspace 16 8.4 0.014 0.021 4.411 4.415 ls_scf_post 1 4.0 0.000 0.000 3.697 3.701 ls_scf_store_result 1 5.0 0.000 0.000 3.377 3.520 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.147 3.380 dbcsr_matrix_vector_mult_local 304 10.0 2.745 3.223 2.747 3.225 mp_allgather_i34 111 8.7 0.824 3.008 0.824 3.008 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.143 2.878 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.619 2.706 dbcsr_data_new 4174 10.1 2.110 2.417 2.110 2.417 make_images_pack 222 9.7 1.824 2.114 1.827 2.117 dbcsr_sort_data 436 11.2 1.837 2.111 1.837 2.111 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.856 1.858 rebuild_ks_matrix 3 7.3 0.000 0.000 1.844 1.846 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.844 1.846 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.950000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1770.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 2.929316E+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 0.026 0.045 93.327 93.328 qs_energies 1 2.0 0.000 0.000 92.694 92.698 ls_scf 1 3.0 0.000 0.000 91.281 91.286 dbcsr_multiply_generic 111 6.7 0.016 0.017 76.006 76.307 ls_scf_main 1 4.0 0.000 0.000 56.989 56.993 multiply_cannon 111 7.7 0.035 0.079 52.472 56.300 multiply_cannon_loop 111 8.7 0.117 0.132 49.717 54.170 density_matrix_trs4 2 5.0 0.002 0.003 51.138 51.349 mp_waitall_1 7281 11.0 23.856 34.415 23.856 34.415 ls_scf_init_scf 1 4.0 0.000 0.000 30.614 30.617 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.435 29.498 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.979 26.996 multiply_cannon_multrec 888 9.7 12.710 15.363 21.312 24.650 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 11.218 24.199 make_m2s 222 7.7 0.006 0.007 16.448 17.180 make_images 222 8.7 1.580 1.856 16.410 17.141 make_images_data 222 9.7 0.004 0.004 9.527 10.679 hybrid_alltoall_any 227 10.6 0.639 2.949 9.083 10.383 dbcsr_mm_accdrv_process 3754 10.4 0.322 0.489 8.129 9.336 mp_sum_l 887 5.1 5.401 8.888 5.401 8.888 dbcsr_mm_accdrv_process_sort 3754 11.4 7.674 8.847 7.674 8.847 multiply_cannon_sync_h2d 888 9.7 6.044 7.611 6.044 7.611 multiply_cannon_metrocomm4 777 9.7 0.004 0.006 2.516 7.171 mp_irecv_dv 2335 11.1 2.500 7.096 2.500 7.096 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.228 7.003 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.756 6.959 arnoldi_extremal 4 6.8 0.000 0.000 5.103 5.126 arnoldi_normal_ev 4 7.8 0.001 0.005 5.103 5.125 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.826 5.055 calculate_norms 1584 9.8 4.406 4.886 4.406 4.886 build_subspace 16 8.4 0.014 0.020 4.803 4.808 mp_allgather_i34 111 8.7 0.866 3.842 0.866 3.842 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.446 3.786 ls_scf_post 1 4.0 0.000 0.000 3.678 3.682 dbcsr_matrix_vector_mult_local 304 10.0 3.030 3.612 3.033 3.614 ls_scf_store_result 1 5.0 0.000 0.000 3.427 3.503 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.814 2.909 dbcsr_data_new 4116 9.9 2.109 2.452 2.109 2.452 make_images_sizes 222 9.7 0.000 0.000 1.077 2.311 mp_alltoall_i44 222 10.7 1.077 2.311 1.077 2.311 dbcsr_sort_data 325 11.1 1.887 2.131 1.887 2.131 dbcsr_finalize 304 7.8 0.026 0.032 1.600 1.894 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.882 1.884 make_images_pack 222 9.7 1.626 1.876 1.629 1.878 rebuild_ks_matrix 3 7.3 0.000 0.000 1.865 1.867 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.865 1.867 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.328000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2242.