=== 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: dfc565023ea3187a044d14b931fefe7b58c72354 ################# 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 2.9.0, 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: 13.09.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.12.0 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/01 job id: 49063605 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/02 job id: 49063606 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/03 job id: 49063607 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/04 job id: 49063608 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/05 job id: 49063609 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/06 job id: 49063610 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/07 job id: 49063615 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/08 job id: 49063616 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/09 job id: 49063617 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/10 job id: 49063619 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/11 job id: 49063620 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/12 job id: 49063624 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/13 job id: 49063625 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/14 job id: 49063626 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/15 job id: 49063627 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/16 job id: 49063628 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/17 job id: 49063629 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/18 job id: 49063630 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/19 job id: 49063632 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/20 job id: 49063634 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/21 job id: 49063635 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/22 job id: 49063636 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/23 job id: 49063637 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/24 job id: 49063639 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/25 job id: 49063641 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/26 job id: 49063643 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/27 job id: 49063644 --- 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/dfc565023ea3187a044d14b931fefe7b58c72354_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 0.035 133.517 133.518 farming_run 1 2.0 132.927 132.928 133.483 133.486 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.504940E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.025 114.666 114.666 qs_energies 1 2.0 0.000 0.000 114.471 114.473 mp2_main 1 3.0 0.000 0.000 112.375 112.377 mp2_gpw_main 1 4.0 0.027 0.034 111.376 111.378 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 92.441 92.736 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.004 54.604 54.900 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.166 40.999 46.006 get_2c_integrals 1 6.0 0.008 0.009 37.060 37.664 integrate_v_rspace 273 8.0 0.439 0.454 24.585 29.458 pw_transfer 6555 10.6 0.374 0.389 26.783 27.428 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 25.393 25.831 grid_integrate_task_list 273 9.0 20.499 25.819 20.499 25.819 fft_wrap_pw1pw2_100 2178 12.4 1.127 1.365 22.959 23.436 compute_2c_integrals 1 7.0 0.002 0.003 19.456 19.457 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.620 19.182 mp2_eri_2c_integrate_gpw 1 9.0 2.374 2.427 18.618 19.180 rpa_ri_compute_en 1 5.0 0.020 0.022 18.818 18.974 cp_fm_cholesky_decompose 12 8.2 17.537 18.146 17.537 18.146 cholesky_decomp 1 7.0 0.000 0.000 16.456 17.065 fft3d_s 5443 13.4 16.136 16.380 16.159 16.404 ao_to_mo_and_store_B_mult_1 272 7.0 10.771 15.308 10.771 15.308 calculate_wavefunction 272 8.0 5.419 5.569 12.323 12.987 rpa_num_int 1 6.0 0.000 0.001 10.627 10.628 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.610 10.621 calc_mat_Q 8 8.0 0.000 0.000 9.485 9.571 contract_S_to_Q 8 9.0 0.000 0.000 8.904 8.989 calc_potential_gpw 544 9.5 0.006 0.007 8.225 8.758 parallel_gemm_fm 14 9.1 0.000 0.000 8.493 8.590 parallel_gemm_fm_cosma 14 10.1 8.493 8.590 8.493 8.590 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.087 8.455 potential_pw2rs 545 10.0 0.107 0.109 7.495 8.160 create_integ_mat 1 6.0 0.022 0.027 7.802 7.803 collocate_single_gaussian 272 10.0 0.040 0.043 7.354 7.646 array2fm 1 7.0 0.000 0.000 6.758 7.167 pw_scatter_s 2720 13.7 4.334 4.491 4.334 4.491 pw_gather_s 2722 13.2 3.398 3.781 3.398 3.781 array2fm_buffer_send 1 8.0 2.890 3.130 2.890 3.130 pw_poisson_solve 545 10.5 1.124 1.201 2.221 2.481 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.374013, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2806.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_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 0.025 0.039 423.155 423.156 farming_run 1 2.0 422.391 422.394 423.113 423.119 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.235603E+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 2391237. 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 0.017 0.035 217.618 217.619 qs_energies 1 2.0 0.000 0.000 217.390 217.396 scf_env_do_scf 1 3.0 0.000 0.000 114.977 114.977 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.059 114.065 rebuild_ks_matrix 4 6.0 0.000 0.000 114.058 114.064 qs_ks_build_kohn_sham_matrix 4 7.0 0.053 0.059 114.058 114.064 hfx_ks_matrix 4 8.0 0.001 0.001 113.713 113.718 integrate_four_center 4 9.0 0.154 0.469 113.712 113.717 integrate_four_center_main 4 10.0 0.096 0.553 102.231 104.973 integrate_four_center_bin 265 11.0 102.135 104.710 102.135 104.710 mp2_main 1 3.0 0.000 0.000 102.100 102.106 mp2_gpw_main 1 4.0 0.049 0.070 101.123 101.133 init_scf_loop 1 4.0 0.000 0.000 97.058 97.058 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.102 75.156 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.825 54.878 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.157 41.530 46.562 integrate_v_rspace 95 8.0 0.396 0.556 28.004 32.876 pw_transfer 2240 10.6 0.145 0.198 29.532 29.962 fft_wrap_pw1pw2 1868 11.4 0.018 0.024 28.532 28.985 mp2_ri_gpw_compute_en 1 5.0 0.056 0.062 26.861 28.551 grid_integrate_task_list 95 9.0 23.424 28.499 23.424 28.499 ao_to_mo_and_store_B_mult_1 91 7.0 10.613 27.440 10.613 27.440 fft_wrap_pw1pw2_100 730 12.4 1.300 1.531 26.290 26.738 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.842 1.893 25.128 25.141 get_2c_integrals 1 6.0 0.000 0.000 20.189 20.214 compute_2c_integrals 1 7.0 0.003 0.003 19.164 19.169 compute_2c_integrals_loop_lm 1 8.0 0.002 0.007 18.688 19.035 mp2_eri_2c_integrate_gpw 1 9.0 1.728 1.895 18.687 19.034 fft3d_s 1823 13.4 18.489 18.845 18.504 18.859 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 17.917 17.917 calculate_wavefunction 91 8.0 2.009 2.032 9.615 9.795 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.582 8.752 9.465 potential_pw2rs 186 10.0 0.033 0.035 8.445 8.950 local_gemm 172 8.0 8.196 8.903 8.196 8.903 mp2_ri_gpw_compute_en_comm 22 7.0 0.493 0.521 8.143 8.575 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.194 8.571 calc_potential_gpw 182 9.5 0.002 0.003 7.920 8.198 collocate_single_gaussian 91 10.0 0.017 0.021 7.779 8.040 integrate_four_center_load 4 10.0 0.000 0.000 6.746 6.750 hfx_load_balance 1 11.0 0.000 0.000 6.746 6.750 mp_sendrecv_dm3 2068 8.0 6.195 6.655 6.195 6.655 mp_sync 37 10.5 3.466 6.614 3.466 6.614 mp2_ri_gpw_compute_en_ener 172 7.0 6.346 6.425 6.346 6.425 pw_gather_s 912 13.2 4.481 4.944 4.481 4.944 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.115009, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1495.