=== 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: f67586f46b2d9d1fb2ad49532ceb7ab94a86b150 ################# 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/01 job id: 48660698 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/02 job id: 48660699 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/03 job id: 48660700 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/04 job id: 48660701 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/05 job id: 48660703 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/06 job id: 48660704 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/07 job id: 48660705 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/08 job id: 48660706 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/09 job id: 48660707 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/10 job id: 48660708 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/11 job id: 48660709 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/12 job id: 48660711 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/13 job id: 48660713 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/14 job id: 48660714 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/15 job id: 48660715 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/16 job id: 48660716 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/17 job id: 48660717 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/18 job id: 48660718 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/19 job id: 48660719 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/20 job id: 48660720 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/21 job id: 48660721 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/22 job id: 48660723 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/23 job id: 48660724 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/24 job id: 48660725 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/25 job id: 48660727 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/26 job id: 48660728 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/27 job id: 48660729 --- 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/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 0.019 0.036 137.115 137.115 farming_run 1 2.0 136.427 136.429 137.077 137.081 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.502999E+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 0.010 0.027 115.949 115.949 qs_energies 1 2.0 0.000 0.000 115.637 115.643 mp2_main 1 3.0 0.000 0.000 113.068 113.075 mp2_gpw_main 1 4.0 0.029 0.041 111.790 111.796 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 92.632 92.720 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.574 54.660 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.164 40.957 45.712 get_2c_integrals 1 6.0 0.008 0.009 36.894 37.884 integrate_v_rspace 273 8.0 0.435 0.452 24.556 29.015 pw_transfer 6555 10.6 0.381 0.398 26.736 26.987 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 25.349 25.585 grid_integrate_task_list 273 9.0 20.466 25.338 20.466 25.338 fft_wrap_pw1pw2_100 2178 12.4 1.081 1.162 22.890 23.113 compute_2c_integrals 1 7.0 0.012 0.012 19.145 19.147 rpa_ri_compute_en 1 5.0 0.028 0.031 19.036 19.131 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.569 18.708 mp2_eri_2c_integrate_gpw 1 9.0 2.391 2.423 18.566 18.707 cp_fm_cholesky_decompose 12 8.2 17.833 18.669 17.833 18.669 cholesky_decomp 1 7.0 0.000 0.000 16.588 17.579 fft3d_s 5443 13.4 16.074 16.317 16.096 16.337 ao_to_mo_and_store_B_mult_1 272 7.0 10.778 15.319 10.778 15.319 calculate_wavefunction 272 8.0 5.425 5.516 12.320 13.054 rpa_num_int 1 6.0 0.002 0.025 10.954 10.972 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.750 10.937 calc_mat_Q 8 8.0 0.000 0.000 9.478 9.588 contract_S_to_Q 8 9.0 0.000 0.000 8.902 9.012 calc_potential_gpw 544 9.5 0.005 0.005 8.159 8.681 parallel_gemm_fm 14 9.1 0.000 0.000 8.486 8.621 parallel_gemm_fm_cosma 14 10.1 8.486 8.621 8.486 8.621 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.045 8.306 potential_pw2rs 545 10.0 0.107 0.109 7.528 8.196 create_integ_mat 1 6.0 0.014 0.027 7.894 7.903 collocate_single_gaussian 272 10.0 0.040 0.042 7.327 7.591 array2fm 1 7.0 0.000 0.000 6.632 7.034 pw_scatter_s 2720 13.7 4.324 4.493 4.324 4.493 pw_gather_s 2722 13.2 3.455 3.768 3.455 3.768 array2fm_buffer_send 1 8.0 2.864 3.031 2.864 3.031 pw_poisson_solve 545 10.5 1.123 1.207 2.214 2.364 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.796379, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2804.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 0.024 0.037 424.645 424.649 farming_run 1 2.0 423.838 423.848 424.603 424.609 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.243234E+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 29701377. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 217.392 217.395 qs_energies 1 2.0 0.000 0.000 217.145 217.153 scf_env_do_scf 1 3.0 0.000 0.000 114.998 114.999 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.902 113.911 rebuild_ks_matrix 4 6.0 0.000 0.000 113.901 113.909 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.064 113.901 113.909 hfx_ks_matrix 4 8.0 0.001 0.001 113.504 113.509 integrate_four_center 4 9.0 0.153 0.470 113.503 113.508 integrate_four_center_main 4 10.0 0.136 0.656 102.117 105.182 integrate_four_center_bin 265 11.0 101.980 105.180 101.980 105.180 mp2_main 1 3.0 0.000 0.000 101.822 101.829 mp2_gpw_main 1 4.0 0.049 0.066 100.606 100.617 init_scf_loop 1 4.0 0.000 0.000 97.055 97.056 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 73.954 74.983 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.764 54.793 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.162 41.520 46.385 integrate_v_rspace 95 8.0 0.397 0.562 28.033 32.782 pw_transfer 2240 10.6 0.145 0.182 29.508 30.031 ao_to_mo_and_store_B_mult_1 91 7.0 10.566 29.187 10.566 29.187 fft_wrap_pw1pw2 1868 11.4 0.018 0.024 28.519 29.023 mp2_ri_gpw_compute_en 1 5.0 0.056 0.066 26.467 28.594 grid_integrate_task_list 95 9.0 23.404 28.263 23.404 28.263 fft_wrap_pw1pw2_100 730 12.4 1.278 1.515 26.260 26.723 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.833 1.893 24.799 24.810 get_2c_integrals 1 6.0 0.000 0.000 20.105 20.124 compute_2c_integrals 1 7.0 0.002 0.003 19.087 19.090 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.612 18.924 mp2_eri_2c_integrate_gpw 1 9.0 1.739 1.851 18.611 18.924 fft3d_s 1823 13.4 18.495 18.816 18.509 18.829 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.940 17.940 calculate_wavefunction 91 8.0 2.018 2.047 9.623 9.835 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.590 8.722 9.123 potential_pw2rs 186 10.0 0.033 0.035 8.453 9.094 local_gemm 172 8.0 8.165 8.570 8.165 8.570 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.145 8.527 mp2_ri_gpw_compute_en_comm 22 7.0 0.502 0.523 7.865 8.305 calc_potential_gpw 182 9.5 0.002 0.002 7.913 8.172 collocate_single_gaussian 91 10.0 0.017 0.019 7.741 7.968 integrate_four_center_load 4 10.0 0.000 0.000 6.786 6.789 hfx_load_balance 1 11.0 0.000 0.000 6.786 6.789 mp2_ri_gpw_compute_en_ener 172 7.0 6.334 6.412 6.334 6.412 mp_sendrecv_dm3 2068 8.0 5.886 6.343 5.886 6.343 mp_sync 37 10.5 3.289 6.056 3.289 6.056 pw_gather_s 912 13.2 4.472 5.060 4.472 5.060 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.604281, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1502.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.563520E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.107 58.600 58.602 qs_mol_dyn_low 1 2.0 0.004 0.010 58.181 58.189 qs_forces 11 3.9 0.002 0.003 57.999 58.002 qs_energies 11 4.9 0.012 0.019 56.387 56.401 scf_env_do_scf 11 5.9 0.001 0.002 49.845 49.846 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 47.225 47.226 qs_scf_new_mos 108 7.5 0.000 0.001 35.308 35.568 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.308 35.568 dbcsr_multiply_generic 2286 12.5 0.093 0.098 34.248 34.852 ot_scf_mini 108 9.5 0.002 0.003 33.616 33.800 velocity_verlet 10 3.0 0.001 0.002 29.260 29.262 multiply_cannon 2286 13.5 0.193 0.205 26.374 28.039 multiply_cannon_loop 2286 14.5 1.783 1.896 25.244 27.091 ot_mini 108 10.5 0.001 0.001 19.797 20.039 qs_ot_get_derivative 108 11.5 0.001 0.001 16.807 16.965 mp_waitall_1 245248 16.5 8.976 14.923 8.976 14.923 multiply_cannon_metrocomm3 54864 15.5 0.075 0.081 5.897 13.078 multiply_cannon_multrec 54864 15.5 3.743 5.842 7.573 11.060 qs_ot_get_p 119 10.4 0.001 0.002 9.016 9.287 rebuild_ks_matrix 119 8.3 0.000 0.001 8.833 9.006 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 8.833 9.005 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.786 7.940 mp_sum_l 7287 12.8 5.033 6.787 5.033 6.787 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.128 6.576 multiply_cannon_sync_h2d 54864 15.5 5.146 6.418 5.146 6.418 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.792 5.847 qs_rho_update_rho_low 119 7.7 0.000 0.001 5.373 5.484 calculate_rho_elec 119 8.7 0.011 0.017 5.373 5.483 dbcsr_mm_accdrv_process 76910 16.1 1.726 2.731 3.744 5.370 sum_up_and_integrate 119 10.3 0.001 0.001 5.236 5.247 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.143 5.241 integrate_v_rspace 119 11.3 0.002 0.002 5.225 5.237 init_scf_run 11 5.9 0.000 0.001 5.005 5.005 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.005 5.005 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.951 4.951 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.608 4.608 cp_fm_redistribute_end 50 14.0 2.340 4.563 2.355 4.569 cp_fm_diag_elpa_base 50 14.0 2.200 4.434 2.208 4.450 multiply_cannon_metrocomm1 54864 15.5 0.058 0.064 1.778 3.524 density_rs2pw 119 9.7 0.004 0.004 3.248 3.333 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.982 3.174 apply_single 119 13.6 0.000 0.000 2.981 3.173 calculate_dm_sparse 119 9.5 0.000 0.000 2.933 3.058 ot_diis_step 108 11.5 0.006 0.006 2.733 2.734 pw_transfer 1439 11.6 0.051 0.057 2.564 2.701 make_m2s 4572 13.5 0.054 0.056 2.584 2.698 acc_transpose_blocks 54864 15.5 0.221 0.248 2.123 2.643 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 2.488 2.629 make_images 4572 14.5 0.133 0.139 2.501 2.615 init_scf_loop 11 6.9 0.001 0.004 2.583 2.585 potential_pw2rs 119 12.3 0.004 0.004 2.530 2.579 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.503 2.560 jit_kernel_multiply 13 15.8 1.956 2.550 1.956 2.550 calculate_first_density_matrix 1 7.0 0.000 0.003 2.461 2.465 wfi_extrapolate 11 7.9 0.001 0.001 2.457 2.458 fft3d_ps 1201 14.6 0.374 0.488 2.208 2.348 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.325 2.327 transfer_rs2pw 487 10.6 0.005 0.006 2.018 2.124 grid_integrate_task_list 119 12.3 1.997 2.117 1.997 2.117 fft_wrap_pw1pw2_140 487 13.2 0.136 0.150 1.911 2.051 mp_alltoall_d11v 2130 13.8 1.865 2.025 1.865 2.025 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.914 1.966 mp_sum_d 4135 12.0 1.331 1.888 1.331 1.888 transfer_pw2rs 487 13.2 0.005 0.006 1.796 1.844 mp_waitany 12084 13.8 1.411 1.618 1.411 1.618 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.428 1.455 make_images_sizes 4572 15.5 0.004 0.005 1.156 1.442 mp_alltoall_i44 4572 16.5 1.151 1.438 1.151 1.438 grid_collocate_task_list 119 9.7 1.348 1.410 1.348 1.410 mp_alltoall_z22v 1201 16.6 1.292 1.404 1.292 1.404 prepare_preconditioner 11 7.9 0.000 0.000 1.344 1.376 make_preconditioner 11 8.9 0.000 0.001 1.344 1.376 acc_transpose_blocks_sync 164592 16.5 1.137 1.374 1.137 1.374 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.264 1.304 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=58.602000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.454545, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 489.656320E+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 1879707. 