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 3.368493E+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 0.099 0.211 97.930 97.938 qs_energies 1 2.0 0.000 0.000 97.015 97.025 ls_scf 1 3.0 0.000 0.001 95.248 95.259 dbcsr_multiply_generic 111 6.7 0.030 0.076 78.653 78.945 ls_scf_main 1 4.0 0.000 0.001 58.872 58.877 multiply_cannon 111 7.7 0.044 0.084 51.496 56.267 density_matrix_trs4 2 5.0 0.004 0.027 52.697 52.823 multiply_cannon_loop 111 8.7 0.153 0.167 46.460 49.837 ls_scf_init_scf 1 4.0 0.000 0.001 33.138 33.140 ls_scf_init_matrix_S 1 5.0 0.000 0.002 31.912 31.982 mp_waitall_1 6369 11.0 22.522 29.618 22.522 29.618 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.185 29.200 multiply_cannon_multrec 1332 9.7 14.193 17.631 22.137 25.629 make_m2s 222 7.7 0.007 0.012 21.282 22.652 make_images 222 8.7 3.139 3.599 21.231 22.603 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 8.994 17.455 make_images_data 222 9.7 0.004 0.005 11.858 13.540 hybrid_alltoall_any 227 10.6 0.797 3.772 11.258 13.154 dbcsr_mm_accdrv_process 3641 10.4 0.288 0.488 7.595 9.123 dbcsr_mm_accdrv_process_sort 3641 11.4 7.118 8.592 7.118 8.592 mp_sum_l 887 5.1 4.287 7.492 4.287 7.492 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.092 6.057 multiply_cannon_sync_h2d 1332 9.7 5.425 6.003 5.425 6.003 mp_irecv_dv 3229 10.9 2.066 5.977 2.066 5.977 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.251 5.816 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.500 5.732 arnoldi_extremal 4 6.8 0.000 0.000 5.276 5.302 arnoldi_normal_ev 4 7.8 0.003 0.007 5.276 5.301 build_subspace 16 8.4 0.014 0.021 4.936 4.946 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.660 4.864 calculate_norms 2376 9.8 4.192 4.562 4.192 4.562 mp_allgather_i34 111 8.7 2.150 4.465 2.150 4.465 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.626 3.895 dbcsr_matrix_vector_mult_local 304 10.0 3.203 3.692 3.205 3.694 dbcsr_sort_data 658 11.4 3.117 3.418 3.117 3.418 ls_scf_post 1 4.0 0.002 0.008 3.238 3.248 dbcsr_special_finalize 555 9.7 0.006 0.007 2.866 3.171 dbcsr_merge_single_wm 555 10.7 0.539 0.640 2.858 3.163 ls_scf_dm_to_ks 2 5.0 0.001 0.009 3.014 3.066 ls_scf_store_result 1 5.0 0.000 0.000 2.966 3.035 dbcsr_data_release 10477 10.7 1.568 2.397 1.568 2.397 dbcsr_finalize 304 7.8 0.049 0.061 1.803 1.973 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.938000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2686.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 4.732015E+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 0.035 0.055 92.330 92.331 qs_energies 1 2.0 0.000 0.000 91.564 91.568 ls_scf 1 3.0 0.000 0.000 89.605 89.609 dbcsr_multiply_generic 111 6.7 0.018 0.019 71.010 71.184 ls_scf_main 1 4.0 0.000 0.000 56.669 56.669 multiply_cannon 111 7.7 0.091 0.172 52.592 56.430 multiply_cannon_loop 111 8.7 0.090 0.115 50.025 51.406 density_matrix_trs4 2 5.0 0.002 0.003 49.591 49.672 ls_scf_init_scf 1 4.0 0.000 0.000 29.498 29.500 mp_waitall_1 5436 11.0 24.621 29.314 24.621 29.314 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.260 28.289 matrix_sqrt_Newton_Schulz 2 6.5 0.063 0.069 26.176 26.190 multiply_cannon_multrec 444 9.7 13.659 16.325 20.761 23.190 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.187 15.573 make_m2s 222 7.7 0.005 0.005 13.715 14.637 make_images 222 8.7 2.039 2.471 13.648 14.568 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 5.839 13.468 hybrid_alltoall_any 227 10.6 0.803 3.857 8.205 9.853 make_images_data 222 9.7 0.003 0.004 8.415 9.734 multiply_cannon_sync_h2d 444 9.7 6.692 8.357 6.692 8.357 dbcsr_mm_accdrv_process 3003 10.4 0.344 0.429 6.794 7.892 dbcsr_mm_accdrv_process_sort 3003 11.4 6.423 7.491 6.423 7.491 arnoldi_extremal 4 6.8 0.000 0.000 5.949 5.961 arnoldi_normal_ev 4 7.8 0.005 0.010 5.948 5.961 build_subspace 16 8.4 0.015 0.020 5.542 5.551 mp_sum_l 887 5.1 2.703 4.839 2.703 4.839 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.519 4.775 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.244 4.457 dbcsr_matrix_vector_mult_local 304 10.0 3.735 4.197 3.737 4.200 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.551 3.982 mp_irecv_dv 1241 11.2 1.536 3.960 1.536 3.960 calculate_norms 792 9.8 3.608 3.776 3.608 3.776 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.873 3.756 mp_allgather_i34 111 8.7 1.144 