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.730880E+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 0.008 0.064 53.873 53.909 qs_mol_dyn_low 1 2.0 0.003 0.004 53.647 53.675 qs_forces 11 3.9 0.002 0.002 53.578 53.588 qs_energies 11 4.9 0.001 0.002 52.069 52.094 scf_env_do_scf 11 5.9 0.000 0.001 46.141 46.142 scf_env_do_scf_inner_loop 108 6.5 0.002 0.025 44.078 44.078 qs_scf_new_mos 108 7.5 0.000 0.001 33.993 34.284 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.993 34.284 dbcsr_multiply_generic 2286 12.5 0.102 0.108 33.556 34.002 ot_scf_mini 108 9.5 0.002 0.002 32.293 32.461 multiply_cannon 2286 13.5 0.190 0.199 26.144 27.768 multiply_cannon_loop 2286 14.5 1.830 1.930 25.473 27.115 velocity_verlet 10 3.0 0.001 0.002 26.356 26.356 ot_mini 108 10.5 0.001 0.001 19.468 19.718 qs_ot_get_derivative 108 11.5 0.001 0.001 16.430 16.619 mp_waitall_1 245248 16.5 8.456 14.895 8.456 14.895 multiply_cannon_metrocomm3 54864 15.5 0.074 0.079 6.032 13.154 multiply_cannon_multrec 54864 15.5 3.639 5.637 7.561 10.809 qs_ot_get_p 119 10.4 0.001 0.001 8.133 8.434 rebuild_ks_matrix 119 8.3 0.000 0.000 7.864 8.003 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.864 8.003 mp_sum_l 7287 12.8 5.422 7.149 5.422 7.149 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.922 7.051 multiply_cannon_sync_h2d 54864 15.5 5.121 6.452 5.121 6.452 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.541 6.042 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.478 5.588 dbcsr_mm_accdrv_process 76910 16.1 1.840 2.892 3.836 5.481 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.189 5.244 init_scf_run 11 5.9 0.000 0.001 4.667 4.668 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.667 4.667 sum_up_and_integrate 119 10.3 0.001 0.001 4.440 4.446 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.439 4.440 integrate_v_rspace 119 11.3 0.002 0.002 4.428 4.436 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.190 4.290 calculate_rho_elec 119 8.7 0.011 0.016 4.189 4.289 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.238 4.239 cp_fm_redistribute_end 50 14.0 2.163 4.214 2.169 4.217 cp_fm_diag_elpa_base 50 14.0 2.042 4.111 2.046 4.118 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.927 3.139 apply_single 119 13.6 0.000 0.000 2.927 3.138 calculate_dm_sparse 119 9.5 0.000 0.000 2.918 3.075 multiply_cannon_metrocomm1 54864 15.5 0.055 0.061 1.772 2.917 acc_transpose_blocks 54864 15.5 0.234 0.260 2.235 2.761 ot_diis_step 108 11.5 0.006 0.006 2.727 2.727 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.484 2.537 jit_kernel_multiply 13 15.8 1.934 2.504 1.934 2.504 calculate_first_density_matrix 1 7.0 0.000 0.001 2.379 2.386 density_rs2pw 119 9.7 0.004 0.005 2.167 2.282 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.245 2.247 wfi_extrapolate 11 7.9 0.001 0.001 2.215 2.215 grid_integrate_task_list 119 12.3 2.004 2.131 2.004 2.131 init_scf_loop 11 6.9 0.000 0.000 2.047 2.048 mp_sum_d 4135 12.0 1.394 2.025 1.394 2.025 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.924 1.987 potential_pw2rs 119 12.3 0.004 0.004 1.823 1.836 pw_transfer 1439 11.6 0.051 0.056 1.675 1.746 make_m2s 4572 13.5 0.055 0.057 1.665 1.716 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.600 1.672 make_images 4572 14.5 0.133 0.139 1.581 1.631 mp_alltoall_d11v 2130 13.8 1.365 1.578 1.365 1.578 transfer_rs2pw 487 10.6 0.005 0.006 1.370 1.483 acc_transpose_blocks_sync 164592 16.5 1.216 1.455 1.216 1.455 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 1.385 1.429 grid_collocate_task_list 119 9.7 1.352 1.418 1.352 1.418 mp_waitany 12084 13.8 1.265 1.401 1.265 1.401 fft3d_ps 1201 14.6 0.369 0.472 1.315 1.380 transfer_pw2rs 487 13.2 0.006 0.007 1.331 1.340 fft_wrap_pw1pw2_140 487 13.2 0.140 0.154 1.240 1.316 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.774 1.166 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.909000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.272727, yerr=0.962091 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.120320E+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 1404126. 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 0.009 0.032 40.391 40.392 qs_mol_dyn_low 1 2.0 0.003 0.003 40.209 40.219 qs_forces 11 3.9 0.002 0.003 40.149 40.150 qs_energies 11 4.9 0.002 0.002 38.469 38.474 scf_env_do_scf 11 5.9 0.001 0.002 32.961 32.962 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 30.368 30.368 dbcsr_multiply_generic 2286 12.5 0.135 0.204 22.906 23.250 qs_scf_new_mos 108 7.5 0.001 0.001 21.340 21.567 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.340 21.566 ot_scf_mini 108 9.5 0.002 0.003 20.393 20.559 velocity_verlet 10 3.0 0.001 0.002 19.273 19.274 multiply_cannon 2286 13.5 0.211 0.219 17.282 18.678 multiply_cannon_loop 2286 14.5 1.197 1.257 16.045 17.530 ot_mini 108 10.5 0.001 0.001 12.686 12.920 mp_waitall_1 200699 16.5 5.879 11.178 5.879 11.178 qs_ot_get_derivative 108 11.5 0.001 0.001 10.290 10.459 multiply_cannon_metrocomm3 27432 15.5 0.072 0.074 4.143 9.759 multiply_cannon_multrec 27432 15.5 1.824 4.200 6.318 9.191 rebuild_ks_matrix 119 8.3 0.000 0.000 6.968 7.105 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.967 7.105 dbcsr_mm_accdrv_process 47894 16.0 3.555 5.970 4.414 6.607 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.149 6.274 qs_ot_get_p 119 10.4 0.001 0.001 4.711 4.932 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.945 4.862 init_scf_run 11 5.9 0.000 0.001 4.317 4.317 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.317 4.317 mp_sum_l 7287 12.8 2.137 4.146 2.137 4.146 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.103 4.088 apply_single 119 13.6 0.000 0.000 3.102 4.088 sum_up_and_integrate 119 10.3 0.001 0.001 3.873 3.879 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.840 3.868 calculate_rho_elec 119 8.7 0.021 0.024 3.840 3.867 integrate_v_rspace 119 11.3 0.002 0.003 3.859 3.866 make_m2s 4572 13.5 0.053 0.054 3.001 3.280 make_images 4572 14.5 0.207 0.245 2.912 3.190 