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 0.101 0.298 47.636 47.638 qs_mol_dyn_low 1 2.0 0.003 0.003 47.072 47.254 qs_forces 11 3.9 0.002 0.002 46.120 46.121 qs_energies 11 4.9 0.003 0.006 44.263 44.270 scf_env_do_scf 11 5.9 0.001 0.001 38.236 38.237 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 35.012 35.014 dbcsr_multiply_generic 2286 12.5 0.105 0.120 24.385 25.265 qs_scf_new_mos 108 7.5 0.001 0.001 24.085 24.330 qs_scf_loop_do_ot 108 8.5 0.001 0.001 24.084 24.329 ot_scf_mini 108 9.5 0.003 0.003 23.035 23.207 velocity_verlet 10 3.0 0.001 0.002 23.138 23.158 multiply_cannon 2286 13.5 0.211 0.220 17.566 20.337 multiply_cannon_loop 2286 14.5 1.198 1.261 16.080 19.033 ot_mini 108 10.5 0.001 0.001 14.015 14.261 mp_waitall_1 200699 16.5 7.131 12.628 7.131 12.628 qs_ot_get_derivative 108 11.5 0.001 0.001 11.373 11.545 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.425 10.276 multiply_cannon_multrec 27432 15.5 1.787 4.279 6.133 8.885 rebuild_ks_matrix 119 8.3 0.000 0.000 8.040 8.328 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 8.040 8.327 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.084 7.347 dbcsr_mm_accdrv_process 47894 16.0 3.571 6.049 4.266 6.389 qs_ot_get_p 119 10.4 0.001 0.001 5.843 6.075 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.190 5.112 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.031 5.101 calculate_rho_elec 119 8.7 0.021 0.024 5.030 5.100 mp_sum_l 7287 12.8 2.867 4.962 2.867 4.962 sum_up_and_integrate 119 10.3 0.001 0.001 4.811 4.818 integrate_v_rspace 119 11.3 0.002 0.003 4.797 4.803 init_scf_run 11 5.9 0.000 0.001 4.650 4.650 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.650 4.650 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.167 4.346 apply_single 119 13.6 0.000 0.000 3.167 4.346 make_m2s 4572 13.5 0.052 0.054 3.508 3.842 make_images 4572 14.5 0.207 0.243 3.419 3.751 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.699 3.726 density_rs2pw 119 9.7 0.004 0.004 3.272 3.341 cp_dbcsr_syevd 50 12.0 0.003 0.003 3.252 3.253 init_scf_loop 11 6.9 0.001 0.004 3.189 3.191 qs_ot_get_derivative_diag 49 12.0 0.001 0.004 2.854 2.945 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.765 2.766 cp_fm_redistribute_end 50 14.0 1.394 2.715 1.400 2.718 calculate_first_density_matrix 1 7.0 0.003 0.014 2.702 2.705 cp_fm_diag_elpa_base 50 14.0 1.275 2.590 1.308 2.635 potential_pw2rs 119 12.3 0.006 0.006 2.511 2.533 ot_diis_step 108 11.5 0.011 0.011 2.526 2.527 pw_transfer 1439 11.6 0.065 0.069 2.382 2.433 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.290 2.344 calculate_dm_sparse 119 9.5 0.000 0.001 2.246 2.332 transfer_rs2pw 487 10.6 0.005 0.005 2.216 2.312 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.303 2.306 multiply_cannon_sync_h2d 27432 15.5 1.621 2.271 1.621 2.271 make_images_data 4572 15.5 0.048 0.054 1.542 2.209 prepare_preconditioner 11 7.9 0.000 0.000 2.141 2.167 make_preconditioner 11 8.9 0.000 0.002 2.141 2.167 make_full_inverse_cholesky 11 9.9 0.000 0.001 2.011 2.074 fft3d_ps 1201 14.6 0.520 0.579 1.928 1.979 hybrid_alltoall_any 4725 16.4 0.054 0.117 1.365 1.973 grid_integrate_task_list 119 12.3 1.833 1.952 1.833 1.952 acc_transpose_blocks 27432 15.5 0.116 0.122 1.620 1.951 wfi_extrapolate 11 7.9 0.001 0.001 1.867 1.867 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 1.749 1.833 transfer_pw2rs 487 13.2 0.004 0.005 1.759 1.780 mp_alltoall_d11v 2130 13.8 1.585 1.759 1.585 1.759 fft_wrap_pw1pw2_140 487 13.2 0.158 0.169 1.572 1.628 jit_kernel_multiply 8 16.3 0.637 1.516 0.637 1.516 make_images_sizes 4572 15.5 0.005 0.005 1.076 1.477 mp_alltoall_i44 4572 16.5 1.071 1.472 1.071 1.472 mp_allgather_i34 2286 14.5 0.903 1.458 0.903 1.458 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.374 1.410 mp_sendrecv_dv 16779 12.7 1.351 1.386 1.351 1.386 grid_collocate_task_list 119 9.7 1.293 1.354 1.293 1.354 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.332 1.345 mp_sum_d 4135 12.0 0.824 1.267 0.824 1.267 mp_alltoall_z22v 1201 16.6 1.047 1.155 1.047 1.155 dbcsr_complete_redistribute 329 12.2 0.189 0.312 1.052 1.137 md_output 10 3.0 0.000 0.000 0.026 1.109 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.016 1.040 qs_energies_init_hamiltonians 11 5.9 0.002 0.009 1.037 1.038 mp_sum_dm 438 4.9 0.986 1.034 0.986 1.034 transfer_rs2pw_50 119 11.7 0.185 0.196 0.906 1.032 cp_fm_cholesky_invert 11 10.9 0.979 0.985 0.979 0.985 dbcsr_dot_sd 1205 11.9 0.071 0.087 0.565 0.965 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=47.638000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.818182, yerr=1.526623 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 520.732672E+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 0.015 0.035 40.391 40.392 qs_mol_dyn_low 1 2.0 0.003 0.003 40.113 40.121 qs_forces 11 3.9 0.011 0.073 39.384 39.386 qs_energies 11 4.9 0.003 0.013 37.612 37.620 scf_env_do_scf 11 5.9 0.001 0.002 32.196 32.197 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 28.852 28.853 velocity_verlet 10 3.0 0.001 0.002 19.933 19.955 dbcsr_multiply_generic 2286 12.5 0.098 0.101 19.243 19.676 qs_scf_new_mos 108 7.5 0.001 0.001 18.784 18.804 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.784 18.804 ot_scf_mini 108 9.5 0.002 0.003 17.844 17.864 multiply_cannon 2286 13.5 0.196 0.201 14.466 15.652 multiply_cannon_loop 2286 14.5 0.863 0.916 13.381 14.678 ot_mini 108 10.5 0.001 0.001 10.703 10.746 qs_ot_get_derivative 108 11.5 0.001 0.001 8.959 8.978 rebuild_ks_matrix 119 8.3 0.000 0.000 7.282 7.404 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.282 7.403 multiply_cannon_multrec 18288 15.5 1.845 2.842 6.932 7.291 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.457 6.568 mp_waitall_1 158411 16.6 4.306 6.249 4.306 6.249 dbcsr_mm_accdrv_process 38222 16.0 4.914 5.764 4.996 5.831 sum_up_and_integrate 119 10.3 0.001 0.002 4.751 4.762 integrate_v_rspace 119 11.3 0.003 0.004 4.738 4.749 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.608 4.615 calculate_rho_elec 119 8.7 0.031 0.031 4.608 4.615 qs_ot_get_p 119 10.4 0.001 0.002 4.539 4.583 init_scf_run 11 5.9 0.000 0.001 3.991 3.991 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.991 3.991 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.326 3.900 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.886 3.528 init_scf_loop 11 6.9 0.001 0.003 3.315 3.317 density_rs2pw 119 9.7 0.004 0.004 3.099 3.192 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.899 2.907 make_m2s 4572 13.5 0.044 0.046 2.439 2.625 mp_sum_l 7287 12.8 1.950 2.576 1.950 2.576 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.547 2.549 make_images 4572 14.5 0.194 0.205 2.354 2.539 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.050 2.535 apply_single 119 13.6 0.000 0.000 2.050 2.535 potential_pw2rs 119 12.3 0.007 0.009 2.473 2.516 prepare_preconditioner 11 7.9 0.000 0.000 2.408 2.413 make_preconditioner 11 8.9 0.000 0.001 2.408 2.413 pw_transfer 1439 11.6 0.065 0.068 2.317 2.371 calculate_first_density_matrix 1 7.0 0.000 0.002 2.340 2.341 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.224 2.327 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.224 2.279 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 2.206 2.215 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.178 2.188 cp_fm_diag_elpa_base 50 14.0 2.149 2.164 2.176 2.186 transfer_rs2pw 487 10.6 0.005 0.005 2.040 2.131 calculate_dm_sparse 119 9.5 0.000 0.000 2.041 2.078 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.901 1.904 grid_integrate_task_list 119 12.3 1.792 1.896 1.792 1.896 fft3d_ps 1201 14.6 0.529 0.547 1.796 1.854 transfer_pw2rs 487 13.2 0.004 0.004 1.756 1.793 acc_transpose_blocks 18288 15.5 0.082 0.084 1.637 1.762 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 1.615 1.659 ot_diis_step 108 11.5 0.011 0.012 1.644 1.644 fft_wrap_pw1pw2_140 487 13.2 0.209 0.216 1.563 1.620 wfi_extrapolate 11 7.9 0.001 0.001 1.583 1.583 make_images_data 4572 15.5 0.048 0.053 0.982 1.353 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.262 1.326 grid_collocate_task_list 119 9.7 1.234 1.321 1.234 1.321 mp_sendrecv_dv 11067 12.7 1.204 1.260 1.204 1.260 hybrid_alltoall_any 4725 16.4 0.058 0.120 0.848 1.240 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.550 1.228 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.210 1.223 mp_alltoall_d11v 2130 13.8 1.029 1.209 1.029 1.209 multiply_cannon_sync_h2d 18288 15.5 0.982 1.188 0.982 1.188 cp_fm_cholesky_invert 11 10.9 1.166 1.174 1.166 1.174 mp_alltoall_z22v 1201 16.6 1.060 1.157 1.060 1.157 qs_energies_init_hamiltonians 11 5.9 0.015 0.025 1.047 1.050 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.898 0.913 mp_allgather_i34 2286 14.5 0.533 0.910 0.533 0.910 make_images_sizes 4572 15.5 0.005 0.005 0.705 0.907 mp_alltoall_i44 4572 16.5 0.700 0.902 0.700 0.902 transfer_rs2pw_50 119 11.7 0.155 0.160 0.758 0.886 acc_transpose_blocks_sync 54864 16.5 0.744 0.860 0.744 0.860 mp_sum_dm 438 4.9 0.794 0.830 0.794 0.830 dbcsr_complete_redistribute 329 12.2 0.093 0.105 0.700 0.823 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=40.392000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.272727, yerr=1.212879 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 558.235648E+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 0.029 0.075 43.414 43.415 qs_mol_dyn_low 1 2.0 0.003 0.004 43.014 43.043 qs_forces 11 3.9 0.002 0.002 42.666 42.667 qs_energies 11 4.9 0.005 0.009 40.831 40.838 scf_env_do_scf 11 5.9 0.001 0.001 34.929 34.931 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 30.612 30.612 velocity_verlet 10 3.0 0.002 0.002 22.164 22.178 dbcsr_multiply_generic 2286 12.5 0.099 0.103 21.843 21.994 qs_scf_new_mos 108 7.5 0.001 0.001 20.764 20.841 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.764 20.840 ot_scf_mini 108 9.5 0.002 0.003 19.619 19.673 multiply_cannon 2286 13.5 0.220 0.232 16.795 17.362 multiply_cannon_loop 2286 14.5 1.548 1.634 15.514 16.306 ot_mini 108 10.5 0.001 0.001 11.618 11.688 qs_ot_get_derivative 108 11.5 0.001 0.001 9.575 9.626 multiply_cannon_multrec 27432 15.5 2.452 3.115 9.005 9.279 dbcsr_mm_accdrv_process 47916 15.9 6.142 7.647 6.450 7.800 rebuild_ks_matrix 119 8.3 0.000 0.000 7.358 7.427 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.358 7.427 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.531 6.595 qs_ot_get_p 119 10.4 0.001 0.001 4.838 4.914 mp_waitall_1 137007 16.6 3.238 4.479 3.238 4.479 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.399 4.455 calculate_rho_elec 119 8.7 0.040 0.046 4.398 4.454 sum_up_and_integrate 119 10.3 0.001 0.001 4.388 4.397 integrate_v_rspace 119 11.3 0.003 0.003 4.377 4.386 init_scf_run 11 5.9 0.000 0.001 4.357 4.357 scf_env_initial_rho_setup 11 6.9 0.001 0.003 4.357 4.357 init_scf_loop 11 6.9 0.000 0.002 4.283 4.284 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.548 3.949 make_m2s 4572 13.5 0.054 0.057 3.290 3.593 make_images 4572 14.5 0.274 0.342 3.180 3.482 prepare_preconditioner 11 7.9 0.000 0.000 3.328 3.336 make_preconditioner 11 8.9 0.000 0.001 3.328 3.336 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.895 3.244 qs_ot_p2m_diag 50 11.0 0.018 0.046 3.010 3.020 density_rs2pw 119 9.7 0.004 0.004 2.787 2.864 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.735 2.761 acc_transpose_blocks 27432 15.5 0.122 0.127 2.482 2.607 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.536 2.540 calculate_first_density_matrix 1 7.0 0.001 0.011 2.536 2.539 pw_transfer 1439 11.6 0.065 0.069 2.386 2.429 calculate_dm_sparse 119 9.5 0.000 0.000 2.323 2.380 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.223 2.342 apply_single 119 13.6 0.000 0.000 2.222 2.342 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.294 2.339 potential_pw2rs 119 12.3 0.009 0.009 2.123 2.138 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.126 2.134 cp_fm_diag_elpa_base 50 14.0 2.084 2.104 2.124 2.132 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.904 2.031 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.014 2.016 ot_diis_step 108 11.5 0.012 0.012 1.978 1.979 grid_integrate_task_list 119 12.3 1.823 1.911 1.823 1.911 fft3d_ps 1201 14.6 0.558 0.614 1.813 1.845 transfer_rs2pw 487 10.6 0.004 0.005 1.737 1.842 make_images_data 4572 15.5 0.047 0.051 1.394 1.802 fft_wrap_pw1pw2_140 487 13.2 0.244 0.258 1.742 1.790 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 1.732 1.752 wfi_extrapolate 11 7.9 0.001 0.001 1.751 1.751 mp_sum_l 7287 12.8 1.344 1.703 1.344 1.703 acc_transpose_blocks_sync 82296 16.5 1.467 1.588 1.467 1.588 hybrid_alltoall_any 4725 16.4 0.066 0.156 1.164 1.536 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.486 1.519 transfer_pw2rs 487 13.2 0.004 0.004 1.405 1.419 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.321 1.333 grid_collocate_task_list 119 9.7 1.249 1.323 1.249 1.323 dbcsr_complete_redistribute 329 12.2 0.122 0.155 1.014 1.307 cp_fm_upper_to_full 72 14.2 0.863 1.244 0.863 1.244 mp_alltoall_d11v 2130 13.8 1.100 1.204 1.100 1.204 qs_energies_init_hamiltonians 11 5.9 0.001 0.007 1.191 1.194 cp_fm_cholesky_invert 11 10.9 1.167 1.171 1.167 1.171 mp_alltoall_z22v 1201 16.6 1.137 1.168 1.137 1.168 jit_kernel_multiply 6 16.8 0.241 1.065 0.241 1.065 make_images_sizes 4572 15.5 0.005 0.005 0.738 1.045 mp_alltoall_i44 4572 16.5 0.733 1.040 0.733 1.040 copy_fm_to_dbcsr 176 11.2 0.001 0.002 0.688 0.977 mp_sendrecv_dv 8211 12.7 0.928 0.953 0.928 0.953 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.919 0.926 acc_transpose_blocks_kernels 27432 16.5 0.272 0.279 0.866 0.879 build_core_hamiltonian_matrix_ 11 4.9 0.002 0.013 0.800 0.872 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=43.415000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.909091, yerr=4.187743 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 631.791616E+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 0.010 0.032 36.043 36.044 qs_mol_dyn_low 1 2.0 0.003 0.003 35.759 35.767 qs_forces 11 3.9 0.002 0.003 35.446 35.452 qs_energies 11 4.9 0.002 0.002 33.600 33.608 scf_env_do_scf 11 5.9 0.000 0.001 28.038 28.038 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.664 24.665 velocity_verlet 10 3.0 0.002 0.002 18.682 18.701 dbcsr_multiply_generic 2286 12.5 0.093 0.098 15.776 15.887 qs_scf_new_mos 108 7.5 0.001 0.001 15.180 15.236 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.180 15.235 ot_scf_mini 108 9.5 0.002 0.002 14.317 14.355 multiply_cannon 2286 13.5 0.224 0.228 11.846 12.692 multiply_cannon_loop 2286 14.5 0.647 0.677 10.544 10.848 ot_mini 108 10.5 0.001 0.001 7.962 8.013 rebuild_ks_matrix 119 8.3 0.000 0.000 6.930 6.976 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.930 6.976 qs_ot_get_derivative 108 11.5 0.001 0.001 6.483 6.520 multiply_cannon_multrec 9144 15.5 1.733 1.917 6.273 6.473 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.183 6.224 dbcsr_mm_accdrv_process 12550 15.8 3.710 4.417 4.428 4.508 sum_up_and_integrate 119 10.3 0.001 0.001 4.396 4.400 integrate_v_rspace 119 11.3 0.003 0.003 4.386 4.390 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.293 4.310 calculate_rho_elec 119 8.7 0.060 0.061 4.293 4.309 qs_ot_get_p 119 10.4 0.001 0.001 4.122 4.171 