3.665 1.144 3.665 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.564 3.646 ls_scf_post 1 4.0 0.000 0.000 3.438 3.443 make_images_sizes 222 9.7 0.000 0.000 0.914 3.407 mp_alltoall_i44 222 10.7 0.914 3.407 0.914 3.407 ls_scf_store_result 1 5.0 0.000 0.000 3.228 3.264 dbcsr_data_new 4608 9.7 1.776 2.317 1.776 2.317 dbcsr_finalize 304 7.8 0.062 0.077 2.199 2.310 dbcsr_merge_all 275 8.9 0.480 0.527 2.057 2.147 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.027 2.028 rebuild_ks_matrix 3 7.3 0.000 0.000 1.995 1.996 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.995 1.996 qs_energies_init_hamiltonians 1 3.0 0.020 0.038 1.943 1.943 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.331000, 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/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 8.887628E+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 0.084 0.093 105.734 105.734 qs_energies 1 2.0 0.000 0.000 104.324 104.331 ls_scf 1 3.0 0.000 0.000 101.390 101.396 dbcsr_multiply_generic 111 6.7 0.024 0.028 74.994 75.161 ls_scf_main 1 4.0 0.000 0.000 63.966 63.967 density_matrix_trs4 2 5.0 0.002 0.003 55.003 55.068 multiply_cannon 111 7.7 0.108 0.158 48.798 50.954 multiply_cannon_loop 111 8.7 0.097 0.099 45.800 46.294 ls_scf_init_scf 1 4.0 0.000 0.000 33.602 33.603 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.109 32.124 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.395 29.403 mp_waitall_1 4527 11.1 22.124 25.957 22.124 25.957 make_m2s 222 7.7 0.005 0.005 22.599 23.684 make_images 222 8.7 3.576 3.871 22.491 23.575 multiply_cannon_multrec 444 9.7 17.886 18.556 22.523 23.088 hybrid_alltoall_any 227 10.6 1.655 3.640 12.782 15.680 make_images_data 222 9.7 0.003 0.004 12.982 15.131 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.552 11.111 multiply_cannon_sync_h2d 444 9.7 8.801 8.849 8.801 8.849 arnoldi_extremal 4 6.8 0.000 0.000 7.535 7.544 arnoldi_normal_ev 4 7.8 0.003 0.009 7.535 7.543 build_subspace 16 8.4 0.026 0.036 6.978 6.990 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.594 5.754 dbcsr_matrix_vector_mult_local 304 10.0 5.124 5.436 5.126 5.438 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.067 5.160 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.901 5.150 dbcsr_mm_accdrv_process 1814 10.4 0.289 0.350 4.460 4.589 dbcsr_mm_accdrv_process_sort 1814 11.4 4.124 4.259 4.124 4.259 ls_scf_post 1 4.0 0.000 0.000 3.821 3.828 mp_allgather_i34 111 8.7 1.108 3.622 1.108 3.622 make_images_sizes 222 9.7 0.000 0.000 1.475 3.621 mp_alltoall_i44 222 10.7 1.474 3.621 1.474 3.621 ls_scf_store_result 1 5.0 0.000 0.000 3.532 3.570 calculate_norms 792 9.8 3.226 3.265 3.226 3.265 dbcsr_finalize 304 7.8 0.082 0.089 3.071 3.167 dbcsr_merge_all 275 8.9 0.883 0.920 2.857 2.943 qs_energies_init_hamiltonians 1 3.0 0.007 0.011 2.904 2.904 dbcsr_complete_redistribute 5 7.6 1.440 1.477 2.750 2.872 matrix_ls_to_qs 2 6.0 0.000 0.000 2.393 2.525 dbcsr_sort_data 325 11.1 2.439 2.493 2.439 2.493 dbcsr_data_new 6591 9.6 1.892 2.464 1.892 2.464 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.382 2.384 rebuild_ks_matrix 3 7.3 0.000 0.000 2.315 2.318 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.315 2.318 dbcsr_new_transposed 4 7.5 0.242 0.251 2.297 2.314 dbcsr_frobenius_norm 74 6.6 2.058 2.135 2.192 2.229 dbcsr_data_release 12724 10.6 1.977 2.228 1.977 2.228 dbcsr_add_d 103 6.2 0.000 0.000 2.124 2.204 dbcsr_add_anytype 103 7.2 0.858 0.890 2.124 2.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.734000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7031.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954_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 592.932864E+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 4771631. 