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.995 3.013 calculate_first_density_matrix 1 7.0 0.000 0.001 2.753 2.756 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.593 2.593 init_scf_loop 11 6.9 0.000 0.000 2.572 2.572 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.362 2.449 ot_diis_step 108 11.5 0.011 0.011 2.347 2.348 calculate_dm_sparse 119 9.5 0.000 0.001 2.184 2.262 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.259 2.259 cp_fm_redistribute_end 50 14.0 1.145 2.233 1.148 2.235 multiply_cannon_sync_h2d 27432 15.5 1.696 2.232 1.696 2.232 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.215 2.218 density_rs2pw 119 9.7 0.004 0.004 2.133 2.199 jit_kernel_multiply 10 16.3 0.802 2.190 0.802 2.190 cp_fm_diag_elpa_base 50 14.0 1.055 2.144 1.083 2.177 acc_transpose_blocks 27432 15.5 0.115 0.119 1.650 1.995 grid_integrate_task_list 119 12.3 1.832 1.950 1.832 1.950 pw_transfer 1439 11.6 0.065 0.069 1.908 1.941 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.838 1.876 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.816 1.852 make_images_data 4572 15.5 0.048 0.054 1.350 1.748 prepare_preconditioner 11 7.9 0.000 0.000 1.646 1.673 make_preconditioner 11 8.9 0.000 0.000 1.646 1.673 potential_pw2rs 119 12.3 0.006 0.006 1.609 1.619 hybrid_alltoall_any 4725 16.4 0.055 0.117 1.184 1.598 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.505 1.561 wfi_extrapolate 11 7.9 0.001 0.001 1.500 1.500 fft3d_ps 1201 14.6 0.518 0.572 1.452 1.482 mp_alltoall_d11v 2130 13.8 1.324 1.460 1.324 1.460 fft_wrap_pw1pw2_140 487 13.2 0.160 0.168 1.404 1.440 grid_collocate_task_list 119 9.7 1.287 1.357 1.287 1.357 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.346 1.355 transfer_rs2pw 487 10.6 0.005 0.006 1.279 1.352 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.230 1.277 mp_allgather_i34 2286 14.5 0.656 1.097 0.656 1.097 make_images_sizes 4572 15.5 0.005 0.005 0.781 1.044 mp_alltoall_i44 4572 16.5 0.776 1.040 0.776 1.040 transfer_pw2rs 487 13.2 0.005 0.006 1.030 1.039 mp_sum_d 4135 12.0 0.542 1.034 0.542 1.034 acc_transpose_blocks_sync 82296 16.5 0.819 0.947 0.819 0.947 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.925 0.927 acc_transpose_blocks_kernels 27432 16.5 0.188 0.277 0.690 0.904 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.875 0.888 dbcsr_complete_redistribute 329 12.2 0.171 0.269 0.764 0.831 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.392000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.363636, yerr=1.871933 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 519.778304E+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 0.009 0.027 34.697 34.698 qs_mol_dyn_low 1 2.0 0.003 0.003 34.488 34.496 qs_forces 11 3.9 0.002 0.003 34.421 34.422 qs_energies 11 4.9 0.002 0.002 32.842 32.845 scf_env_do_scf 11 5.9 0.001 0.001 28.069 28.070 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 25.323 25.323 dbcsr_multiply_generic 2286 12.5 0.097 0.100 17.776 17.921 qs_scf_new_mos 108 7.5 0.001 0.001 17.117 17.134 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.117 17.134 ot_scf_mini 108 9.5 0.002 0.002 16.333 16.346 velocity_verlet 10 3.0 0.002 0.002 16.272 16.274 multiply_cannon 2286 13.5 0.195 0.199 13.946 14.908 multiply_cannon_loop 2286 14.5 0.860 0.905 13.097 14.121 ot_mini 108 10.5 0.001 0.001 9.863 9.884 qs_ot_get_derivative 108 11.5 0.001 0.001 8.310 8.323 multiply_cannon_multrec 18288 15.5 1.872 2.923 6.966 7.318 rebuild_ks_matrix 119 8.3 0.000 0.000 6.365 6.381 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.364 6.381 dbcsr_mm_accdrv_process 38222 16.0 4.898 5.773 5.001 5.836 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.633 5.650 mp_waitall_1 158411 16.6 3.201 4.365 3.201 4.365 qs_ot_get_p 119 10.4 0.001 0.001 4.120 4.143 sum_up_and_integrate 119 10.3 0.001 0.001 3.920 3.926 integrate_v_rspace 119 11.3 0.003 0.003 3.907 3.915 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.201 3.816 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.528 3.540 calculate_rho_elec 119 8.7 0.031 0.031 3.528 3.539 init_scf_run 11 5.9 0.000 0.001 3.519 3.519 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.519 3.519 multiply_cannon_metrocomm3 18288 15.5 0.048 0.049 1.659 2.851 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.812 2.821 init_scf_loop 11 6.9 0.000 0.000 2.728 2.728 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.482 2.483 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.054 2.391 apply_single 119 13.6 0.000 0.000 2.054 2.390 density_rs2pw 119 9.7 0.004 0.004 2.076 2.225 calculate_first_density_matrix 1 7.0 0.000 0.001 2.222 2.222 make_m2s 4572 13.5 0.045 0.047 2.111 2.221 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.162 2.172 cp_fm_diag_elpa_base 50 14.0 2.138 2.151 2.160 2.170 make_images 4572 14.5 0.194 0.206 2.026 2.135 pw_transfer 1439 11.6 0.065 0.069 2.044 2.066 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.951 1.976 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.902 1.924 grid_integrate_task_list 119 12.3 1.808 1.900 1.808 1.900 calculate_dm_sparse 119 9.5 0.000 0.001 1.885 1.899 prepare_preconditioner 11 7.9 0.000 0.000 1.880 1.883 make_preconditioner 11 8.9 0.000 0.000 1.880 1.883 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.732 1.816 mp_sum_l 7287 12.8 1.326 1.767 1.326 1.767 acc_transpose_blocks 18288 15.5 0.081 0.082 1.641 1.713 potential_pw2rs 119 12.3 0.007 0.008 1.704 1.710 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.704 1.706 fft3d_ps 1201 14.6 0.527 0.547 1.520 1.546 ot_diis_step 108 11.5 0.011 0.011 1.522 1.523 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.507 1.513 fft_wrap_pw1pw2_140 487 13.2 0.211 0.217 1.459 1.479 grid_collocate_task_list 119 9.7 1.234 1.353 1.234 1.353 transfer_rs2pw 487 10.6 0.005 0.005 1.206 1.336 wfi_extrapolate 11 7.9 0.001 0.001 1.238 1.238 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.202 1.209 multiply_cannon_sync_h2d 18288 15.5 1.010 1.165 1.010 1.165 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.433 1.092 make_images_data 4572 15.5 0.048 0.052 0.929 1.087 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.036 1.062 transfer_pw2rs 487 13.2 0.004 0.005 1.036 1.046 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.003 1.004 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.808 0.974 mp_alltoall_d11v 2130 13.8 0.778 0.914 0.778 0.914 mp_alltoall_z22v 1201 16.6 0.816 0.906 0.816 0.906 cp_fm_cholesky_invert 11 10.9 0.853 0.858 0.853 0.858 acc_transpose_blocks_sync 54864 16.5 0.745 0.811 0.745 0.811 acc_transpose_blocks_kernels 18288 16.5 0.217 0.223 0.795 0.810 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.804 0.806 jit_kernel_multiply 4 16.2 0.047 0.779 0.047 0.779 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.676 0.743 transfer_rs2pw_140 130 11.5 0.117 0.121 