init_scf_run 11 5.9 0.000 0.001 3.970 3.970 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.969 3.970 mp_waitall_1 115863 16.7 3.064 3.919 3.064 3.919 init_scf_loop 11 6.9 0.000 0.000 3.343 3.346 make_m2s 4572 13.5 0.034 0.035 2.739 3.145 make_images 4572 14.5 0.269 0.301 2.650 3.056 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.662 2.666 density_rs2pw 119 9.7 0.003 0.004 2.556 2.630 pw_transfer 1439 11.6 0.065 0.068 2.521 2.535 prepare_preconditioner 11 7.9 0.000 0.000 2.462 2.467 make_preconditioner 11 8.9 0.000 0.000 2.462 2.467 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.428 2.443 calculate_first_density_matrix 1 7.0 0.000 0.000 2.420 2.421 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.349 2.377 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.292 2.347 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.327 2.329 grid_integrate_task_list 119 12.3 1.924 2.058 1.924 2.058 calculate_dm_sparse 119 9.5 0.000 0.000 2.017 2.045 potential_pw2rs 119 12.3 0.010 0.011 2.034 2.041 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.924 1.927 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.914 1.920 cp_fm_diag_elpa_base 50 14.0 1.881 1.900 1.911 1.918 fft3d_ps 1201 14.6 0.565 0.577 1.842 1.856 fft_wrap_pw1pw2_140 487 13.2 0.322 0.333 1.820 1.838 make_images_data 4572 15.5 0.042 0.046 1.301 1.836 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.811 1.827 hybrid_alltoall_any 4725 16.4 0.065 0.175 1.215 1.783 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.651 1.664 acc_transpose_blocks 9144 15.5 0.042 0.043 1.566 1.620 jit_kernel_multiply 8 15.8 0.678 1.569 0.678 1.569 wfi_extrapolate 11 7.9 0.001 0.001 1.486 1.486 transfer_rs2pw 487 10.6 0.004 0.004 1.385 1.459 ot_diis_step 108 11.5 0.013 0.013 1.442 1.442 grid_collocate_task_list 119 9.7 1.349 1.415 1.349 1.415 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.354 1.396 apply_single 119 13.6 0.000 0.000 1.354 1.396 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.291 1.297 cp_fm_cholesky_invert 11 10.9 1.279 1.284 1.279 1.284 mp_alltoall_d11v 2130 13.8 1.194 1.280 1.194 1.280 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.271 1.276 mp_allgather_i34 2286 14.5 0.615 1.275 0.615 1.275 transfer_pw2rs 487 13.2 0.003 0.004 1.248 1.255 mp_alltoall_z22v 1201 16.6 1.149 1.193 1.149 1.193 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 0.386 1.072 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.053 1.072 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.825 1.042 make_images_sizes 4572 15.5 0.005 0.005 0.653 1.020 mp_alltoall_i44 4572 16.5 0.648 1.015 0.648 1.015 mp_sum_l 7287 12.8 0.755 0.937 0.755 0.937 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.874 0.925 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.872 0.877 dbcsr_complete_redistribute 329 12.2 0.166 0.180 0.817 0.873 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 0.784 0.787 acc_transpose_blocks_sync 27432 16.5 0.725 0.779 0.725 0.779 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.718 0.761 mp_sendrecv_dv 5355 12.7 0.703 0.721 0.703 0.721 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=36.044000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=594.272727, yerr=10.963501 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 783.564800E+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 0.013 0.031 47.863 47.864 qs_mol_dyn_low 1 2.0 0.003 0.003 47.660 47.668 qs_forces 11 3.9 0.002 0.002 47.566 47.567 qs_energies 11 4.9 0.002 0.002 45.428 45.431 scf_env_do_scf 11 5.9 0.001 0.001 39.374 39.374 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 30.586 30.587 velocity_verlet 10 3.0 0.002 0.002 27.004 27.014 dbcsr_multiply_generic 2286 12.5 0.101 0.105 20.653 20.792 qs_scf_new_mos 108 7.5 0.001 0.001 19.678 19.793 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.677 19.792 ot_scf_mini 108 9.5 0.002 0.002 18.436 18.550 multiply_cannon 2286 13.5 0.300 0.309 15.529 16.538 multiply_cannon_loop 2286 14.5 0.879 0.893 14.098 14.996 ot_mini 108 10.5 0.001 0.001 10.978 11.105 multiply_cannon_multrec 9144 15.5 3.525 5.010 8.964 9.110 qs_ot_get_derivative 108 11.5 0.001 0.001 8.834 8.945 init_scf_loop 11 6.9 0.000 0.000 8.759 8.762 rebuild_ks_matrix 119 8.3 0.000 0.000 8.055 8.213 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 8.055 8.212 prepare_preconditioner 11 7.9 0.000 0.000 7.655 7.669 make_preconditioner 11 8.9 0.000 0.000 7.655 7.669 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.117 7.525 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.223 7.364 dbcsr_mm_accdrv_process 12550 15.8 4.615 6.595 5.297 6.642 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.932 4.936 calculate_rho_elec 119 8.7 0.118 0.121 4.931 4.935 cp_fm_upper_to_full 72 14.2 3.200 4.670 3.200 4.670 mp_waitall_1 94719 16.7 3.404 4.649 3.404 4.649 sum_up_and_integrate 119 10.3 0.001 0.001 4.491 4.497 integrate_v_rspace 119 11.3 0.003 0.003 4.481 4.487 qs_ot_get_p 119 10.4 0.001 0.001 4.267 4.422 init_scf_run 11 5.9 0.000 0.001 3.931 3.931 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.931 3.931 make_m2s 4572 13.5 0.038 0.038 3.415 3.647 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.147 3.546 make_images 4572 14.5 0.351 0.384 3.295 3.527 pw_transfer 1439 11.6 0.068 0.069 3.234 3.238 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.135 3.140 dbcsr_complete_redistribute 329 12.2 0.283 0.285 2.230 3.066 density_rs2pw 119 9.7 0.004 0.004 2.803 2.814 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.842 2.670 qs_ot_p2m_diag 50 11.0 0.044 0.045 2.580 2.583 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.275 2.572 apply_single 119 13.6 0.000 0.000 2.275 2.571 fft_wrap_pw1pw2_140 487 13.2 0.575 0.581 2.556 2.563 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.372 2.430 mp_alltoall_i22 627 13.8 1.521 2.386 1.521 2.386 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.532 2.351 calculate_dm_sparse 119 9.5 0.000 0.000 2.328 2.340 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.399 2.323 fft3d_ps 1201 14.6 0.599 0.609 2.231 2.236 calculate_first_density_matrix 1 7.0 0.000 0.000 2.180 2.182 grid_integrate_task_list 119 12.3 2.166 2.177 2.166 2.177 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.171 2.172 make_images_data 4572 15.5 0.046 0.048 1.706 2.152 hybrid_alltoall_any 4725 16.4 0.090 0.152 1.660 2.141 ot_diis_step 108 11.5 0.014 0.014 2.115 2.116 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.985 1.987 mp_sum_l 7287 12.8 1.165 1.949 1.165 1.949 acc_transpose_blocks 9144 15.5 0.045 0.046 1.829 1.850 cp_fm_cholesky_invert 11 10.9 1.830 1.834 1.830 1.834 potential_pw2rs 119 12.3 0.014 0.014 1.816 1.820 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.804 1.804 cp_fm_diag_elpa_base 50 14.0 1.646 1.707 1.801 1.801 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.786 1.787 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.680 1.721 wfi_extrapolate 11 7.9 0.001 0.001 1.675 1.675 grid_collocate_task_list 119 9.7 1.581 1.590 1.581 1.590 mp_alltoall_d11v 2130 13.8 1.532 1.552 1.532 1.552 mp_alltoall_z22v 1201 16.6 1.494 1.514 1.494 1.514 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.430 1.453 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.269 1.275 transfer_rs2pw 487 10.6 0.005 0.005 1.150 1.179 acc_transpose_blocks_sync 27432 16.5 1.117 1.136 1.117 1.136 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.118 1.132 make_images_sizes 4572 15.5 0.005 0.005 0.605 1.079 mp_alltoall_i44 4572 16.5 0.601 1.075 0.601 1.075 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.045 1.062 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.993 1.023 yz_to_x 606 15.1 0.058 0.059 0.950 0.960 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=47.864000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=727.818182, yerr=18.516912 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 503.300096E+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 2058142. 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 0.019 0.057 92.167 92.172 qs_mol_dyn_low 1 2.0 0.003 0.004 91.843 91.861 qs_forces 11 3.9 0.003 0.007 91.768 91.773 qs_energies 11 4.9 0.002 0.003 88.700 88.721 scf_env_do_scf 11 5.9 0.000 0.001 78.980 78.983 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 71.948 71.948 dbcsr_multiply_generic 2055 12.4 0.107 0.108 53.584 53.877 qs_scf_new_mos 99 7.5 0.000 0.001 52.307 52.419 qs_scf_loop_do_ot 99 8.5 0.000 0.001 52.306 52.418 ot_scf_mini 99 9.5 0.002 0.003 49.724 49.864 velocity_verlet 10 3.0 0.005 0.009 48.072 48.073 multiply_cannon 2055 13.4 0.185 0.189 43.470 44.444 multiply_cannon_loop 2055 14.4 1.805 1.861 42.123 43.100 ot_mini 99 10.5 0.001 0.001 28.312 28.419 qs_ot_get_derivative 99 11.5 0.001 0.001 21.407 21.544 multiply_cannon_multrec 49320 15.4 11.210 11.903 17.298 17.877 rebuild_ks_matrix 110 8.3 0.000 0.000 16.003 16.170 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.014 16.002 16.170 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.055 14.209 mp_waitall_1 220248 16.4 12.726 13.639 12.726 13.639 qs_ot_get_p 110 10.4 0.001 0.001 12.544 12.696 multiply_cannon_sync_h2d 49320 15.4 9.340 10.048 9.340 10.048 qs_ot_p2m_diag 48 11.0 0.012 0.019 8.634 8.675 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.065 8.616 sum_up_and_integrate 110 10.3 0.002 0.003 8.221 8.233 integrate_v_rspace 110 11.3 0.003 0.003 8.196 8.214 multiply_cannon_metrocomm3 49320 15.4 0.083 0.087 6.747 7.988 qs_rho_update_rho_low 110 7.6 0.000 0.001 7.815 7.969 calculate_rho_elec 110 8.6 0.021 0.024 7.814 7.969 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.282 7.743 apply_single 110 13.6 0.000 0.000 7.282 7.743 cp_dbcsr_syevd 48 12.0 0.002 0.003 7.612 7.612 init_scf_run 11 5.9 0.000 0.001 7.532 7.532 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.532 7.532 init_scf_loop 11 6.9 0.000 0.000 6.996 6.997 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.783 6.805 cp_fm_diag_elpa_base 48 14.0 6.756 6.783 6.780 6.802 ot_diis_step 99 11.5 0.006 0.013 6.677 6.678 dbcsr_mm_accdrv_process 87628 16.1 3.047 3.159 5.959 6.213 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.082 6.146 mp_sum_l 6594 12.7 4.680 5.655 4.680 5.655 make_m2s 4110 13.4 0.061 0.065 5.000 5.118 make_images 4110 14.4 0.178 0.191 4.901 5.022 density_rs2pw 110 9.6 0.004 0.005 4.787 4.959 prepare_preconditioner 11 7.9 0.000 0.000 4.655 4.675 make_preconditioner 11 8.9 0.000 0.000 4.655 4.675 wfi_extrapolate 11 7.9 0.001 0.001 4.510 4.510 pw_transfer 1331 11.6 0.056 0.065 4.387 4.496 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.418 4.484 multiply_cannon_metrocomm1 49320 15.4 0.066 0.068 3.041 4.479 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.297 4.409 calculate_dm_sparse 110 9.5 0.001 0.001 3.965 4.073 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.813 3.818 potential_pw2rs 110 12.3 0.006 0.006 3.751 3.786 fft3d_ps 1111 14.6 0.794 0.891 3.599 3.692 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.529 3.629 fft_wrap_pw1pw2_140 451 13.1 0.384 0.425 3.424 3.538 grid_integrate_task_list 110 12.3 3.242 3.411 3.242 3.411 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.355 3.407 calculate_first_density_matrix 1 7.0 0.000 0.002 2.887 2.894 transfer_rs2pw 451 10.6 0.005 0.006 2.682 2.875 mp_alltoall_d11v 2046 13.8 2.414 2.872 2.414 2.872 jit_kernel_multiply 13 15.9 2.632 2.719 2.632 2.719 acc_transpose_blocks 49320 15.4 0.218 0.227 2.582 2.670 cp_fm_cholesky_invert 11 10.9 2.574 2.579 2.574 2.579 transfer_pw2rs 451 13.1 0.006 0.007 2.532 2.566 mp_alltoall_z22v 1111 16.6 2.337 2.505 2.337 2.505 make_images_data 4110 15.4 0.044 0.047 2.126 2.331 grid_collocate_task_list 110 9.6 2.161 2.289 2.161 2.289 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.082 2.097 mp_waitany 14300 13.8 1.897 2.095 1.897 2.095 hybrid_alltoall_any 4261 16.3 0.084 0.477 1.805 2.086 mp_sum_d 3889 11.9 1.469 1.981 1.469 1.981 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=92.172000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.727273, yerr=2.766386 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 592.293888E+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 1467894. 