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 0.022 0.102 233.177 233.203 qs_mol_dyn_low 1 2.0 0.003 0.014 231.859 231.886 qs_forces 5 3.8 0.007 0.104 231.764 231.767 qs_energies 5 4.8 0.036 0.067 228.650 228.680 scf_env_do_scf 5 5.8 0.000 0.002 212.760 212.763 scf_env_do_scf_inner_loop 105 6.6 0.003 0.018 185.058 185.068 qs_scf_new_mos 105 7.6 0.000 0.001 143.842 144.047 qs_scf_loop_do_ot 105 8.6 0.001 0.001 143.842 144.047 dbcsr_multiply_generic 1445 12.2 0.127 0.135 134.598 135.031 ot_scf_mini 105 9.6 0.003 0.004 133.826 133.949 multiply_cannon 1445 13.2 0.276 0.293 115.239 117.435 multiply_cannon_loop 1445 14.2 2.844 3.005 113.332 114.806 velocity_verlet 4 3.0 0.001 0.002 108.352 108.352 ot_mini 105 10.6 0.001 0.001 60.187 60.309 multiply_cannon_multrec 69360 15.2 29.540 34.629 39.396 45.447 mp_waitall_1 488190 16.1 35.600 44.636 35.600 44.636 qs_ot_get_p 112 10.4 0.001 0.001 43.607 43.954 qs_ot_get_derivative 55 11.6 0.001 0.001 38.417 38.572 multiply_cannon_metrocomm3 69360 15.2 0.202 0.213 26.046 34.816 multiply_cannon_sync_h2d 69360 15.2 28.944 33.368 28.944 33.368 qs_ot_p2m_diag 40 11.0 0.020 0.031 32.547 32.666 rebuild_ks_matrix 110 8.4 0.000 0.000 29.702 29.893 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.013 29.701 29.892 cp_dbcsr_syevd 40 12.0 0.002 0.003 29.282 29.283 init_scf_loop 7 6.6 0.000 0.004 27.671 27.673 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.296 27.470 cp_fm_syevd 40 13.0 0.000 0.000 24.095 24.237 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.085 23.282 apply_single 62 13.6 0.000 0.000 23.085 23.282 prepare_preconditioner 7 7.6 0.000 0.000 22.742 22.777 make_preconditioner 7 8.6 0.001 0.026 22.742 22.777 ot_new_cg_direction 55 11.6 0.001 0.001 21.082 21.083 cp_fm_redistribute_end 40 14.0 9.678 19.307 9.684 19.308 cp_fm_syevd_base 40 14.0 9.615 19.242 9.615 19.242 qs_rho_update_rho_low 110 7.6 0.001 0.001 17.435 17.887 calculate_rho_elec 110 8.6 0.030 0.032 17.434 17.886 make_full_inverse_cholesky 7 9.6 0.000 0.000 15.423 15.508 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.783 14.942 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.630 13.722 init_scf_run 5 5.8 0.000 0.001 12.997 12.998 scf_env_initial_rho_setup 5 6.8 0.014 0.021 12.997 12.998 mp_sum_l 4764 12.2 12.020 12.809 12.020 12.809 pw_transfer 1645 12.4 0.079 0.097 12.528 12.701 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 12.389 12.567 density_rs2pw 110 9.6 0.005 0.007 11.580 11.986 calculate_dm_sparse 110 9.5 0.000 0.001 11.559 11.746 dbcsr_mm_accdrv_process 154766 15.8 6.196 6.375 9.726 11.706 fft_wrap_pw1pw2_240 915 15.0 1.134 1.230 10.851 11.018 qs_vxc_create 110 10.4 0.002 0.004 10.952 10.999 cp_fm_cholesky_invert 7 10.6 10.554 10.563 10.554 10.563 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.456 10.521 check_diag 80 13.5 8.685 8.935 9.583 9.714 fft3d_pb 915 16.0 2.395 2.651 9.063 9.304 acc_transpose_blocks 69360 15.2 0.358 0.374 8.187 8.916 multiply_cannon_metrocomm1 69360 15.2 0.095 0.100 4.809 8.865 sum_up_and_integrate 60 10.3 0.001 0.002 8.670 8.682 integrate_v_rspace 60 11.3 0.002 0.002 8.653 8.665 transfer_rs2pw 445 10.6 0.007 0.008 8.086 8.568 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 8.429 8.445 calculate_first_density_matrix 1 7.0 0.000 0.002 8.002 8.040 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.915 7.970 xc_rho_set_and_dset_create 110 12.4 0.076 0.097 7.672 7.923 make_m2s 2890 13.2 0.079 0.088 6.879 7.450 xc_vxc_pw_create 60 11.3 0.039 0.051 7.295 7.340 make_images 2890 14.2 0.240 0.260 6.771 7.335 make_full_single_inverse 7 9.6 0.001 0.002 7.009 7.045 xc_pw_derive 510 13.4 0.005 0.007 6.680 6.748 mp_alltoall_z22v 2340 17.7 6.120 6.413 6.120 6.413 acc_transpose_blocks_kernels 69360 16.2 0.848 0.893 5.401 6.058 jit_kernel_transpose 5 15.0 4.553 5.178 4.553 5.178 mp_waitany 7680 13.5 4.495 5.063 4.495 5.063 potential_pw2rs 60 12.3 0.003 0.003 4.865 4.882 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=233.203000, 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: edf7f2fb6bc475bf1356f57a1cc5c6d8009d0954 Summary: empty Status: OK