0.576 0.710 mp_waitany 9880 13.7 0.597 0.708 0.597 0.708 mp_allgather_i34 2286 14.5 0.297 0.705 0.297 0.705 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.698000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 557.625344E+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 0.012 0.029 37.041 37.042 qs_mol_dyn_low 1 2.0 0.003 0.004 36.864 36.871 qs_forces 11 3.9 0.002 0.003 36.804 36.806 qs_energies 11 4.9 0.002 0.003 35.076 35.083 scf_env_do_scf 11 5.9 0.000 0.001 30.025 30.026 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 26.531 26.531 dbcsr_multiply_generic 2286 12.5 0.100 0.103 19.983 20.127 velocity_verlet 10 3.0 0.002 0.002 18.880 18.882 qs_scf_new_mos 108 7.5 0.001 0.001 18.273 18.327 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.272 18.326 ot_scf_mini 108 9.5 0.002 0.003 17.230 17.279 multiply_cannon 2286 13.5 0.219 0.229 16.075 16.696 multiply_cannon_loop 2286 14.5 1.507 1.619 15.064 15.503 ot_mini 108 10.5 0.001 0.001 10.486 10.550 multiply_cannon_multrec 27432 15.5 2.445 3.113 8.976 9.304 qs_ot_get_derivative 108 11.5 0.001 0.001 8.595 8.644 dbcsr_mm_accdrv_process 47916 15.9 5.900 7.604 6.429 7.765 rebuild_ks_matrix 119 8.3 0.000 0.000 6.424 6.479 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.423 6.479 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.698 5.746 qs_ot_get_p 119 10.4 0.001 0.001 3.870 3.943 init_scf_run 11 5.9 0.000 0.001 3.692 3.693 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.692 3.693 sum_up_and_integrate 119 10.3 0.001 0.001 3.595 3.602 integrate_v_rspace 119 11.3 0.003 0.003 3.584 3.591 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.198 3.591 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.472 3.501 calculate_rho_elec 119 8.7 0.040 0.046 3.472 3.500 init_scf_loop 11 6.9 0.000 0.000 3.469 3.470 prepare_preconditioner 11 7.9 0.000 0.000 2.623 2.631 make_preconditioner 11 8.9 0.000 0.000 2.623 2.631 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.234 2.552 acc_transpose_blocks 27432 15.5 0.121 0.124 2.375 2.546 make_m2s 4572 13.5 0.055 0.058 2.412 2.543 mp_waitall_1 137007 16.6 1.901 2.485 1.901 2.485 make_images 4572 14.5 0.274 0.336 2.303 2.432 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.393 2.418 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.375 2.388 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.120 2.248 apply_single 119 13.6 0.000 0.000 2.119 2.248 calculate_first_density_matrix 1 7.0 0.000 0.001 2.225 2.229 calculate_dm_sparse 119 9.5 0.000 0.000 2.158 2.216 pw_transfer 1439 11.6 0.065 0.069 1.993 2.034 density_rs2pw 119 9.7 0.004 0.004 1.925 2.013 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.958 1.959 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.901 1.945 grid_integrate_task_list 119 12.3 1.820 1.893 1.820 1.893 ot_diis_step 108 11.5 0.012 0.012 1.850 1.850 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.816 1.817 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.685 1.696 cp_fm_diag_elpa_base 50 14.0 1.650 1.669 1.682 1.693 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.623 1.635 fft_wrap_pw1pw2_140 487 13.2 0.245 0.260 1.542 1.589 acc_transpose_blocks_sync 82296 16.5 1.409 1.565 1.409 1.565 fft3d_ps 1201 14.6 0.555 0.606 1.421 1.446 jit_kernel_multiply 8 16.4 0.465 1.421 0.465 1.421 wfi_extrapolate 11 7.9 0.001 0.001 1.411 1.411 potential_pw2rs 119 12.3 0.008 0.009 1.396 1.401 mp_sum_l 7287 12.8 1.068 1.387 1.068 1.387 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.766 1.342 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.314 1.332 grid_collocate_task_list 119 9.7 1.249 1.331 1.249 1.331 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.248 1.260 cp_fm_upper_to_full 72 14.2 0.802 1.140 0.802 1.140 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.086 1.088 dbcsr_complete_redistribute 329 12.2 0.124 0.155 0.814 1.082 transfer_rs2pw 487 10.6 0.005 0.005 0.993 1.080 make_images_data 4572 15.5 0.048 0.052 0.903 1.040 hybrid_alltoall_any 4725 16.4 0.066 0.155 0.773 0.956 mp_alltoall_d11v 2130 13.8 0.803 0.911 0.803 0.911 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.875 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.589 0.854 transfer_pw2rs 487 13.2 0.004 0.004 0.823 0.828 acc_transpose_blocks_kernels 27432 16.5 0.270 0.277 0.815 0.825 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.817 0.822 cp_fm_cholesky_invert 11 10.9 0.811 0.814 0.811 0.814 mp_alltoall_z22v 1201 16.6 0.748 0.777 0.748 0.777 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.042000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.909091, yerr=3.146558 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_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 615.460864E+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 0.009 0.027 31.168 31.169 qs_mol_dyn_low 1 2.0 0.003 0.003 30.854 30.863 qs_forces 11 3.9 0.002 0.002 30.786 30.787 qs_energies 11 4.9 0.001 0.001 29.067 29.070 scf_env_do_scf 11 5.9 0.000 0.001 24.074 24.074 scf_env_do_scf_inner_loop 108 6.5 0.004 0.013 21.220 21.221 velocity_verlet 10 3.0 0.002 0.002 15.927 15.931 dbcsr_multiply_generic 2286 12.5 0.093 0.098 13.873 13.976 qs_scf_new_mos 108 7.5 0.001 0.001 13.136 13.171 qs_scf_loop_do_ot 108 8.5 0.001 0.001 13.135 13.170 ot_scf_mini 108 9.5 0.002 0.002 12.376 12.407 multiply_cannon 2286 13.5 0.224 0.230 10.801 11.326 multiply_cannon_loop 2286 14.5 0.647 0.666 9.798 10.023 ot_mini 108 10.5 0.001 0.001 6.962 7.002 multiply_cannon_multrec 9144 15.5 1.712 1.892 6.026 6.249 rebuild_ks_matrix 119 8.3 0.000 0.000 6.014 6.041 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.013 6.040 qs_ot_get_derivative 108 11.5 0.001 0.001 5.591 5.620 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.376 5.401 dbcsr_mm_accdrv_process 12550 15.8 3.500 4.227 4.204 4.296 sum_up_and_integrate 119 10.3 0.001 0.001 3.634 3.639 integrate_v_rspace 119 11.3 0.003 0.003 3.624 3.628 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.592 3.602 calculate_rho_elec 119 8.7 0.060 0.061 3.591 3.601 init_scf_run 11 5.9 0.000 0.001 3.523 3.523 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.523 3.523 qs_ot_get_p 119 10.4 0.001 0.001 3.435 3.477 init_scf_loop 11 6.9 0.000 0.000 2.830 2.832 mp_waitall_1 115863 16.7 1.828 2.373 1.828 2.373 make_m2s 4572 13.5 0.035 0.035 2.100 2.306 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.237 2.239 calculate_first_density_matrix 1 7.0 0.000 0.000 2.234 2.235 make_images 4572 14.5 0.270 0.301 2.010 2.215 pw_transfer 1439 11.6 0.065 0.068 2.139 2.149 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.046 2.058 prepare_preconditioner 11 7.9 0.000 0.000 2.040 2.046 make_preconditioner 11 8.9 0.000 0.000 2.040 2.046 density_rs2pw 119 9.7 0.004 0.004 1.947 2.006 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.967 1.968 