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 0.032 0.086 79.646 79.648 qs_mol_dyn_low 1 2.0 0.004 0.009 79.246 79.257 qs_forces 11 3.9 0.004 0.013 79.165 79.168 qs_energies 11 4.9 0.003 0.013 75.597 75.611 scf_env_do_scf 11 5.9 0.001 0.001 66.150 66.153 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 56.448 56.449 dbcsr_multiply_generic 2055 12.4 0.118 0.122 41.145 41.796 velocity_verlet 10 3.0 0.002 0.002 41.264 41.265 qs_scf_new_mos 99 7.5 0.001 0.001 38.719 38.828 qs_scf_loop_do_ot 99 8.5 0.001 0.001 38.718 38.827 ot_scf_mini 99 9.5 0.008 0.043 36.882 36.972 multiply_cannon 2055 13.4 0.224 0.244 32.517 33.912 multiply_cannon_loop 2055 14.4 1.173 1.201 30.751 32.819 ot_mini 99 10.5 0.001 0.001 20.936 21.053 multiply_cannon_multrec 24660 15.4 6.840 8.610 14.070 15.670 qs_ot_get_derivative 99 11.5 0.001 0.001 14.909 15.012 rebuild_ks_matrix 110 8.3 0.000 0.000 14.747 14.858 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 14.747 14.858 mp_waitall_1 176588 16.5 9.378 14.060 9.378 14.060 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.979 13.075 multiply_cannon_metrocomm3 24660 15.4 0.073 0.076 5.807 10.163 init_scf_loop 11 6.9 0.001 0.003 9.651 9.652 qs_ot_get_p 110 10.4 0.001 0.001 8.854 9.011 multiply_cannon_sync_h2d 24660 15.4 6.234 7.501 6.234 7.501 prepare_preconditioner 11 7.9 0.000 0.000 7.472 7.493 make_preconditioner 11 8.9 0.000 0.001 7.472 7.493 sum_up_and_integrate 110 10.3 0.002 0.002 7.466 7.479 integrate_v_rspace 110 11.3 0.003 0.017 7.439 7.449 dbcsr_mm_accdrv_process 52282 16.1 5.461 6.145 7.064 7.431 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.573 7.216 apply_single 110 13.6 0.000 0.001 6.573 7.216 make_full_inverse_cholesky 11 9.9 0.000 0.001 7.030 7.203 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.927 6.970 calculate_rho_elec 110 8.6 0.039 0.047 6.926 6.969 init_scf_run 11 5.9 0.000 0.001 6.853 6.854 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.853 6.854 qs_ot_p2m_diag 48 11.0 0.029 0.044 6.343 6.374 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.419 6.131 make_m2s 4110 13.4 0.057 0.059 5.237 5.916 ot_diis_step 99 11.5 0.010 0.010 5.886 5.886 make_images 4110 14.4 0.406 0.458 5.126 5.802 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.754 5.754 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.943 4.955 cp_fm_diag_elpa_base 48 14.0 4.881 4.904 4.940 4.952 pw_transfer 1331 11.6 0.066 0.074 4.197 4.368 density_rs2pw 110 9.6 0.004 0.005 4.109 4.297 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.089 4.264 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 4.151 4.203 wfi_extrapolate 11 7.9 0.001 0.001 3.953 3.953 cp_fm_cholesky_invert 11 10.9 3.641 3.652 3.641 3.652 mp_sum_l 6594 12.7 2.790 3.616 2.790 3.616 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.515 3.518 make_images_data 4110 15.4 0.049 0.053 2.750 3.465 fft3d_ps 1111 14.6 1.112 1.340 3.191 3.373 grid_integrate_task_list 110 12.3 3.139 3.319 3.139 3.319 potential_pw2rs 110 12.3 0.008 0.009 3.274 3.311 hybrid_alltoall_any 4261 16.3 0.107 0.452 2.361 3.296 fft_wrap_pw1pw2_140 451 13.1 0.455 0.469 3.018 3.197 calculate_dm_sparse 110 9.5 0.001 0.001 3.157 3.185 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.071 3.116 calculate_first_density_matrix 1 7.0 0.001 0.006 2.789 2.792 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.319 2.374 mp_alltoall_d11v 2046 13.8 2.140 2.367 2.140 2.367 transfer_rs2pw 451 10.6 0.007 0.007 2.147 2.341 grid_collocate_task_list 110 9.6 2.171 2.335 2.171 2.335 mp_allgather_i34 2055 14.4 1.073 2.199 1.073 2.199 cp_fm_cholesky_decompose 22 10.9 2.140 2.150 2.140 2.150 acc_transpose_blocks 24660 15.4 0.118 0.121 1.989 2.011 transfer_pw2rs 451 13.1 0.005 0.006 1.945 1.976 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.928 1.944 jit_kernel_multiply 9 16.2 1.240 1.923 1.240 1.923 qs_energies_init_hamiltonians 11 5.9 0.002 0.007 1.900 1.903 multiply_cannon_metrocomm4 22605 15.4 0.078 0.083 0.799 1.897 mp_alltoall_z22v 1111 16.6 1.742 1.844 1.742 1.844 dbcsr_complete_redistribute 325 12.2 0.235 0.290 1.416 1.716 mp_irecv_dv 57340 16.2 0.670 1.672 0.670 1.672 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.581 1.594 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=79.648000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=559.818182, yerr=6.952020 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 666.308608E+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 0.020 0.053 70.998 71.002 qs_mol_dyn_low 1 2.0 0.003 0.004 70.691 70.709 qs_forces 11 3.9 0.003 0.003 70.555 70.559 qs_energies 11 4.9 0.002 0.002 67.192 67.206 scf_env_do_scf 11 5.9 0.000 0.001 58.567 58.567 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 47.359 47.360 velocity_verlet 10 3.0 0.002 0.002 38.378 38.381 dbcsr_multiply_generic 2055 12.4 0.109 0.112 32.251 32.979 qs_scf_new_mos 99 7.5 0.001 0.001 31.225 31.393 qs_scf_loop_do_ot 99 8.5 0.001 0.001 31.224 31.393 ot_scf_mini 99 9.5 0.002 0.002 29.752 29.921 multiply_cannon 2055 13.4 0.213 0.222 23.976 26.693 multiply_cannon_loop 2055 14.4 0.827 0.871 22.409 24.460 ot_mini 99 10.5 0.001 0.001 16.310 16.524 rebuild_ks_matrix 110 8.3 0.000 0.000 13.104 13.388 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.104 13.387 mp_waitall_1 139946 16.5 9.021 12.760 9.021 12.760 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.548 11.800 qs_ot_get_derivative 99 11.5 0.001 0.001 11.555 11.735 multiply_cannon_multrec 16440 15.4 3.576 5.058 9.816 11.183 init_scf_loop 11 6.9 0.000 0.000 11.162 11.163 multiply_cannon_metrocomm3 16440 15.4 0.046 0.048 5.302 9.279 prepare_preconditioner 11 7.9 0.000 0.000 9.230 9.264 make_preconditioner 11 8.9 0.000 0.000 9.230 9.264 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.518 8.885 qs_ot_get_p 110 10.4 0.001 0.001 8.068 8.305 sum_up_and_integrate 110 10.3 0.001 0.002 7.091 7.103 integrate_v_rspace 110 11.3 0.003 0.003 7.065 7.079 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.505 6.521 calculate_rho_elec 110 8.6 0.058 0.059 6.504 6.520 dbcsr_mm_accdrv_process 34862 16.1 5.344 5.720 6.094 6.272 init_scf_run 11 5.9 0.000 0.001 6.040 6.040 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.040 6.040 qs_ot_p2m_diag 48 11.0 0.042 0.044 5.721 5.735 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.080 5.675 apply_single 110 13.6 0.000 0.000 5.080 5.675 make_m2s 4110 13.4 0.050 0.052 5.076 5.405 make_images 4110 14.4 0.398 0.518 4.960 5.288 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.231 5.232 ot_diis_step 99 11.5 0.011 0.011 4.649 4.650 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.782 4.454 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.419 4.429 cp_fm_diag_elpa_base 48 14.0 4.330 4.372 4.416 4.426 cp_fm_cholesky_invert 11 10.9 4.144 4.155 4.144 4.155 pw_transfer 1331 11.6 0.065 0.072 3.900 3.916 multiply_cannon_sync_h2d 16440 15.4 3.161 3.887 3.161 3.887 density_rs2pw 110 9.6 0.004 0.005 3.662 3.815 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.793 3.811 mp_sum_l 6594 12.7 2.582 3.606 2.582 3.606 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.340 3.448 wfi_extrapolate 11 7.9 0.001 0.001 3.409 3.409 grid_integrate_task_list 110 12.3 3.171 3.373 3.171 3.373 make_images_data 4110 15.4 0.045 0.050 2.774 3.259 hybrid_alltoall_any 4261 16.3 0.109 0.383 2.461 3.143 fft_wrap_pw1pw2_140 451 13.1 0.577 0.584 3.036 3.056 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.983 2.984 potential_pw2rs 110 12.3 0.011 0.012 2.882 2.893 calculate_dm_sparse 110 9.5 0.001 0.001 2.794 2.826 fft3d_ps 1111 14.6 1.089 1.099 2.736 2.751 multiply_cannon_metrocomm4 14385 15.4 0.048 0.053 0.892 2.612 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.505 2.598 calculate_first_density_matrix 1 7.0 0.000 0.000 2.517 2.518 mp_alltoall_d11v 2046 13.8 2.180 2.511 2.180 2.511 mp_irecv_dv 48980 15.7 0.816 2.474 0.816 2.474 cp_fm_cholesky_decompose 22 10.9 2.413 2.440 2.413 2.440 grid_collocate_task_list 110 9.6 2.217 2.380 2.217 2.380 dbcsr_complete_redistribute 325 12.2 0.323 0.381 1.766 2.279 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.071 2.075 transfer_rs2pw 451 10.6 0.005 0.006 1.830 1.996 mp_allgather_i34 2055 14.4 0.878 1.993 0.878 1.993 cp_fm_upper_to_full 70 14.2 1.453 1.866 1.453 1.866 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.751 1.782 transfer_pw2rs 451 13.1 0.005 0.005 1.763 1.778 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.690 1.737 acc_transpose_blocks 16440 15.4 0.080 0.082 1.603 1.693 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.152 1.646 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.533 1.557 mp_waitany 17072 13.8 1.341 1.542 1.341 1.542 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.383 1.470 mp_alltoall_z22v 1111 16.6 1.433 1.453 1.433 1.453 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=71.002000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=632.545455, yerr=8.478461 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 742.236160E+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 0.050 0.132 74.155 74.161 qs_mol_dyn_low 1 2.0 0.003 0.003 73.788 73.803 qs_forces 11 3.9 0.003 0.003 73.714 73.715 qs_energies 11 4.9 0.026 0.054 70.115 70.120 scf_env_do_scf 11 5.9 0.001 0.001 61.175 61.177 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 47.032 47.034 velocity_verlet 10 3.0 0.013 0.013 41.753 41.755 dbcsr_multiply_generic 2055 12.4 0.114 0.119 32.226 32.406 qs_scf_new_mos 99 7.5 0.001 0.001 31.095 31.194 qs_scf_loop_do_ot 99 8.5 0.001 0.001 31.094 31.194 ot_scf_mini 99 9.5 0.003 0.003 29.323 29.427 multiply_cannon 2055 13.4 0.237 0.252 24.168 25.134 multiply_cannon_loop 2055 14.4 1.437 1.465 22.333 22.810 ot_mini 99 10.5 0.001 0.001 16.084 16.215 multiply_cannon_multrec 24660 15.4 4.081 6.680 13.024 14.272 init_scf_loop 11 6.9 0.001 0.003 14.096 14.098 rebuild_ks_matrix 110 8.3 0.000 0.000 12.783 12.890 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.782 12.890 prepare_preconditioner 11 7.9 0.000 0.000 12.250 12.264 make_preconditioner 11 8.9 0.000 0.001 12.250 12.264 make_full_inverse_cholesky 11 9.9 0.000 0.001 10.388 11.919 qs_ot_get_derivative 99 11.5 0.001 0.001 11.808 11.916 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.260 11.357 dbcsr_mm_accdrv_process 52304 16.0 7.875 9.241 8.785 9.719 qs_ot_get_p 110 10.4 0.001 0.001 7.866 8.018 sum_up_and_integrate 110 10.3 0.001 0.002 6.995 7.008 integrate_v_rspace 110 11.3 0.003 0.003 6.969 6.983 mp_waitall_1 121746 16.5 5.092 6.941 5.092 6.941 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.648 6.661 calculate_rho_elec 110 8.6 0.078 0.081 6.648 6.660 make_m2s 4110 13.4 0.060 0.061 6.133 6.611 make_images 4110 14.4 0.575 0.698 5.990 6.464 init_scf_run 11 5.9 0.000 0.001 6.021 6.021 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.020 6.021 qs_ot_p2m_diag 48 11.0 0.055 0.064 5.637 5.652 cp_fm_upper_to_full 70 14.2 3.474 5.053 3.474 5.053 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.010 5.011 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.268 4.276 cp_fm_diag_elpa_base 48 14.0 4.093 4.162 4.265 4.274 dbcsr_complete_redistribute 325 12.2 0.464 0.527 3.018 4.268 ot_diis_step 99 11.5 0.011 0.011 4.216 4.218 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.097 4.192 apply_single 110 13.6 0.000 0.000 4.096 4.191 cp_fm_cholesky_invert 11 10.9 4.162 4.173 4.162 4.173 pw_transfer 1331 11.6 0.065 0.074 4.080 4.145 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.973 4.042 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.854 3.907 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.852 3.903 density_rs2pw 110 9.6 0.004 0.004 3.714 3.870 make_images_data 4110 15.4 0.048 0.053 2.958 3.618 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.384 3.611 hybrid_alltoall_any 4261 16.3 0.123 0.462 2.511 3.538 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.457 3.489 grid_integrate_task_list 110 12.3 3.268 3.448 3.268 3.448 wfi_extrapolate 11 7.9 0.001 0.001 3.371 3.371 fft_wrap_pw1pw2_140 451 13.1 0.609 0.631 3.211 3.287 calculate_dm_sparse 110 9.5 0.001 0.001 3.089 3.107 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.853 3.064 mp_alltoall_i22 605 13.7 1.751 3.019 1.751 3.019 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.984 2.985 fft3d_ps 1111 14.6 1.084 1.114 2.855 2.903 potential_pw2rs 110 12.3 0.012 0.013 2.789 2.821 multiply_cannon_sync_h2d 24660 15.4 2.367 2.549 2.367 2.549 calculate_first_density_matrix 1 7.0 0.000 0.002 2.526 2.529 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.489 2.524 acc_transpose_blocks 24660 15.4 0.117 0.120 2.452 2.503 mp_alltoall_d11v 2046 13.8 2.073 2.408 2.073 2.408 cp_fm_cholesky_decompose 22 10.9 2.345 2.392 2.345 2.392 grid_collocate_task_list 110 9.6 2.267 2.383 2.267 2.383 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.281 2.282 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.991 2.030 transfer_rs2pw 451 10.6 0.005 0.005 1.786 1.948 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.767 1.779 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.611 1.717 mp_sum_l 6594 12.7 1.274 1.703 1.274 1.703 mp_alltoall_z22v 1111 16.6 1.555 1.640 1.555 1.640 transfer_pw2rs 451 13.1 0.004 0.005 1.602 1.635 mp_allgather_i34 2055 14.4 0.919 1.620 0.919 1.620 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.581 1.594 multiply_cannon_metrocomm4 20550 15.4 0.061 0.065 0.852 1.540 acc_transpose_blocks_sync 73980 16.4 1.434 1.486 1.434 1.486 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=74.161000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=700.909091, yerr=9.690244 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 861.724672E+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 0.025 0.061 61.589 61.593 qs_mol_dyn_low 1 2.0 0.003 0.003 61.227 61.242 qs_forces 11 3.9 0.003 0.004 61.152 61.153 qs_energies 11 4.9 0.002 0.002 57.411 57.417 scf_env_do_scf 11 5.9 0.000 0.001 48.624 48.624 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 39.042 39.043 velocity_verlet 10 3.0 0.015 0.015 34.941 34.944 dbcsr_multiply_generic 2055 12.4 0.106 0.108 24.772 25.104 qs_scf_new_mos 99 7.5 0.001 0.001 23.475 23.530 qs_scf_loop_do_ot 99 8.5 0.001 0.001 23.474 23.530 ot_scf_mini 99 9.5 0.002 0.002 22.129 22.156 multiply_cannon 2055 13.4 0.238 0.249 18.291 20.365 multiply_cannon_loop 2055 14.4 0.607 0.628 16.975 17.925 rebuild_ks_matrix 110 8.3 0.000 0.000 12.214 12.317 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.214 12.317 ot_mini 99 10.5 0.001 0.001 11.895 11.930 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.861 10.958 init_scf_loop 11 6.9 0.000 0.000 9.526 9.528 multiply_cannon_multrec 8220 15.4 3.219 4.430 7.678 8.567 mp_waitall_1 103326 16.6 6.773 8.351 6.773 8.351 qs_ot_get_derivative 99 11.5 0.001 0.001 8.115 8.148 prepare_preconditioner 11 7.9 0.000 0.000 7.789 7.798 make_preconditioner 11 8.9 0.000 0.000 7.789 7.798 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.305 7.422 sum_up_and_integrate 110 10.3 0.002 0.002 6.710 6.723 integrate_v_rspace 110 11.3 0.003 0.003 6.683 6.696 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.572 6.589 calculate_rho_elec 110 8.6 0.114 0.114 6.571 6.589 qs_ot_get_p 110 10.4 0.001 0.001 6.066 6.105 init_scf_run 11 5.9 0.000 0.001 5.544 5.544 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.543 5.543 dbcsr_mm_accdrv_process 17442 15.9 3.098 4.219 4.319 5.284 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 3.177 5.206 make_m2s 4110 13.4 0.038 0.039 4.670 4.920 make_images 4110 14.4 0.647 0.709 4.540 4.790 qs_ot_p2m_diag 48 11.0 0.081 0.085 4.411 4.419 pw_transfer 1331 11.6 0.066 0.072 4.237 4.258 cp_fm_cholesky_invert 11 10.9 4.229 4.237 4.229 4.237 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.129 4.153 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.033 4.034 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.750 3.862 apply_single 110 13.6 0.000 0.000 3.750 3.862 ot_diis_step 99 11.5 0.012 0.012 3.720 3.720 density_rs2pw 110 9.6 0.004 0.004 3.478 3.627 grid_integrate_task_list 110 12.3 3.368 3.517 3.368 3.517 fft_wrap_pw1pw2_140 451 13.1 0.775 0.786 3.478 3.505 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.332 3.339 cp_fm_diag_elpa_base 48 14.0 3.274 3.298 3.330 3.338 wfi_extrapolate 11 7.9 0.001 0.001 3.101 3.101 hybrid_alltoall_any 4261 16.3 0.200 0.853 2.400 2.842 fft3d_ps 1111 14.6 1.136 1.163 2.796 2.815 make_images_data 4110 15.4 0.040 0.046 2.422 2.788 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.696 2.697 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.635 2.637 calculate_dm_sparse 110 9.5 0.001 0.001 2.590 2.630 grid_collocate_task_list 110 9.6 2.362 2.471 2.362 2.471 multiply_cannon_sync_h2d 8220 15.4 2.363 2.468 2.363 2.468 potential_pw2rs 110 12.3 0.015 0.015 2.426 2.444 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.378 2.405 mp_alltoall_d11v 2046 13.8 2.053 2.337 2.053 2.337 calculate_first_density_matrix 1 7.0 0.000 0.000 2.334 2.335 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.242 2.257 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.192 2.224 cp_fm_cholesky_decompose 22 10.9 2.029 2.047 2.029 2.047 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.788 1.991 dbcsr_complete_redistribute 325 12.2 0.554 0.586 1.677 1.781 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.683 1.690 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 1.066 1.682 qs_env_update_s_mstruct 11 6.9 0.012 0.022 1.533 1.641 transfer_rs2pw 451 10.6 0.005 0.005 1.481 1.630 mp_allgather_i34 2055 14.4 0.487 1.619 0.487 1.619 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.472 1.485 mp_alltoall_z22v 1111 16.6 1.422 1.447 1.422 1.447 acc_transpose_blocks 8220 15.4 0.040 0.042 1.402 1.437 jit_kernel_multiply 8 15.5 0.904 1.403 0.904 1.403 mp_sum_l 6594 12.7 1.087 1.402 1.087 1.402 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.320 1.352 mp_waitany 9240 13.8 1.178 1.348 1.178 1.348 qs_create_task_list 11 7.9 0.000 0.000 1.239 1.339 generate_qs_task_list 11 8.9 0.373 0.442 1.239 1.339 transfer_pw2rs 451 13.1 0.004 0.004 1.249 1.271 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=61.593000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=815.818182, yerr=10.928014 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 1.363919E+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 0.015 0.036 94.630 94.633 qs_mol_dyn_low 1 2.0 0.003 0.003 94.287 94.297 qs_forces 11 3.9 0.003 0.003 94.217 94.218 qs_energies 11 4.9 0.002 0.002 89.988 89.992 scf_env_do_scf 11 5.9 0.001 0.001 79.523 79.523 velocity_verlet 10 3.0 0.008 0.009 60.418 60.424 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 48.894 48.895 dbcsr_multiply_generic 2055 12.4 0.120 0.124 31.465 31.663 init_scf_loop 11 6.9 0.000 0.000 30.537 30.538 qs_scf_new_mos 99 7.5 0.001 0.001 30.109 30.154 qs_scf_loop_do_ot 99 8.5 0.001 0.001 30.109 30.153 prepare_preconditioner 11 7.9 0.000 0.000 28.400 28.407 make_preconditioner 11 8.9 0.000 0.000 28.400 28.406 ot_scf_mini 99 9.5 0.002 0.002 28.214 28.266 make_full_inverse_cholesky 11 9.9 0.000 0.000 22.632 27.828 multiply_cannon 2055 13.4 0.331 0.360 23.312 24.803 multiply_cannon_loop 2055 14.4 0.836 0.849 21.275 22.283 cp_fm_upper_to_full 70 14.2 12.559 18.000 12.559 18.000 ot_mini 99 10.5 0.001 0.001 15.584 15.628 rebuild_ks_matrix 110 8.3 0.000 0.001 14.509 14.597 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.509 14.596 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.145 13.223 qs_ot_get_derivative 99 11.5 0.001 0.001 10.777 10.821 dbcsr_complete_redistribute 325 12.2 1.017 1.046 7.604 10.758 mp_waitall_1 84994 16.7 8.745 10.210 8.745 10.210 multiply_cannon_multrec 8220 15.4 4.062 4.214 9.746 9.861 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.395 9.545 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.752 8.879 mp_alltoall_i22 605 13.7 5.392 8.567 5.392 8.567 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.923 7.965 calculate_rho_elec 110 8.6 0.225 0.225 7.923 7.965 sum_up_and_integrate 110 10.3 0.002 0.002 7.455 7.471 integrate_v_rspace 110 11.3 0.003 0.003 7.427 7.443 qs_ot_get_p 110 10.4 0.001 0.001 6.989 7.028 cp_fm_cholesky_invert 11 10.9 6.765 6.769 6.765 6.769 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 5.462 6.513 make_m2s 4110 13.4 0.043 0.045 5.971 6.350 make_images 4110 14.4 0.887 0.948 5.782 6.160 init_scf_run 11 5.9 0.000 0.001 6.095 6.095 scf_env_initial_rho_setup 11 6.9 0.000 0.000 6.094 6.094 dbcsr_mm_accdrv_process 11614 15.7 4.009 4.761 5.536 5.744 pw_transfer 1331 11.6 0.075 0.075 5.661 5.679 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.544 5.562 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.842 5.233 apply_single 110 13.6 0.000 0.000 4.842 5.233 qs_ot_p2m_diag 48 11.0 0.151 0.156 5.137 5.142 ot_diis_step 99 11.5 0.015 0.016 4.738 4.738 fft_wrap_pw1pw2_140 451 13.1 1.275 1.286 4.715 4.731 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.579 4.581 density_rs2pw 110 9.6 0.004 0.004 4.278 4.295 hybrid_alltoall_any 4261 16.3 0.265 0.574 3.074 3.881 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.444 3.877 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.870 3.870 cp_fm_diag_elpa_base 48 14.0 3.313 3.509 3.867 3.867 make_images_data 4110 15.4 0.045 0.048 3.097 3.807 grid_integrate_task_list 110 12.3 3.692 3.758 3.692 3.758 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.690 3.690 wfi_extrapolate 11 7.9 0.001 0.001 3.682 3.682 fft3d_ps 1111 14.6 1.304 1.323 3.600 3.616 calculate_dm_sparse 110 9.5 0.001 0.001 3.296 3.351 multiply_cannon_sync_h2d 8220 15.4 3.110 3.136 3.110 3.136 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.038 3.056 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.017 3.020 potential_pw2rs 110 12.3 0.021 0.021 2.828 2.839 cp_fm_cholesky_decompose 22 10.9 2.712 2.731 2.712 2.731 grid_collocate_task_list 110 9.6 2.679 2.704 2.679 2.704 mp_alltoall_d11v 2046 13.8 2.572 2.632 2.572 2.632 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.423 2.459 calculate_first_density_matrix 1 7.0 0.000 0.000 2.287 2.288 qs_env_update_s_mstruct 11 6.9 0.006 0.011 2.207 2.258 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.096 2.179 mp_alltoall_z22v 1111 16.6 2.042 2.062 2.042 2.062 copy_dbcsr_to_fm 151 11.3 0.003 0.003 2.001 2.027 mp_sum_l 6594 12.7 1.281 2.014 1.281 2.014 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.965 1.975 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.890 1.938 qs_create_task_list 11 7.9 0.000 0.000 1.890 1.937 generate_qs_task_list 11 8.9 0.731 0.785 1.890 1.937 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=94.633000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1236.000000, yerr=54.602697 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 630.882304E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 3104781. 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 0.035 0.064 218.522 218.525 qs_mol_dyn_low 1 2.0 0.005 0.015 217.678 217.692 qs_forces 11 3.9 0.005 0.007 217.547 217.551 qs_energies 11 4.9 0.001 0.002 211.789 211.804 scf_env_do_scf 11 5.9 0.001 0.001 194.265 194.269 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 171.209 171.211 velocity_verlet 10 3.0 0.018 0.047 130.138 130.139 qs_scf_new_mos 117 7.6 0.001 0.001 128.771 129.157 qs_scf_loop_do_ot 117 8.6 0.001 0.001 128.770 129.156 dbcsr_multiply_generic 2507 12.6 0.179 0.185 125.635 126.533 ot_scf_mini 117 9.6 0.003 0.003 122.102 122.496 multiply_cannon 2507 13.6 0.243 0.250 101.740 103.438 multiply_cannon_loop 2507 14.6 2.345 2.479 99.127 100.503 ot_mini 117 10.6 0.001 0.001 66.074 66.456 multiply_cannon_multrec 60168 15.6 32.273 34.198 41.901 43.890 qs_ot_get_derivative 117 11.6 0.001 0.001 40.974 41.362 rebuild_ks_matrix 128 8.3 0.001 0.001 35.136 35.451 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 35.136 35.451 qs_ot_get_p 128 10.4 0.001 0.001 33.712 34.054 mp_waitall_1 267128 16.5 29.230 32.276 29.230 32.276 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.576 31.883 multiply_cannon_sync_h2d 60168 15.6 27.026 29.577 27.026 29.577 qs_ot_p2m_diag 83 11.4 0.079 0.092 26.624 26.739 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.586 25.533 apply_single 128 13.6 0.001 0.001 24.586 25.533 ot_diis_step 117 11.6 0.008 0.008 24.869 24.870 cp_dbcsr_syevd 83 12.4 0.005 0.005 24.095 24.097 init_scf_loop 11 6.9 0.000 0.000 22.965 22.966 cp_fm_diag_elpa 83 13.4 0.000 0.000 20.572 20.607 cp_fm_diag_elpa_base 83 14.4 20.470 20.515 20.568 20.603 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.738 19.039 prepare_preconditioner 11 7.9 0.000 0.000 18.229 18.271 make_preconditioner 11 8.9 0.000 0.000 18.229 18.271 multiply_cannon_metrocomm3 60168 15.6 0.120 0.128 15.154 17.706 make_full_inverse_cholesky 11 9.9 0.000 0.000 17.460 17.636 make_m2s 5014 13.6 0.104 0.110 15.048 15.372 sum_up_and_integrate 128 10.3 0.002 0.004 15.295 15.313 integrate_v_rspace 128 11.3 0.004 0.004 15.236 15.257 make_images 5014 14.6 0.399 0.419 14.865 15.201 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.034 15.151 calculate_rho_elec 128 8.7 0.045 0.064 15.033 15.151 init_scf_run 11 5.9 0.000 0.001 13.120 13.120 scf_env_initial_rho_setup 11 6.9 0.000 0.001 13.120 13.120 cp_fm_cholesky_invert 11 10.9 10.606 10.614 10.606 10.614 dbcsr_mm_accdrv_process 124484 16.2 4.418 4.854 9.174 9.942 wfi_extrapolate 11 7.9 0.001 0.001 9.530 9.530 density_rs2pw 128 9.7 0.006 0.007 8.673 9.331 mp_sum_l 7950 12.9 7.588 9.192 7.588 9.192 pw_transfer 1547 11.6 0.074 0.094 8.626 8.937 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 8.424 8.736 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.444 8.600 calculate_dm_sparse 128 9.5 0.001 0.001 8.486 8.590 make_images_data 5014 15.6 0.067 0.072 7.277 8.233 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.030 8.191 multiply_cannon_metrocomm1 60168 15.6 0.097 0.102 6.173 8.095 hybrid_alltoall_any 5200 16.5 0.294 2.268 6.286 7.546 grid_integrate_task_list 128 12.3 6.974 7.526 6.974 7.526 fft_wrap_pw1pw2_140 523 13.2 1.111 1.163 6.870 7.175 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.083 7.093 fft3d_ps 1291 14.7 2.195 2.849 6.445 6.928 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.212 6.318 mp_alltoall_d11v 2415 14.1 4.695 5.909 4.695 5.909 potential_pw2rs 128 12.3 0.009 0.010 5.705 5.821 cp_fm_cholesky_decompose 22 10.9 5.419 5.431 5.419 5.431 transfer_rs2pw 523 10.6 0.007 0.009 4.363 5.225 grid_collocate_task_list 128 9.7 4.838 5.196 4.838 5.196 mp_sum_d 4465 12.1 3.566 4.372 3.566 4.372 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=218.525000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.545455, yerr=6.257557 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 838.184960E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834560144 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 5197907. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.066 206.216 206.218 qs_mol_dyn_low 1 2.0 0.003 0.004 205.740 205.754 qs_forces 11 3.9 0.005 0.005 205.648 205.650 qs_energies 11 4.9 0.023 0.046 198.723 198.748 scf_env_do_scf 11 5.9 0.001 0.001 181.671 181.681 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 144.724 144.726 velocity_verlet 10 3.0 0.010 0.018 128.926 128.928 qs_scf_new_mos 117 7.6 0.001 0.001 103.827 104.372 qs_scf_loop_do_ot 117 8.6 0.001 0.001 103.826 104.371 dbcsr_multiply_generic 2507 12.6 0.191 0.197 101.019 102.790 ot_scf_mini 117 9.6 0.004 0.004 98.930 99.557 multiply_cannon 2507 13.6 0.504 0.564 78.355 83.409 multiply_cannon_loop 2507 14.6 1.579 1.646 74.466 79.041 ot_mini 117 10.6 0.001 0.001 53.060 53.700 mp_waitall_1 214728 16.6 27.004 44.058 27.004 44.058 init_scf_loop 11 6.9 0.000 0.000 36.847 36.848 multiply_cannon_multrec 30084 15.6 20.498 24.450 31.062 35.518 rebuild_ks_matrix 128 8.3 0.001 0.001 33.685 34.442 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 33.685 34.442 multiply_cannon_metrocomm3 30084 15.6 0.098 0.106 16.995 34.334 prepare_preconditioner 11 7.9 0.000 0.000 32.370 32.454 make_preconditioner 11 8.9 0.000 0.000 32.370 32.454 make_full_inverse_cholesky 11 9.9 0.000 0.000 31.010 31.595 qs_ot_get_derivative 117 11.6 0.001 0.002 30.782 31.392 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.307 30.980 qs_ot_get_p 128 10.4 0.001 0.001 27.337 28.079 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.219 23.643 apply_single 128 13.6 0.001 0.001 22.218 23.642 qs_ot_p2m_diag 83 11.4 0.189 0.217 21.958 22.056 ot_diis_step 117 11.6 0.014 0.015 22.051 22.054 multiply_cannon_sync_h2d 30084 15.6 17.536 20.527 17.536 20.527 cp_dbcsr_syevd 83 12.4 0.006 0.006 20.491 20.493 cp_fm_cholesky_invert 11 10.9 19.031 19.047 19.031 19.047 make_m2s 5014 13.6 0.089 0.093 15.493 17.193 make_images 5014 14.6 1.173 1.380 15.284 16.982 cp_fm_diag_elpa 83 13.4 0.000 0.001 16.779 16.806 cp_fm_diag_elpa_base 83 14.4 16.479 16.619 16.773 16.795 sum_up_and_integrate 128 10.3 0.002 0.004 15.104 15.132 integrate_v_rspace 128 11.3 0.003 0.004 15.043 15.074 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.643 14.691 calculate_rho_elec 128 8.7 0.087 0.104 14.642 14.690 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 12.697 13.159 init_scf_run 11 5.9 0.000 0.001 12.115 12.116 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.114 12.116 multiply_cannon_metrocomm4 27577 15.6 0.105 0.121 3.978 11.246 make_images_data 5014 15.6 0.067 0.076 9.080 11.134 mp_irecv_dv 69486 16.3 3.770 10.836 3.770 10.836 dbcsr_mm_accdrv_process 62242 16.2 5.453 6.090 10.039 10.610 hybrid_alltoall_any 5200 16.5 0.351 1.507 7.566 10.420 pw_transfer 1547 11.6 0.085 0.093 9.063 9.128 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.839 8.909 wfi_extrapolate 11 7.9 0.001 0.001 8.846 8.846 density_rs2pw 128 9.7 0.006 0.007 8.254 8.809 cp_fm_cholesky_decompose 22 10.9 8.377 8.465 8.377 8.465 mp_sum_l 7950 12.9 5.378 8.027 5.378 8.027 grid_integrate_task_list 128 12.3 7.166 7.651 7.166 7.651 fft_wrap_pw1pw2_140 523 13.2 1.213 1.233 7.528 7.609 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.662 7.409 calculate_dm_sparse 128 9.5 0.001 0.001 6.576 6.723 fft3d_ps 1291 14.7 2.842 3.024 6.398 6.449 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.315 6.323 mp_allgather_i34 2507 14.6 2.409 5.945 2.409 5.945 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.688 5.774 mp_alltoall_d11v 2415 14.1 4.526 5.575 4.526 5.575 potential_pw2rs 128 12.3 0.016 0.018 5.532 5.564 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.291 5.472 grid_collocate_task_list 128 9.7 5.008 5.367 5.008 5.367 dbcsr_complete_redistribute 395 12.7 0.807 0.923 3.557 4.479 mp_sum_d 4465 12.1 2.830 4.311 2.830 4.311 transfer_rs2pw 523 10.6 0.007 0.008 3.832 4.300 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=206.218000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=799.000000, yerr=3.162278 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 955.719680E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 58094. MP_Allreduce 11102 999. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.065 190.034 190.035 qs_mol_dyn_low 1 2.0 0.003 0.004 189.550 189.576 qs_forces 11 3.9 0.004 0.004 189.152 189.156 qs_energies 11 4.9 0.006 0.010 182.469 182.482 scf_env_do_scf 11 5.9 0.001 0.001 166.267 166.269 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 126.992 126.995 velocity_verlet 10 3.0 0.002 0.002 121.632 121.644 qs_scf_new_mos 117 7.6 0.001 0.001 88.215 88.570 qs_scf_loop_do_ot 117 8.6 0.001 0.001 88.214 88.569 dbcsr_multiply_generic 2507 12.6 0.181 0.186 84.677 86.357 ot_scf_mini 117 9.6 0.003 0.004 83.855 84.220 multiply_cannon 2507 13.6 0.510 0.534 63.248 67.832 multiply_cannon_loop 2507 14.6 1.135 1.199 59.739 63.672 ot_mini 117 10.6 0.001 0.001 44.603 44.984 init_scf_loop 11 6.9 0.000 0.000 39.159 39.161 mp_waitall_1 170520 16.6 26.543 38.585 26.543 38.585 prepare_preconditioner 11 7.9 0.000 0.000 35.008 35.059 make_preconditioner 11 8.9 0.000 0.000 35.008 35.059 make_full_inverse_cholesky 11 9.9 0.000 0.000 32.499 33.995 rebuild_ks_matrix 128 8.3 0.001 0.001 31.262 31.816 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 31.261 31.815 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.171 28.676 multiply_cannon_metrocomm3 20056 15.6 0.063 0.067 15.862 27.496 multiply_cannon_multrec 20056 15.6 12.975 16.016 22.319 25.450 qs_ot_get_derivative 117 11.6 0.002 0.002 24.813 25.188 qs_ot_get_p 128 10.4 0.001 0.001 24.187 24.619 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.783 20.666 apply_single 128 13.6 0.001 0.001 19.782 20.666 ot_diis_step 117 11.6 0.018 0.018 19.645 19.645 qs_ot_p2m_diag 83 11.4 0.266 0.272 19.536 19.585 cp_dbcsr_syevd 83 12.4 0.005 0.006 18.288 18.290 cp_fm_cholesky_invert 11 10.9 17.074 17.091 17.074 17.091 make_m2s 5014 13.6 0.079 0.083 15.565 16.834 make_images 5014 14.6 1.150 1.246 15.333 16.600 multiply_cannon_sync_h2d 20056 15.6 13.492 14.897 13.492 14.897 sum_up_and_integrate 128 10.3 0.002 0.003 14.781 14.808 integrate_v_rspace 128 11.3 0.003 0.004 14.721 14.744 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.500 14.549 calculate_rho_elec 128 8.7 0.130 0.145 14.499 14.548 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.510 14.531 cp_fm_diag_elpa_base 83 14.4 14.042 14.240 14.506 14.528 make_images_data 5014 15.6 0.062 0.070 9.326 11.093 init_scf_run 11 5.9 0.000 0.001 11.059 11.060 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.059 11.059 hybrid_alltoall_any 5200 16.5 0.451 2.045 8.126 10.426 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.117 10.385 multiply_cannon_metrocomm4 17549 15.6 0.066 0.078 3.522 9.774 mp_irecv_dv 50230 16.2 3.393 9.509 3.393 9.509 pw_transfer 1547 11.6 0.086 0.103 9.069 9.201 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.844 8.982 dbcsr_mm_accdrv_process 41502 16.2 5.626 5.958 8.825 8.973 cp_fm_cholesky_decompose 22 10.9 8.594 8.654 8.594 8.654 density_rs2pw 128 9.7 0.006 0.006 7.805 8.228 wfi_extrapolate 11 7.9 0.001 0.001 7.928 7.928 grid_integrate_task_list 128 12.3 7.290 7.899 7.290 7.899 fft_wrap_pw1pw2_140 523 13.2 1.302 1.332 7.617 7.773 cp_fm_upper_to_full 105 14.8 5.799 7.456 5.799 7.456 dbcsr_complete_redistribute 395 12.7 1.172 1.203 4.952 6.751 fft3d_ps 1291 14.7 2.756 2.991 6.221 6.320 calculate_dm_sparse 128 9.5 0.001 0.001 6.031 6.117 mp_sum_l 7950 12.9 4.049 5.791 4.049 5.791 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.601 5.608 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.910 5.598 mp_alltoall_d11v 2415 14.1 4.686 5.577 4.686 5.577 grid_collocate_task_list 128 9.7 5.156 5.483 5.156 5.483 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.664 5.462 mp_allgather_i34 2507 14.6 2.009 5.267 2.009 5.267 potential_pw2rs 128 12.3 0.020 0.022 5.156 5.199 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.617 4.766 transfer_fm_to_dbcsr 11 9.9 0.018 0.024 2.490 4.252 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.191 4.220 mp_alltoall_i22 716 14.1 2.147 4.071 2.147 4.071 transfer_rs2pw 523 10.6 0.006 0.007 3.406 3.825 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=190.035000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=901.727273, yerr=12.714279 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.166275E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58348. MP_Allreduce 11057 1083. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.087 0.136 200.607 200.608 qs_mol_dyn_low 1 2.0 0.003 0.003 199.798 199.811 qs_forces 11 3.9 0.004 0.004 199.396 199.403 qs_energies 11 4.9 0.007 0.029 192.190 192.201 scf_env_do_scf 11 5.9 0.001 0.001 174.465 174.478 velocity_verlet 10 3.0 0.002 0.003 132.428 132.443 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 124.704 124.705 qs_scf_new_mos 117 7.6 0.001 0.001 86.696 87.050 qs_scf_loop_do_ot 117 8.6 0.001 0.001 86.695 87.049 dbcsr_multiply_generic 2507 12.6 0.187 0.195 82.412 83.369 ot_scf_mini 117 9.6 0.003 0.003 82.092 82.412 multiply_cannon 2507 13.6 0.547 0.583 56.107 58.857 multiply_cannon_loop 2507 14.6 1.859 1.920 52.203 54.580 init_scf_loop 11 6.9 0.001 0.005 49.628 49.629 prepare_preconditioner 11 7.9 0.000 0.001 45.398 45.421 make_preconditioner 11 8.9 0.000 0.001 45.398 45.421 ot_mini 117 10.6 0.001 0.001 44.411 44.732 make_full_inverse_cholesky 11 9.9 0.012 0.020 39.046 44.089 multiply_cannon_multrec 30084 15.6 13.612 19.038 26.484 31.661 rebuild_ks_matrix 128 8.3 0.001 0.001 30.394 30.690 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 30.393 30.690 mp_waitall_1 147882 16.7 18.620 28.441 18.620 28.441 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.459 27.725 qs_ot_get_derivative 117 11.6 0.002 0.002 24.675 25.003 qs_ot_get_p 128 10.4 0.001 0.001 23.583 23.942 make_m2s 5014 13.6 0.096 0.099 21.646 22.913 make_images 5014 14.6 1.958 2.260 21.335 22.602 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.194 19.753 apply_single 128 13.6 0.001 0.001 19.194 19.753 ot_diis_step 117 11.6 0.017 0.018 19.616 19.618 qs_ot_p2m_diag 83 11.4 0.343 0.390 18.999 19.051 cp_fm_cholesky_invert 11 10.9 18.049 18.058 18.049 18.058 cp_dbcsr_syevd 83 12.4 0.005 0.006 17.457 17.458 cp_fm_upper_to_full 105 14.8 11.603 17.070 11.603 17.070 sum_up_and_integrate 128 10.3 0.002 0.003 14.880 14.906 multiply_cannon_metrocomm3 30084 15.6 0.050 0.053 6.271 14.855 integrate_v_rspace 128 11.3 0.003 0.004 14.820 14.847 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.650 14.708 calculate_rho_elec 128 8.7 0.173 0.189 14.649 14.707 cp_fm_diag_elpa 83 13.4 0.000 0.000 13.999 14.014 make_images_data 5014 15.6 0.066 0.071 11.775 14.011 cp_fm_diag_elpa_base 83 14.4 12.979 13.311 13.992 14.005 dbcsr_complete_redistribute 395 12.7 1.498 1.615 9.192 13.003 dbcsr_mm_accdrv_process 62264 16.2 8.436 9.200 12.440 12.960 hybrid_alltoall_any 5200 16.5 0.529 2.204 10.309 12.735 copy_fm_to_dbcsr 209 11.7 0.002 0.003 7.695 11.523 init_scf_run 11 5.9 0.000 0.001 11.251 11.252 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.250 11.252 multiply_cannon_sync_h2d 30084 15.6 10.340 11.151 10.340 11.151 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.461 10.702 transfer_fm_to_dbcsr 11 9.9 0.001 0.007 6.328 10.044 pw_transfer 1547 11.6 0.086 0.102 9.517 9.598 mp_alltoall_i22 716 14.1 5.624 9.418 5.624 9.418 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.290 9.376 cp_fm_cholesky_decompose 22 10.9 8.756 8.854 8.756 8.854 fft_wrap_pw1pw2_140 523 13.2 1.444 1.478 8.086 8.187 density_rs2pw 128 9.7 0.006 0.006 7.570 8.098 wfi_extrapolate 11 7.9 0.001 0.001 8.096 8.097 grid_integrate_task_list 128 12.3 7.539 7.798 7.539 7.798 multiply_cannon_metrocomm4 25070 15.6 0.083 0.094 2.838 6.866 mp_irecv_dv 76098 16.2 2.681 6.586 2.681 6.586 fft3d_ps 1291 14.7 2.850 2.921 6.471 6.520 calculate_dm_sparse 128 9.5 0.001 0.001 6.342 6.432 mp_alltoall_d11v 2415 14.1 5.361 5.745 5.361 5.745 grid_collocate_task_list 128 9.7 5.284 5.682 5.284 5.682 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.608 5.642 potential_pw2rs 128 12.3 0.023 0.023 4.995 5.015 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.743 4.842 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 4.607 4.608 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.422 4.528 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.308 4.368 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=200.608000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1102.363636, yerr=19.434144 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.540694E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532237E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58346. MP_Allreduce 11057 1167. MP_Sync 87 MP_Alltoall 1712 18838186. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.070 0.076 185.982 185.991 qs_mol_dyn_low 1 2.0 0.003 0.003 185.190 185.215 qs_forces 11 3.9 0.004 0.004 185.092 185.095 qs_energies 11 4.9 0.010 0.015 177.640 177.649 scf_env_do_scf 11 5.9 0.001 0.001 159.477 159.489 velocity_verlet 10 3.0 0.005 0.006 121.232 121.237 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 118.991 118.992 qs_scf_new_mos 117 7.6 0.001 0.001 80.353 80.450 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.352 80.449 ot_scf_mini 117 9.6 0.003 0.004 75.866 75.926 dbcsr_multiply_generic 2507 12.6 0.181 0.191 73.306 73.746 multiply_cannon 2507 13.6 0.560 0.586 53.886 58.516 multiply_cannon_loop 2507 14.6 0.811 0.851 50.805 52.232 init_scf_loop 11 6.9 0.000 0.000 40.324 40.326 ot_mini 117 10.6 0.001 0.001 38.901 39.000 prepare_preconditioner 11 7.9 0.000 0.000 36.399 36.427 make_preconditioner 11 8.9 0.000 0.000 36.399 36.427 make_full_inverse_cholesky 11 9.9 0.016 0.025 34.247 34.559 mp_waitall_1 125778 16.7 25.262 31.151 25.262 31.151 rebuild_ks_matrix 128 8.3 0.001 0.001 29.981 30.067 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 29.981 30.067 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.385 27.461 qs_ot_get_p 128 10.4 0.001 0.001 24.140 24.250 multiply_cannon_multrec 10028 15.6 10.415 14.282 18.226 22.172 qs_ot_get_derivative 117 11.6 0.001 0.002 21.903 21.977 cp_fm_cholesky_invert 11 10.9 21.812 21.821 21.812 21.821 qs_ot_p2m_diag 83 11.4 0.496 0.501 20.072 20.101 multiply_cannon_metrocomm3 10028 15.6 0.026 0.027 12.781 19.973 cp_dbcsr_syevd 83 12.4 0.005 0.006 18.787 18.788 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 16.980 17.344 apply_single 128 13.6 0.001 0.001 16.980 17.343 ot_diis_step 117 11.6 0.020 0.021 16.885 16.886 make_m2s 5014 13.6 0.065 0.068 15.011 15.931 make_images 5014 14.6 2.170 2.607 14.707 15.622 cp_fm_diag_elpa 83 13.4 0.000 0.000 15.173 15.184 cp_fm_diag_elpa_base 83 14.4 14.915 15.003 15.164 15.174 sum_up_and_integrate 128 10.3 0.002 0.002 15.114 15.165 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.110 15.134 calculate_rho_elec 128 8.7 0.256 