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.920 1.949 grid_integrate_task_list 119 12.3 1.874 1.939 1.874 1.939 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.912 1.936 calculate_dm_sparse 119 9.5 0.000 0.000 1.842 1.861 fft_wrap_pw1pw2_140 487 13.2 0.322 0.330 1.633 1.648 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.614 1.621 cp_fm_diag_elpa_base 50 14.0 1.586 1.602 1.612 1.620 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.600 1.602 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.531 1.546 acc_transpose_blocks 9144 15.5 0.042 0.043 1.460 1.483 fft3d_ps 1201 14.6 0.561 0.577 1.460 1.469 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.406 1.416 potential_pw2rs 119 12.3 0.010 0.011 1.406 1.411 grid_collocate_task_list 119 9.7 1.300 1.395 1.300 1.395 jit_kernel_multiply 8 15.6 0.665 1.377 0.665 1.377 ot_diis_step 108 11.5 0.013 0.013 1.349 1.349 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.265 1.297 apply_single 119 13.6 0.000 0.000 1.265 1.297 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 1.241 1.242 wfi_extrapolate 11 7.9 0.001 0.001 1.234 1.234 hybrid_alltoall_any 4725 16.4 0.066 0.175 0.919 1.202 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.183 1.189 make_images_data 4572 15.5 0.042 0.046 0.927 1.188 cp_fm_cholesky_invert 11 10.9 1.096 1.098 1.096 1.098 mp_alltoall_d11v 2130 13.8 0.947 1.032 0.947 1.032 transfer_rs2pw 487 10.6 0.004 0.004 0.944 1.009 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.873 0.925 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.903 0.909 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 0.393 0.849 mp_alltoall_z22v 1201 16.6 0.773 0.806 0.773 0.806 mp_allgather_i34 2286 14.5 0.316 0.805 0.316 0.805 transfer_pw2rs 487 13.2 0.004 0.004 0.773 0.775 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.765 0.767 acc_transpose_blocks_sync 27432 16.5 0.728 0.749 0.728 0.749 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.687 0.737 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.675 0.678 dbcsr_complete_redistribute 329 12.2 0.168 0.176 0.621 0.649 mp_sum_l 7287 12.8 0.524 0.649 0.524 0.649 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=31.169000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=584.909091, yerr=3.848366 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_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 757.665792E+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 0.012 0.033 43.261 43.262 qs_mol_dyn_low 1 2.0 0.003 0.003 43.044 43.054 qs_forces 11 3.9 0.002 0.002 42.984 42.985 qs_energies 11 4.9 0.001 0.001 40.943 40.945 scf_env_do_scf 11 5.9 0.001 0.001 35.169 35.170 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 27.225 27.226 velocity_verlet 10 3.0 0.002 0.002 24.285 24.290 dbcsr_multiply_generic 2286 12.5 0.100 0.103 19.179 19.412 qs_scf_new_mos 108 7.5 0.001 0.001 17.670 17.782 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.670 17.782 ot_scf_mini 108 9.5 0.002 0.002 16.497 16.605 multiply_cannon 2286 13.5 0.298 0.305 15.063 15.993 multiply_cannon_loop 2286 14.5 0.870 0.899 13.788 14.749 ot_mini 108 10.5 0.001 0.001 9.960 10.077 multiply_cannon_multrec 9144 15.5 3.354 4.769 8.782 8.893 qs_ot_get_derivative 108 11.5 0.001 0.001 7.913 8.019 init_scf_loop 11 6.9 0.000 0.000 7.913 7.914 rebuild_ks_matrix 119 8.3 0.000 0.000 7.202 7.345 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.201 7.344 prepare_preconditioner 11 7.9 0.000 0.000 6.946 6.958 make_preconditioner 11 8.9 0.000 0.000 6.945 6.958 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.490 6.834 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.465 6.595 dbcsr_mm_accdrv_process 12550 15.8 4.516 5.830 5.293 6.551 cp_fm_upper_to_full 72 14.2 3.187 4.591 3.187 4.591 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.150 4.166 calculate_rho_elec 119 8.7 0.118 0.121 4.149 4.166 sum_up_and_integrate 119 10.3 0.001 0.001 3.857 3.862 integrate_v_rspace 119 11.3 0.003 0.004 3.846 3.852 qs_ot_get_p 119 10.4 0.001 0.001 3.565 3.703 init_scf_run 11 5.9 0.000 0.001 3.692 3.692 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.692 3.692 mp_waitall_1 94719 16.7 2.450 3.494 2.450 3.494 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.869 3.262 dbcsr_complete_redistribute 329 12.2 0.286 0.290 2.007 2.851 make_m2s 4572 13.5 0.039 0.039 2.546 2.700 pw_transfer 1439 11.6 0.068 0.068 2.670 2.675 make_images 4572 14.5 0.356 0.387 2.424 2.577 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.572 2.576 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.693 2.529 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.189 2.498 apply_single 119 13.6 0.000 0.000 2.188 2.497 calculate_dm_sparse 119 9.5 0.000 0.001 2.266 2.281 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.450 2.281 mp_alltoall_i22 627 13.8 1.434 2.267 1.434 2.267 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.186 2.238 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.295 2.232 calculate_first_density_matrix 1 7.0 0.000 0.001 2.198 2.202 fft_wrap_pw1pw2_140 487 13.2 0.575 0.579 2.168 2.173 density_rs2pw 119 9.7 0.004 0.004 2.155 2.169 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.124 2.126 grid_integrate_task_list 119 12.3 2.083 2.101 2.083 2.101 ot_diis_step 108 11.5 0.014 0.014 2.024 2.025 acc_transpose_blocks 9144 15.5 0.044 0.045 1.830 1.876 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.827 1.828 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.777 1.778 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.757 1.759 mp_sum_l 7287 12.8 1.023 1.723 1.023 1.723 fft3d_ps 1201 14.6 0.596 0.605 1.669 1.673 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.562 1.600 grid_collocate_task_list 119 9.7 1.524 1.555 1.524 1.555 cp_fm_cholesky_invert 11 10.9 1.542 1.546 1.542 1.546 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.459 1.460 cp_fm_diag_elpa_base 50 14.0 1.306 1.367 1.456 1.458 hybrid_alltoall_any 4725 16.4 0.091 0.150 1.169 1.428 wfi_extrapolate 11 7.9 0.001 0.001 1.425 1.425 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.342 1.362 make_images_data 4572 15.5 0.047 0.050 1.123 1.348 potential_pw2rs 119 12.3 0.014 0.014 1.345 1.347 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.259 1.267 mp_alltoall_d11v 2130 13.8 1.229 1.249 1.229 1.249 acc_transpose_blocks_sync 27432 16.5 1.114 1.158 1.114 1.158 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.106 1.124 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.998 1.051 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.948 0.962 mp_alltoall_z22v 1201 16.6 0.937 0.953 0.937 0.953 qs_create_task_list 11 7.9 0.000 0.000 0.938 0.950 generate_qs_task_list 11 8.9 0.372 0.392 0.938 0.949 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.262000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=710.636364, yerr=14.386058 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.616640E+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 1383689. 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 0.030 0.043 83.164 83.166 qs_mol_dyn_low 1 2.0 0.003 0.003 82.865 82.876 qs_forces 11 3.9 0.003 0.003 82.797 82.798 qs_energies 11 4.9 0.001 0.002 79.853 79.867 scf_env_do_scf 11 5.9 0.000 0.001 70.847 70.849 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 65.123 65.123 dbcsr_multiply_generic 2055 12.4 0.107 0.109 51.253 51.501 qs_scf_new_mos 99 7.5 0.000 0.001 47.728 47.855 qs_scf_loop_do_ot 99 8.5 0.000 0.001 47.727 47.855 ot_scf_mini 99 9.5 0.002 0.002 45.271 45.409 velocity_verlet 10 3.0 0.001 0.002 44.021 44.022 multiply_cannon 2055 13.4 0.185 0.190 42.667 43.695 multiply_cannon_loop 2055 14.4 1.780 1.843 41.634 42.661 ot_mini 99 10.5 0.001 0.001 26.578 26.704 qs_ot_get_derivative 99 11.5 0.001 0.001 19.685 19.794 multiply_cannon_multrec 49320 15.4 11.386 12.066 17.359 17.989 rebuild_ks_matrix 110 8.3 0.000 0.000 14.700 14.829 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.700 14.828 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.851 12.955 mp_waitall_1 220248 16.4 11.071 11.992 11.071 11.992 qs_ot_get_p 110 10.4 0.001 0.001 10.063 10.200 multiply_cannon_sync_h2d 49320 15.4 9.582 10.140 9.582 10.140 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.584 8.126 multiply_cannon_metrocomm3 49320 15.4 0.084 0.088 6.484 7.711 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.259 7.665 apply_single 110 13.6 0.000 0.001 7.259 7.665 sum_up_and_integrate 110 10.3 0.002 0.003 7.131 7.143 integrate_v_rspace 110 11.3 0.003 0.003 7.105 7.124 init_scf_run 11 5.9 0.000 0.001 6.892 6.892 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.891 6.892 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.623 6.778 calculate_rho_elec 110 8.6 0.020 0.024 6.622 6.777 ot_diis_step 99 11.5 0.006 0.006 6.660 6.660 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.631 6.648 dbcsr_mm_accdrv_process 87628 16.1 3.007 3.102 5.842 6.123 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.823 5.824 init_scf_loop 11 6.9 0.000 0.000 5.698 5.699 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.294 5.356 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.152 5.179 cp_fm_diag_elpa_base 48 14.0 5.139 5.167 5.150 5.177 mp_sum_l 6594 12.7 3.885 4.671 3.885 4.671 make_m2s 4110 13.4 0.062 0.067 4.246 4.349 make_images 4110 14.4 0.177 0.190 4.148 4.256 wfi_extrapolate 11 7.9 0.001 0.001 4.142 4.142 calculate_dm_sparse 110 9.5 0.001 0.001 3.859 3.970 density_rs2pw 110 9.6 0.004 0.005 3.571 3.703 multiply_cannon_metrocomm1 49320 15.4 0.064 0.067 2.544 3.674 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.572 3.576 prepare_preconditioner 11 7.9 0.000 0.000 3.424 3.443 make_preconditioner 11 8.9 0.000 0.000 3.424 3.443 grid_integrate_task_list 110 12.3 3.274 3.407 3.274 3.407 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.308 3.352 pw_transfer 1331 11.6 0.055 0.067 3.230 3.314 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.224 3.271 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.142 3.228 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.191 3.226 fft_wrap_pw1pw2_140 451 13.1 0.389 0.428 2.659 2.758 potential_pw2rs 110 12.3 0.005 0.006 2.715 2.733 calculate_first_density_matrix 1 7.0 0.000 0.001 2.650 2.656 acc_transpose_blocks 49320 15.4 0.216 0.225 2.562 2.638 jit_kernel_multiply 13 15.9 2.559 2.578 2.559 2.578 mp_alltoall_d11v 2046 13.8 2.180 2.554 2.180 2.554 fft3d_ps 1111 14.6 0.794 0.881 2.438 2.495 grid_collocate_task_list 110 9.6 2.159 2.323 2.159 2.323 transfer_rs2pw 451 10.6 0.005 0.006 2.084 2.225 mp_waitany 14300 13.8 1.915 2.187 1.915 2.187 make_images_data 4110 15.4 0.044 0.047 1.896 2.052 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.979 2.005 hybrid_alltoall_any 4261 16.3 0.086 0.484 1.655 1.935 mp_sum_d 3889 11.9 1.422 1.920 1.422 1.920 cp_fm_cholesky_invert 11 10.9 1.903 1.908 1.903 1.908 transfer_pw2rs 451 13.1 0.006 0.007 1.853 1.864 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.684 1.725 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.166000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.818182, yerr=2.622219 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_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 592.162816E+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 2770492. 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 0.011 0.030 71.094 71.095 qs_mol_dyn_low 1 2.0 0.003 0.003 70.376 70.892 qs_forces 11 3.9 0.003 0.003 70.300 70.301 qs_energies 11 4.9 0.001 0.002 66.947 66.950 scf_env_do_scf 11 5.9 0.001 0.001 58.369 58.372 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 50.441 50.442 dbcsr_multiply_generic 2055 12.4 0.115 0.120 38.268 38.421 velocity_verlet 10 3.0 0.001 0.002 36.886 36.887 qs_scf_new_mos 99 7.5 0.001 0.001 34.113 34.250 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.112 34.250 ot_scf_mini 99 9.5 0.003 0.003 32.434 32.568 multiply_cannon 2055 13.4 0.226 0.247 31.376 32.344 multiply_cannon_loop 2055 14.4 1.165 1.197 30.054 30.885 ot_mini 99 10.5 0.001 0.001 18.693 18.823 multiply_cannon_multrec 24660 15.4 6.969 8.513 14.032 15.500 rebuild_ks_matrix 110 8.3 0.000 0.001 13.852 13.973 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.852 13.972 qs_ot_get_derivative 99 11.5 0.001 0.001 12.899 13.034 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.207 12.321 mp_waitall_1 176588 16.5 7.764 10.283 7.764 10.283 init_scf_loop 11 6.9 0.000 0.000 7.888 7.888 multiply_cannon_metrocomm3 24660 15.4 0.072 0.075 5.113 7.570 multiply_cannon_sync_h2d 24660 15.4 6.356 7.259 6.356 7.259 dbcsr_mm_accdrv_process 52282 16.1 5.585 6.362 6.893 7.233 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.466 7.215 apply_single 110 13.6 0.000 0.001 6.466 7.215 qs_ot_get_p 110 10.4 0.001 0.001 7.064 7.201 sum_up_and_integrate 110 10.3 0.001 0.003 6.696 6.708 integrate_v_rspace 110 11.3 0.002 0.003 6.669 6.682 init_scf_run 11 5.9 0.000 0.001 6.194 6.195 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.194 6.195 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.102 6.110 calculate_rho_elec 110 8.6 0.039 0.047 6.101 6.110 prepare_preconditioner 11 7.9 0.000 0.000 5.849 5.866 make_preconditioner 11 8.9 0.000 0.000 5.849 5.866 ot_diis_step 99 11.5 0.010 0.011 5.746 5.747 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.851 5.586 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.420 5.576 qs_ot_p2m_diag 48 11.0 0.029 0.044 5.017 5.039 make_m2s 4110 13.4 0.058 0.061 4.437 4.870 make_images 4110 14.4 0.409 0.469 4.327 4.755 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.525 4.525 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.904 3.921 cp_fm_diag_elpa_base 48 