0.267 15.109 15.133 integrate_v_rspace 128 11.3 0.004 0.004 15.054 15.107 multiply_cannon_sync_h2d 10028 15.6 10.824 11.524 10.824 11.524 init_scf_run 11 5.9 0.000 0.001 11.188 11.188 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.188 11.188 pw_transfer 1547 11.6 0.085 0.094 10.020 10.081 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.797 9.864 make_images_data 5014 15.6 0.056 0.064 8.348 9.522 hybrid_alltoall_any 5200 16.5 0.850 3.779 8.180 9.248 cp_fm_cholesky_decompose 22 10.9 8.889 9.015 8.889 9.015 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.725 8.773 fft_wrap_pw1pw2_140 523 13.2 1.803 1.842 8.481 8.553 grid_integrate_task_list 128 12.3 7.847 8.221 7.847 8.221 dbcsr_mm_accdrv_process 20762 16.1 3.462 4.583 7.463 8.162 density_rs2pw 128 9.7 0.005 0.006 7.611 7.884 wfi_extrapolate 11 7.9 0.001 0.001 7.807 7.807 multiply_cannon_metrocomm1 10028 15.6 0.030 0.030 4.372 7.571 fft3d_ps 1291 14.7 2.765 2.846 6.525 6.574 calculate_dm_sparse 128 9.5 0.001 0.001 6.271 6.357 dbcsr_complete_redistribute 395 12.7 2.147 2.215 5.589 5.958 grid_collocate_task_list 128 9.7 5.582 5.803 5.582 5.803 mp_alltoall_d11v 2415 14.1 5.197 5.788 5.197 5.788 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.433 5.458 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.302 5.302 potential_pw2rs 128 12.3 0.026 0.027 5.071 5.112 multiply_cannon_metrocomm4 7521 15.6 0.026 0.029 1.808 4.726 mp_irecv_dv 28860 15.9 1.768 4.638 1.768 4.638 mp_allgather_i34 2507 14.6 1.280 4.463 1.280 4.463 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.162 4.199 copy_fm_to_dbcsr 209 11.7 0.001 0.001 3.641 3.968 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.808 3.887 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.801 3.863 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.584 3.861 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=185.991000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1464.909091, yerr=8.989434 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1979979923456 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.250208E+12 0.0% 0.0% 100.0% flops max/rank 11.788211E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806645728 0.0% 0.0% 100.0% number of processed stacks 1983232 0.0% 0.0% 100.0% average stack size 0.0 0.0 3432.1 marketing flops 145.668111E+12 ------------------------------------------------------------------------------- # multiplications 2537 max memory usage/rank 3.179147E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101480 MPI messages size (bytes): total size 1.145238E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.285357E+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 45968 35701915648 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4050 58461. MP_Allreduce 11202 1500. MP_Sync 87 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 0.025 0.045 316.285 316.286 qs_mol_dyn_low 1 2.0 0.003 0.004 315.635 315.654 qs_forces 11 3.9 0.004 0.004 315.539 315.544 qs_energies 11 4.9 0.002 0.002 306.417 306.423 scf_env_do_scf 11 5.9 0.001 0.001 282.945 282.950 velocity_verlet 10 3.0 0.002 0.002 226.767 226.774 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 150.214 150.216 init_scf_loop 11 6.9 0.000 0.000 132.435 132.437 prepare_preconditioner 11 7.9 0.000 0.000 127.400 127.434 make_preconditioner 11 8.9 0.000 0.000 127.400 127.434 make_full_inverse_cholesky 11 9.9 0.038 0.039 102.654 124.349 qs_scf_new_mos 118 7.6 0.001 0.001 102.760 102.843 qs_scf_loop_do_ot 118 8.6 0.001 0.001 102.759 102.842 ot_scf_mini 118 9.6 0.004 0.004 97.620 97.718 dbcsr_multiply_generic 2537 12.6 0.215 0.227 88.483 89.041 cp_fm_upper_to_full 105 14.8 52.864 76.543 52.864 76.543 multiply_cannon 2537 13.6 0.701 0.758 62.468 64.211 multiply_cannon_loop 2537 14.6 1.077 1.094 57.713 59.485 ot_mini 118 10.6 0.001 0.001 48.966 49.113 dbcsr_complete_redistribute 395 12.7 3.991 4.024 31.215 44.505 copy_fm_to_dbcsr 209 11.7 0.001 0.001 27.384 40.660 cp_fm_cholesky_invert 11 10.9 38.544 38.552 38.544 38.552 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 24.703 37.786 rebuild_ks_matrix 129 8.3 0.001 0.001 36.589 36.747 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 36.588 36.747 mp_waitall_1 104900 16.8 32.198 36.445 32.198 36.445 mp_alltoall_i22 718 14.1 22.345 35.798 22.345 35.798 qs_ks_update_qs_env 129 7.6 0.001 0.001 33.871 34.011 qs_ot_get_p 129 10.4 0.001 0.001 32.736 32.931 qs_ot_get_derivative 118 11.6 0.002 0.002 27.853 27.937 qs_ot_p2m_diag 83 11.4 0.880 0.884 27.839 27.885 cp_dbcsr_syevd 83 12.4 0.006 0.006 25.748 25.750 multiply_cannon_metrocomm3 10148 15.6 0.025 0.026 21.224 23.017 make_m2s 5074 13.6 0.077 0.081 20.547 22.598 make_images 5074 14.6 3.089 3.285 20.061 22.111 cp_fm_diag_elpa 83 13.4 0.000 0.000 21.959 21.960 cp_fm_diag_elpa_base 83 14.4 17.415 19.129 21.953 21.954 ot_diis_step 118 11.6 0.022 0.023 21.043 21.044 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.167 20.562 apply_single 129 13.6 0.001 0.001 20.166 20.562 qs_rho_update_rho_low 129 7.7 0.001 0.001 18.559 18.577 calculate_rho_elec 129 8.7 0.484 0.485 18.558 18.576 multiply_cannon_multrec 10148 15.6 10.461 12.356 18.357 18.424 sum_up_and_integrate 129 10.3 0.002 0.002 17.727 17.814 integrate_v_rspace 129 11.3 0.004 0.004 17.664 17.751 make_images_data 5074 15.6 0.064 0.070 11.943 14.925 hybrid_alltoall_any 5260 16.5 1.324 3.079 11.800 14.727 multiply_cannon_sync_h2d 10148 15.6 14.409 14.466 14.409 14.466 pw_transfer 1559 11.6 0.096 0.096 13.303 13.316 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 13.063 13.076 init_scf_run 11 5.9 0.000 0.001 13.040 13.041 scf_env_initial_rho_setup 11 6.9 0.000 0.001 13.040 13.041 fft_wrap_pw1pw2_140 527 13.2 3.062 3.106 11.473 11.494 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.215 11.290 cp_fm_cholesky_decompose 22 10.9 11.049 11.068 11.049 11.068 wfi_extrapolate 11 7.9 0.001 0.001 9.750 9.750 dbcsr_mm_accdrv_process 20966 16.1 4.260 6.256 7.641 9.587 density_rs2pw 129 9.7 0.005 0.006 8.943 8.994 mp_alltoall_d11v 2423 14.1 8.853 8.968 8.853 8.968 grid_integrate_task_list 129 12.3 8.637 8.824 8.637 8.824 fft3d_ps 1301 14.7 2.876 2.895 8.280 8.330 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.145 8.146 calculate_dm_sparse 129 9.5 0.001 0.001 7.107 7.146 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.607 6.676 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.627 6.667 grid_collocate_task_list 129 9.7 6.475 6.546 6.475 6.546 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=316.286000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2885.454545, yerr=156.522646 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 1.259524E+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 31823980. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.042 85.695 85.696 qs_energies 1 2.0 0.000 0.000 85.081 85.086 ls_scf 1 3.0 0.001 0.008 84.161 84.166 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.422 72.624 multiply_cannon 111 7.7 0.017 0.020 55.878 57.042 multiply_cannon_loop 111 8.7 0.224 0.245 52.465 53.906 ls_scf_main 1 4.0 0.002 0.013 52.309 52.321 density_matrix_trs4 2 5.0 0.002 0.003 46.711 46.803 ls_scf_init_scf 1 4.0 0.000 0.001 28.791 28.795 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.214 27.286 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.093 25.115 mp_waitall_1 11031 10.9 22.222 24.944 22.222 24.944 multiply_cannon_multrec 2664 9.7 8.076 8.800 15.536 17.255 multiply_cannon_sync_h2d 2664 9.7 13.678 15.949 13.678 15.949 make_m2s 222 7.7 0.009 0.011 12.961 13.487 make_images 222 8.7 0.098 0.108 12.939 13.466 multiply_cannon_metrocomm1 2664 9.7 0.010 0.010 9.621 12.481 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.424 9.530 make_images_data 222 9.7 0.004 0.005 7.502 8.054 dbcsr_mm_accdrv_process 4760 10.4 0.585 0.709 7.077 8.047 hybrid_alltoall_any 227 10.6 0.216 1.839 6.505 7.621 dbcsr_mm_accdrv_process_sort 4760 11.4 6.293 7.175 6.293 7.175 calculate_norms 4752 9.8 5.524 6.107 5.524 6.107 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.060 5.239 mp_sum_l 887 5.1 3.073 4.695 3.073 4.695 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.037 3.606 mp_irecv_dv 6231 10.9 2.021 3.582 2.021 3.582 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.257 3.504 arnoldi_extremal 4 6.8 0.000 0.000 3.425 3.457 arnoldi_normal_ev 4 7.8 0.001 0.002 3.425 3.456 make_images_sizes 222 9.7 0.000 0.000 0.687 3.335 mp_alltoall_i44 222 10.7 0.686 3.334 0.686 3.334 build_subspace 16 8.4 0.009 0.012 3.319 3.323 ls_scf_post 1 4.0 0.000 0.001 3.060 3.066 ls_scf_store_result 1 5.0 0.000 0.000 2.888 2.920 dbcsr_special_finalize 555 9.7 0.006 0.006 2.423 2.773 dbcsr_merge_single_wm 555 10.7 0.468 0.595 2.414 2.765 make_images_pack 222 9.7 2.202 2.625 2.204 2.627 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.394 2.622 dbcsr_sort_data 658 11.4 2.204 2.509 2.204 2.509 dbcsr_matrix_vector_mult_local 304 10.0 2.058 2.452 2.060 2.454 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.335 2.449 buffer_matrices_ensure_size 222 8.7 1.772 2.066 1.772 2.066 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.808 1.814 rebuild_ks_matrix 3 7.3 0.000 0.000 1.798 1.805 qs_ks_build_kohn_sham_matrix 3 8.3 0.005 0.033 1.798 1.805 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.696000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 2.186502E+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 0.051 0.093 89.964 89.966 qs_energies 1 2.0 0.000 0.000 89.374 89.381 ls_scf 1 3.0 0.000 0.001 88.031 88.036 dbcsr_multiply_generic 111 6.7 0.016 0.018 73.890 74.153 multiply_cannon 111 7.7 0.028 0.042 52.512 57.467 ls_scf_main 1 4.0 0.000 0.005 54.272 54.276 multiply_cannon_loop 111 8.7 0.135 0.144 49.707 53.486 density_matrix_trs4 2 5.0 0.003 0.014 48.538 48.748 ls_scf_init_scf 1 4.0 0.000 0.001 30.016 30.017 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.827 28.919 mp_waitall_1 9105 10.9 19.787 28.863 19.787 28.863 multiply_cannon_multrec 1332 9.7 13.378 16.897 22.744 27.365 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.448 26.457 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.065 19.992 make_m2s 222 7.7 0.007 0.009 14.530 15.092 make_images 222 8.7 1.373 1.692 14.499 15.063 dbcsr_mm_accdrv_process 4041 10.4 0.365 0.534 8.959 10.554 dbcsr_mm_accdrv_process_sort 4041 11.4 8.479 10.020 8.479 10.020 make_images_data 222 9.7 0.004 0.005 8.149 8.939 hybrid_alltoall_any 227 10.6 0.544 2.547 7.516 8.847 mp_sum_l 887 5.1 5.372 8.420 5.372 8.420 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.176 7.896 mp_irecv_dv 3311 11.0 3.156 7.835 3.156 7.835 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.136 6.894 calculate_norms 2376 9.8 6.082 6.678 6.082 6.678 multiply_cannon_sync_h2d 1332 9.7 4.938 6.076 4.938 6.076 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.939 5.146 arnoldi_extremal 4 6.8 0.000 0.000 4.841 4.853 arnoldi_normal_ev 4 7.8 0.001 0.005 4.841 4.853 build_subspace 16 8.4 0.014 0.021 4.589 4.595 ls_scf_post 1 4.0 0.000 0.001 3.742 3.749 ls_scf_store_result 1 5.0 0.000 0.000 3.434 3.566 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.236 3.462 dbcsr_matrix_vector_mult_local 304 10.0 2.747 3.240 2.749 3.242 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.208 2.939 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.660 2.765 mp_allgather_i34 111 8.7 0.800 2.748 0.800 2.748 dbcsr_data_new 4174 10.1 2.112 2.420 2.112 2.420 make_images_pack 222 9.7 1.824 2.127 1.826 2.129 dbcsr_sort_data 436 11.2 1.829 2.105 1.829 2.105 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.871 1.873 rebuild_ks_matrix 3 7.3 0.000 0.000 1.858 1.860 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.858 1.860 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.966000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1832.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 2.829214E+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 0.032 0.050 91.411 91.412 qs_energies 1 2.0 0.000 0.000 90.821 90.824 ls_scf 1 3.0 0.000 0.001 89.414 89.419 dbcsr_multiply_generic 111 6.7 0.016 0.017 73.852 74.157 ls_scf_main 1 4.0 0.001 0.007 55.647 55.651 multiply_cannon 111 7.7 0.032 0.070 51.451 55.182 multiply_cannon_loop 111 8.7 0.116 0.130 48.731 52.323 density_matrix_trs4 2 5.0 0.002 0.004 49.717 49.865 mp_waitall_1 7281 11.0 22.277 32.145 22.277 32.145 ls_scf_init_scf 1 4.0 0.000 0.001 30.065 30.068 ls_scf_init_matrix_S 1 5.0 0.000 0.002 28.820 28.912 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.495 26.511 multiply_cannon_multrec 888 9.7 12.837 15.468 21.517 24.744 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.201 21.731 make_m2s 222 7.7 0.006 0.007 15.777 16.445 make_images 222 8.7 1.584 1.854 15.738 16.406 make_images_data 222 9.7 0.004 0.004 8.961 9.932 hybrid_alltoall_any 227 10.6 0.642 2.956 8.524 9.468 dbcsr_mm_accdrv_process 3754 10.4 0.311 0.492 8.191 9.441 dbcsr_mm_accdrv_process_sort 3754 11.4 7.757 8.949 7.757 8.949 mp_sum_l 887 5.1 4.939 8.179 4.939 8.179 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.401 7.023 mp_irecv_dv 2335 11.1 2.386 6.985 2.386 6.985 multiply_cannon_sync_h2d 888 9.7 6.008 6.916 6.008 6.916 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.744 6.772 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.737 6.476 arnoldi_extremal 4 6.8 0.000 0.000 5.308 5.331 arnoldi_normal_ev 4 7.8 0.001 0.005 5.308 5.331 build_subspace 16 8.4 0.014 0.020 5.005 5.010 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.745 4.970 calculate_norms 1584 9.8 4.394 4.710 4.394 4.710 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.558 3.890 ls_scf_post 1 4.0 0.000 0.001 3.702 3.705 dbcsr_matrix_vector_mult_local 304 10.0 3.025 3.608 3.027 3.610 ls_scf_store_result 1 5.0 0.000 0.000 3.457 3.525 mp_allgather_i34 111 8.7 0.836 3.459 0.836 3.459 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.862 2.947 dbcsr_data_new 4116 9.9 2.101 2.452 2.101 2.452 dbcsr_sort_data 325 11.1 1.835 2.179 1.835 2.179 make_images_sizes 222 9.7 0.000 0.000 1.023 2.104 mp_alltoall_i44 222 10.7 1.023 2.104 1.023 2.104 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.894 1.896 make_images_pack 222 9.7 1.624 1.885 1.627 1.888 rebuild_ks_matrix 3 7.3 0.000 0.000 1.876 1.877 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.876 1.877 dbcsr_finalize 304 7.8 0.026 0.032 1.602 1.871 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=91.412000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2195.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 3.378188E+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 0.067 0.102 96.595 96.596 qs_energies 1 2.0 0.000 0.000 95.832 95.837 ls_scf 1 3.0 0.000 0.001 94.151 94.156 dbcsr_multiply_generic 111 6.7 0.017 0.023 77.448 77.694 ls_scf_main 1 4.0 0.001 0.008 58.170 58.172 multiply_cannon 111 7.7 0.048 0.121 51.037 55.517 density_matrix_trs4 2 5.0 0.003 0.033 52.043 52.156 multiply_cannon_loop 111 8.7 0.153 0.166 45.924 49.239 ls_scf_init_scf 1 4.0 0.000 0.002 32.727 32.729 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.514 31.584 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.033 29.047 mp_waitall_1 6369 11.0 21.727 28.525 21.727 28.525 multiply_cannon_multrec 1332 9.7 14.193 17.517 22.179 25.333 make_m2s 222 7.7 0.007 0.008 20.893 22.353 make_images 222 8.7 3.144 3.606 20.843 22.305 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 8.532 16.501 make_images_data 222 9.7 0.004 0.005 11.587 13.230 hybrid_alltoall_any 227 10.6 0.799 3.739 10.861 12.688 dbcsr_mm_accdrv_process 3641 10.4 0.294 0.480 7.624 9.177 dbcsr_mm_accdrv_process_sort 3641 11.4 7.185 8.681 7.185 8.681 mp_sum_l 887 5.1 3.926 7.314 3.926 7.314 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.111 6.089 mp_irecv_dv 3229 10.9 2.084 6.006 2.084 6.006 multiply_cannon_sync_h2d 1332 9.7 5.374 5.820 5.374 5.820 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.992 5.749 arnoldi_extremal 4 6.8 0.000 0.000 5.554 5.567 arnoldi_normal_ev 4 7.8 0.001 0.004 5.554 5.567 build_subspace 16 8.4 0.014 0.021 5.217 5.225 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.431 4.832 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.515 4.673 mp_allgather_i34 111 8.7 2.229 4.637 2.229 4.637 calculate_norms 2376 9.8 4.194 4.500 4.194 4.500 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.772 4.003 dbcsr_matrix_vector_mult_local 304 10.0 3.199 3.694 3.201 3.695 dbcsr_sort_data 658 11.4 3.056 3.297 3.056 3.297 ls_scf_post 1 4.0 0.000 0.000 3.255 3.260 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.042 3.089 dbcsr_special_finalize 555 9.7 0.006 0.007 2.797 3.069 dbcsr_merge_single_wm 555 10.7 0.538 0.660 2.788 3.060 ls_scf_store_result 1 5.0 0.000 0.000 2.960 3.012 dbcsr_data_release 10477 10.7 1.589 2.449 1.589 2.449 dbcsr_finalize 304 7.8 0.049 0.061 1.806 1.983 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.596000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2730.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 4.757070E+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 0.061 0.097 91.085 91.085 qs_energies 1 2.0 0.000 0.000 90.054 90.058 ls_scf 1 3.0 0.000 0.000 88.120 88.124 dbcsr_multiply_generic 111 6.7 0.017 0.018 69.091 69.287 ls_scf_main 1 4.0 0.000 0.000 55.577 55.577 multiply_cannon 111 7.7 0.074 0.170 51.579 54.818 multiply_cannon_loop 111 8.7 0.089 0.101 49.028 50.488 density_matrix_trs4 2 5.0 0.002 0.003 48.575 48.658 ls_scf_init_scf 1 4.0 0.000 0.000 29.171 29.174 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.901 27.930 mp_waitall_1 5436 11.0 22.687 26.727 22.687 26.727 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.819 25.831 multiply_cannon_multrec 444 9.7 13.669 16.353 20.831 22.541 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.399 14.611 make_m2s 222 7.7 0.005 0.005 12.927 13.834 make_images 222 8.7 2.038 2.473 12.859 13.768 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 5.455 12.516 make_images_data 222 9.7 0.003 0.004 7.649 9.084 hybrid_alltoall_any 227 10.6 0.805 3.819 7.567 9.049 multiply_cannon_sync_h2d 444 9.7 6.789 8.267 6.789 8.267 dbcsr_mm_accdrv_process 3003 10.4 0.343 0.436 6.852 7.976 dbcsr_mm_accdrv_process_sort 3003 11.4 6.481 7.583 6.481 7.583 arnoldi_extremal 4 6.8 0.000 0.000 6.304 6.318 arnoldi_normal_ev 4 7.8 0.001 0.004 6.304 6.318 build_subspace 16 8.4 0.014 0.019 5.900 5.908 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.401 4.653 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.392 4.574 mp_sum_l 887 5.1 2.618 4.418 2.618 4.418 dbcsr_matrix_vector_mult_local 304 10.0 3.713 4.192 3.715 4.195 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.505 3.869 mp_irecv_dv 1241 11.2 1.490 3.846 1.490 3.846 calculate_norms 792 9.8 3.654 3.790 3.654 3.790 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.573 3.638 ls_scf_post 1 4.0 0.000 0.000 3.372 3.376 mp_allgather_i34 111 8.7 1.136 3.373 1.136 3.373 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.815 3.329 ls_scf_store_result 1 5.0 0.000 0.000 3.161 3.194 make_images_sizes 222 9.7 0.000 0.000 0.895 2.950 mp_alltoall_i44 222 10.7 0.895 2.950 0.895 2.950 dbcsr_finalize 304 7.8 0.062 0.077 2.198 2.304 dbcsr_data_new 4608 9.7 1.796 2.292 1.796 2.292 dbcsr_merge_all 275 8.9 0.479 0.524 2.056 2.144 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.077 2.078 rebuild_ks_matrix 3 7.3 0.000 0.000 2.045 2.046 qs_ks_build_kohn_sham_matrix 3 8.3 0.018 0.028 2.045 2.045 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 1.918 1.918 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.085000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3764.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 8.900481E+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 0.081 0.108 106.215 106.215 qs_energies 1 2.0 0.000 0.000 104.810 104.817 ls_scf 1 3.0 0.000 0.000 101.896 101.903 dbcsr_multiply_generic 111 6.7 0.024 0.027 75.134 75.304 ls_scf_main 1 4.0 0.000 0.000 64.143 64.144 density_matrix_trs4 2 5.0 0.002 0.003 55.091 55.152 multiply_cannon 111 7.7 0.165 0.213 48.634 50.764 multiply_cannon_loop 111 8.7 0.099 0.101 45.579 46.270 ls_scf_init_scf 1 4.0 0.000 0.000 33.957 33.958 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.402 32.432 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.686 29.695 mp_waitall_1 4527 11.1 22.060 25.891 22.060 25.891 make_m2s 222 7.7 0.005 0.005 22.766 23.896 make_images 222 8.7 3.574 3.862 22.659 23.787 multiply_cannon_multrec 444 9.7 17.837 18.526 22.530 23.256 hybrid_alltoall_any 227 10.6 1.654 3.627 12.895 15.803 make_images_data 222 9.7 0.003 0.004 13.151 15.351 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.301 10.596 multiply_cannon_sync_h2d 444 9.7 8.796 8.843 8.796 8.843 arnoldi_extremal 4 6.8 0.000 0.000 7.718 7.729 arnoldi_normal_ev 4 7.8 0.002 0.008 7.718 7.729 build_subspace 16 8.4 0.026 0.036 7.127 7.139 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.661 5.715 dbcsr_matrix_vector_mult_local 304 10.0 5.083 5.389 5.086 5.392 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.184 5.281 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.855 5.129 dbcsr_mm_accdrv_process 1814 10.4 0.291 0.362 4.513 4.642 dbcsr_mm_accdrv_process_sort 1814 11.4 4.176 4.306 4.176 4.306 ls_scf_post 1 4.0 0.000 0.000 3.795 3.802 make_images_sizes 222 9.7 0.000 0.000 1.484 3.701 mp_alltoall_i44 222 10.7 1.484 3.700 1.484 3.700 mp_allgather_i34 111 8.7 1.174 3.648 1.174 3.648 ls_scf_store_result 1 5.0 0.000 0.000 3.515 3.554 calculate_norms 792 9.8 3.226 3.267 3.226 3.267 dbcsr_finalize 304 7.8 0.082 0.089 3.065 3.128 dbcsr_merge_all 275 8.9 0.883 0.916 2.850 2.907 dbcsr_complete_redistribute 5 7.6 1.432 1.466 2.765 2.893 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.884 2.884 matrix_ls_to_qs 2 6.0 0.000 0.000 2.393 2.556 dbcsr_sort_data 325 11.1 2.437 2.504 2.437 2.504 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.474 2.477 rebuild_ks_matrix 3 7.3 0.000 0.000 2.409 2.411 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.004 2.409 2.411 dbcsr_new_transposed 4 7.5 0.243 0.255 2.297 2.313 dbcsr_data_new 6591 9.6 1.821 2.297 1.821 2.297 dbcsr_frobenius_norm 74 6.6 2.055 2.131 2.204 2.262 dbcsr_add_d 103 6.2 0.000 0.000 2.124 2.203 dbcsr_add_anytype 103 7.2 0.859 0.890 2.124 2.202 dbcsr_data_release 12724 10.6 1.970 2.162 1.970 2.162 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.215000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6987.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f67586f46b2d9d1fb2ad49532ceb7ab94a86b150_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 592.965632E+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 3454652. 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 0.054 0.140 274.529 274.531 qs_mol_dyn_low 1 2.0 0.004 0.037 273.215 273.277 qs_forces 5 3.8 0.005 0.021 273.114 273.117 qs_energies 5 4.8 0.003 0.027 269.664 269.685 scf_env_do_scf 5 5.8 0.000 0.002 253.462 253.467 scf_env_do_scf_inner_loop 105 6.6 0.015 0.053 218.876 218.919 qs_scf_new_mos 105 7.6 0.000 0.001 170.688 170.937 qs_scf_loop_do_ot 105 8.6 0.001 0.001 170.687 170.937 ot_scf_mini 105 9.6 0.003 0.004 160.653 160.880 dbcsr_multiply_generic 1445 12.2 0.125 0.131 138.878 139.964 multiply_cannon 1445 13.2 0.278 0.295 117.086 120.414 velocity_verlet 4 3.0 0.003 0.009 119.550 119.551 multiply_cannon_loop 1445 14.2 2.846 3.002 114.575 117.621 qs_ot_get_p 112 10.4 0.001 0.001 66.071 66.377 ot_mini 105 10.6 0.001 0.001 63.769 63.877 qs_ot_p2m_diag 40 11.0 0.020 0.031 53.643 53.750 mp_waitall_1 488190 16.1 39.884 49.998 39.884 49.998 cp_dbcsr_syevd 40 12.0 0.002 0.002 49.958 49.961 multiply_cannon_multrec 69360 15.2 29.387 34.556 39.290 44.814 cp_fm_syevd 40 13.0 0.000 0.001 43.892 44.033 qs_ot_get_derivative 55 11.6 0.001 0.001 41.905 42.085 multiply_cannon_metrocomm3 69360 15.2 0.201 0.214 27.578 39.851 cp_fm_redistribute_end 40 14.0 19.254 38.397 19.264 38.404 cp_fm_syevd_base 40 14.0 19.127 38.270 19.127 38.270 init_scf_loop 7 6.6 0.002 0.030 34.515 34.519 multiply_cannon_sync_h2d 69360 15.2 28.647 33.386 28.647 33.386 rebuild_ks_matrix 110 8.4 0.000 0.000 32.107 32.292 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.014 32.107 32.292 qs_ks_update_qs_env 112 7.6 0.001 0.001 29.423 29.591 prepare_preconditioner 7 7.6 0.000 0.000 29.053 29.089 make_preconditioner 7 8.6 0.002 0.043 29.053 29.089 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.134 23.444 apply_single 62 13.6 0.000 0.000 23.134 23.443 qs_rho_update_rho_low 110 7.6 0.001 0.001 22.743 23.076 calculate_rho_elec 110 8.6 0.030 0.032 22.743 23.076 ot_new_cg_direction 55 11.6 0.001 0.001 21.120 21.121 make_full_inverse_cholesky 7 9.6 0.001 0.008 20.732 20.797 density_rs2pw 110 9.6 0.005 0.007 16.295 16.697 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.501 15.810 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 15.185 15.321 mp_sum_l 4764 12.2 13.142 14.284 13.142 14.284 cp_fm_cholesky_invert 7 10.6 14.190 14.213 14.190 14.213 init_scf_run 5 5.8 0.000 0.001 13.282 13.283 scf_env_initial_rho_setup 5 6.8 0.000 0.001 13.281 13.283 transfer_rs2pw 445 10.6 0.007 0.008 12.538 13.017 pw_transfer 1645 12.4 0.080 0.105 12.610 12.857 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 12.471 12.721 calculate_dm_sparse 110 9.5 0.000 0.000 11.668 11.847 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 11.284 11.356 sum_up_and_integrate 60 10.3 0.001 0.002 11.191 11.201 integrate_v_rspace 60 11.3 0.002 0.014 11.174 11.184 check_diag 80 13.5 8.684 8.937 10.883 11.023 qs_vxc_create 110 10.4 0.002 0.004 10.821 10.888 fft_wrap_pw1pw2_240 915 15.0 1.162 1.257 10.667 10.887 multiply_cannon_metrocomm1 69360 15.2 0.098 0.104 5.472 10.831 dbcsr_mm_accdrv_process 154766 15.8 6.138 6.359 9.772 10.674 fft3d_pb 915 16.0 2.400 2.829 8.823 9.274 make_m2s 2890 13.2 0.079 0.085 8.244 8.759 make_images 2890 14.2 0.241 0.301 8.138 8.650 acc_transpose_blocks 69360 15.2 0.358 0.376 7.621 8.448 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 8.152 8.170 make_full_single_inverse 7 9.6 0.002 0.029 7.893 7.929 calculate_first_density_matrix 1 7.0 0.000 0.005 7.816 7.847 xc_rho_set_and_dset_create 110 12.4 0.077 0.106 7.456 7.703 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.001 7.484 7.555 xc_vxc_pw_create 60 11.3 0.039 0.049 7.281 7.347 potential_pw2rs 60 12.3 0.003 0.003 7.102 7.140 transfer_rs2pw_30 110 11.6 1.268 1.326 6.451 6.974 mp_sendrecv_dv 168740 12.6 6.627 6.829 6.627 6.829 xc_pw_derive 510 13.4 0.006 0.007 6.485 6.554 mp_alltoall_z22v 2340 17.7 5.886 6.347 5.886 6.347 transfer_pw2rs 245 13.2 0.003 0.004 5.911 5.950 cp_fm_cholesky_decompose 14 10.2 5.909 5.918 5.909 5.918 mp_alltoall_d11v 1300 13.8 4.884 5.654 4.884 5.654 acc_transpose_blocks_kernels 69360 16.2 0.849 0.893 4.860 5.599 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=274.531000, 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: f67586f46b2d9d1fb2ad49532ceb7ab94a86b150 Summary: empty Status: OK