14.0 3.853 3.879 3.901 3.918 pw_transfer 1331 11.6 0.066 0.073 3.761 3.903 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.653 3.799 wfi_extrapolate 11 7.9 0.001 0.001 3.585 3.585 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.425 3.495 density_rs2pw 110 9.6 0.004 0.005 3.327 3.480 grid_integrate_task_list 110 12.3 3.157 3.353 3.157 3.353 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.287 3.289 fft_wrap_pw1pw2_140 451 13.1 0.460 0.475 3.001 3.149 calculate_dm_sparse 110 9.5 0.001 0.001 2.993 3.019 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.908 2.952 fft3d_ps 1111 14.6 1.110 1.339 2.751 2.911 make_images_data 4110 15.4 0.049 0.053 2.391 2.849 hybrid_alltoall_any 4261 16.3 0.108 0.461 2.059 2.843 cp_fm_cholesky_invert 11 10.9 2.748 2.755 2.748 2.755 potential_pw2rs 110 12.3 0.008 0.009 2.543 2.563 calculate_first_density_matrix 1 7.0 0.000 0.001 2.513 2.515 mp_sum_l 6594 12.7 1.823 2.451 1.823 2.451 grid_collocate_task_list 110 9.6 2.174 2.325 2.174 2.325 mp_alltoall_d11v 2046 13.8 1.854 2.033 1.854 2.033 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.989 2.010 acc_transpose_blocks 24660 15.4 0.117 0.120 1.982 2.010 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.861 1.862 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.803 1.815 jit_kernel_multiply 9 16.3 0.945 1.795 0.945 1.795 transfer_rs2pw 451 10.6 0.006 0.007 1.544 1.725 mp_allgather_i34 2055 14.4 0.626 1.646 0.626 1.646 multiply_cannon_metrocomm4 22605 15.4 0.078 0.083 0.780 1.590 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.550 1.560 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.387 1.491 dbcsr_complete_redistribute 325 12.2 0.239 0.302 1.221 1.488 mp_irecv_dv 57340 16.2 0.654 1.472 0.654 1.472 transfer_pw2rs 451 13.1 0.006 0.007 1.442 1.462 mp_alltoall_z22v 1111 16.6 1.371 1.451 1.371 1.451 cp_fm_cholesky_decompose 22 10.9 1.440 1.446 1.440 1.446 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.095000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=559.363636, yerr=7.401139 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_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 666.660864E+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 0.015 0.036 61.483 61.483 qs_mol_dyn_low 1 2.0 0.003 0.003 61.215 61.226 qs_forces 11 3.9 0.003 0.003 61.141 61.142 qs_energies 11 4.9 0.001 0.002 57.940 57.943 scf_env_do_scf 11 5.9 0.000 0.001 49.947 49.947 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.020 41.020 velocity_verlet 10 3.0 0.001 0.002 33.368 33.369 dbcsr_multiply_generic 2055 12.4 0.107 0.112 29.087 29.293 qs_scf_new_mos 99 7.5 0.001 0.001 26.059 26.164 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.058 26.163 ot_scf_mini 99 9.5 0.002 0.003 24.793 24.908 multiply_cannon 2055 13.4 0.212 0.218 22.558 23.731 multiply_cannon_loop 2055 14.4 0.821 0.838 21.346 22.390 ot_mini 99 10.5 0.001 0.001 13.962 14.078 rebuild_ks_matrix 110 8.3 0.000 0.000 12.317 12.437 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.317 12.436 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.839 10.948 multiply_cannon_multrec 16440 15.4 3.739 4.824 9.952 10.860 mp_waitall_1 139946 16.5 7.066 9.786 7.066 9.786 qs_ot_get_derivative 99 11.5 0.001 0.001 9.539 9.655 init_scf_loop 11 6.9 0.000 0.000 8.888 8.889 prepare_preconditioner 11 7.9 0.000 0.000 7.128 7.145 make_preconditioner 11 8.9 0.000 0.000 7.128 7.145 multiply_cannon_metrocomm3 16440 15.4 0.046 0.049 4.228 7.032 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.428 6.800 sum_up_and_integrate 110 10.3 0.001 0.002 6.434 6.449 integrate_v_rspace 110 11.3 0.003 0.003 6.409 6.424 dbcsr_mm_accdrv_process 34862 16.1 5.326 5.769 6.058 6.200 qs_ot_get_p 110 10.4 0.001 0.001 5.925 6.076 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.937 5.948 calculate_rho_elec 110 8.6 0.059 0.059 5.936 5.947 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 5.544 5.544 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.891 5.362 apply_single 110 13.6 0.000 0.000 4.890 5.362 make_m2s 4110 13.4 0.050 0.052 4.431 4.793 make_images 4110 14.4 0.399 0.521 4.316 4.678 ot_diis_step 99 11.5 0.011 0.011 4.393 4.393 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.181 4.184 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.253 3.870 multiply_cannon_sync_h2d 16440 15.4 3.214 3.805 3.214 3.805 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.770 3.771 pw_transfer 1331 11.6 0.066 0.073 3.706 3.714 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.599 3.608 grid_integrate_task_list 110 12.3 3.180 3.392 3.180 3.392 density_rs2pw 110 9.6 0.004 0.005 3.162 3.371 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.174 3.183 cp_fm_diag_elpa_base 48 14.0 3.105 3.137 3.172 3.181 fft_wrap_pw1pw2_140 451 13.1 0.578 0.586 3.047 3.061 wfi_extrapolate 11 7.9 0.001 0.001 3.058 3.058 make_images_data 4110 15.4 0.046 0.051 2.471 2.980 hybrid_alltoall_any 4261 16.3 0.110 0.381 2.212 2.895 cp_fm_cholesky_invert 11 10.9 2.860 2.868 2.860 2.868 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.798 2.799 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.590 2.655 calculate_dm_sparse 110 9.5 0.001 0.001 2.572 2.606 fft3d_ps 1111 14.6 1.088 1.098 2.542 2.551 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.437 2.477 grid_collocate_task_list 110 9.6 2.220 2.432 2.220 2.432 calculate_first_density_matrix 1 7.0 0.000 0.000 2.389 2.391 multiply_cannon_metrocomm4 14385 15.4 0.048 0.052 0.864 2.390 potential_pw2rs 110 12.3 0.011 0.011 2.266 2.276 mp_irecv_dv 48980 15.7 0.789 2.258 0.789 2.258 mp_alltoall_d11v 2046 13.8 1.883 2.180 1.883 2.180 mp_sum_l 6594 12.7 1.472 2.146 1.472 2.146 dbcsr_complete_redistribute 325 12.2 0.327 0.377 1.537 2.028 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.989 1.989 cp_fm_upper_to_full 70 14.2 1.420 1.811 1.420 1.811 cp_fm_cholesky_decompose 22 10.9 1.698 1.714 1.698 1.714 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.673 1.689 acc_transpose_blocks 16440 15.4 0.078 0.080 1.594 1.641 mp_allgather_i34 2055 14.4 0.523 1.631 0.523 1.631 transfer_rs2pw 451 10.6 0.005 0.006 1.411 1.612 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.497 1.511 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.030 1.498 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.376 1.475 mp_waitany 17072 13.8 1.156 1.377 1.156 1.377 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.321 1.333 transfer_pw2rs 451 13.1 0.005 0.005 1.256 1.263 mp_alltoall_z22v 1111 16.6 1.236 1.252 1.236 1.252 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.483000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.181818, yerr=9.073620 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 731.021312E+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 0.009 0.029 67.877 67.888 qs_mol_dyn_low 1 2.0 0.003 0.003 67.645 67.656 qs_forces 11 3.9 0.005 0.006 67.580 67.590 qs_energies 11 4.9 0.001 0.002 64.187 64.198 scf_env_do_scf 11 5.9 0.000 0.001 55.782 55.792 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 43.958 43.965 velocity_verlet 10 3.0 0.001 0.002 38.424 38.432 dbcsr_multiply_generic 2055 12.4 0.114 0.122 31.460 31.790 qs_scf_new_mos 99 7.5 0.001 0.001 28.578 28.698 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.578 28.698 ot_scf_mini 99 9.5 0.002 0.003 26.888 27.002 multiply_cannon 2055 13.4 0.239 0.257 23.757 25.171 multiply_cannon_loop 2055 14.4 1.413 1.479 22.341 23.162 ot_mini 99 10.5 0.001 0.001 15.487 15.627 multiply_cannon_multrec 24660 15.4 4.104 6.685 12.993 14.196 rebuild_ks_matrix 110 8.3 0.000 0.000 12.271 12.397 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.270 12.396 init_scf_loop 11 6.9 0.000 0.000 11.783 11.785 qs_ot_get_derivative 99 11.5 0.001 0.001 11.252 11.370 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.879 10.991 prepare_preconditioner 11 7.9 0.000 0.000 10.036 10.052 make_preconditioner 11 8.9 0.000 0.000 10.036 10.052 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.253 9.728 dbcsr_mm_accdrv_process 52304 16.0 7.566 9.107 8.731 9.669 mp_waitall_1 121746 16.5 4.760 6.576 4.760 6.576 sum_up_and_integrate 110 10.3 0.001 0.002 6.457 6.471 integrate_v_rspace 110 11.3 0.003 0.003 6.431 6.443 qs_ot_get_p 110 10.4 0.001 0.001 6.184 6.352 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.315 6.329 calculate_rho_elec 110 8.6 0.078 0.081 6.315 6.329 make_m2s 4110 13.4 0.061 0.063 5.967 6.274 make_images 4110 14.4 0.579 0.700 5.824 6.127 init_scf_run 11 5.9 0.000 0.001 5.641 5.642 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.640 5.642 cp_fm_upper_to_full 70 14.2 3.343 4.860 3.343 4.860 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.222 4.240 ot_diis_step 99 11.5 0.011 0.012 4.197 4.198 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.093 4.190 apply_single 110 13.6 0.000 0.000 4.092 4.190 pw_transfer 1331 11.6 0.066 0.075 3.930 3.964 dbcsr_complete_redistribute 325 12.2 0.421 0.472 2.775 3.916 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.831 3.886 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.823 3.861 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.652 3.652 density_rs2pw 110 9.6 0.004 0.004 3.387 3.564 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.394 3.458 grid_integrate_task_list 110 12.3 3.298 3.415 3.298 3.415 make_images_data 4110 15.4 0.049 0.053 3.028 3.378 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.177 3.311 hybrid_alltoall_any 4261 16.3 0.124 0.460 2.530 3.289 fft_wrap_pw1pw2_140 451 13.1 0.612 0.633 3.218 3.259 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.512 3.207 wfi_extrapolate 11 7.9 0.001 0.001 3.120 3.121 calculate_dm_sparse 110 9.5 0.001 0.001 3.071 3.107 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.091 3.102 cp_fm_diag_elpa_base 48 14.0 2.939 2.999 3.088 3.100 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.774 2.890 mp_alltoall_i22 605 13.7 1.676 2.867 1.676 2.867 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.862 2.864 cp_fm_cholesky_invert 11 10.9 2.787 2.798 2.787 2.798 fft3d_ps 1111 14.6 1.084 1.113 2.703 2.720 acc_transpose_blocks 24660 15.4 0.115 0.118 2.434 2.580 multiply_cannon_sync_h2d 24660 15.4 2.390 2.532 2.390 2.532 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.465 2.513 calculate_first_density_matrix 1 7.0 0.000 0.001 2.425 2.428 grid_collocate_task_list 110 9.6 2.266 2.426 2.266 2.426 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.251 2.252 mp_alltoall_d11v 2046 13.8 2.049 2.184 2.049 2.184 potential_pw2rs 110 12.3 0.012 0.013 2.131 2.139 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.828 1.864 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.615 1.716 cp_fm_cholesky_decompose 22 10.9 1.662 1.695 1.662 1.695 multiply_cannon_metrocomm4 20550 15.4 0.062 0.066 0.844 1.670 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.627 1.648 transfer_rs2pw 451 10.6 0.005 0.006 1.450 1.648 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.633 1.645 mp_sum_l 6594 12.7 1.047 1.604 1.047 1.604 mp_irecv_dv 62702 16.1 0.740 1.587 0.740 1.587 mp_allgather_i34 2055 14.4 0.503 1.526 0.503 1.526 acc_transpose_blocks_sync 73980 16.4 1.408 1.525 1.408 1.525 mp_alltoall_z22v 1111 16.6 1.402 1.458 1.402 1.458 jit_kernel_multiply 8 15.9 0.820 1.440 0.820 1.440 mp_waitany 13376 13.8 1.153 1.403 1.153 1.403 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.888000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=693.909091, yerr=7.856145 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_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 859.680768E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.030 57.748 57.749 qs_mol_dyn_low 1 2.0 0.003 0.003 57.460 57.470 qs_forces 11 3.9 0.003 0.003 57.393 57.394 qs_energies 11 4.9 0.001 0.002 53.699 53.704 scf_env_do_scf 11 5.9 0.000 0.001 45.057 45.058 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 36.920 36.922 velocity_verlet 10 3.0 0.002 0.002 32.375 32.378 dbcsr_multiply_generic 2055 12.4 0.105 0.109 24.093 24.254 qs_scf_new_mos 99 7.5 0.001 0.001 21.561 21.614 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.560 21.613 ot_scf_mini 99 9.5 0.002 0.002 20.277 20.301 multiply_cannon 2055 13.4 0.239 0.247 17.987 19.599 multiply_cannon_loop 2055 14.4 0.602 0.622 16.659 17.070 rebuild_ks_matrix 110 8.3 0.000 0.000 12.237 12.259 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.236 12.259 ot_mini 99 10.5 0.001 0.001 11.125 11.145 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.916 10.934 multiply_cannon_multrec 8220 15.4 3.184 4.338 7.579 8.476 init_scf_loop 11 6.9 0.000 0.000 8.090 8.091 mp_waitall_1 103326 16.6 6.553 8.054 6.553 8.054 qs_ot_get_derivative 99 11.5 0.001 0.001 7.402 7.426 sum_up_and_integrate 110 10.3 0.001 0.002 6.709 6.721 integrate_v_rspace 110 11.3 0.003 0.003 6.682 6.695 prepare_preconditioner 11 7.9 0.000 0.000 6.380 6.384 make_preconditioner 11 8.9 0.000 0.000 6.380 6.384 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.237 6.255 calculate_rho_elec 110 8.6 0.114 0.115 6.236 6.255 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.926 6.020 init_scf_run 11 5.9 0.000 0.001 5.419 5.419 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.419 5.419 dbcsr_mm_accdrv_process 17442 15.9 3.116 4.108 4.255 5.159 qs_ot_get_p 110 10.4 0.001 0.001 4.999 5.037 make_m2s 4110 13.4 0.039 0.040 4.642 4.941 make_images 4110 14.4 0.651 0.711 4.511 4.808 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 2.982 4.359 pw_transfer 1331 11.6 0.066 0.071 4.020 4.043 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.912 3.938 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.741 3.785 apply_single 110 13.6 0.000 0.000 3.741 3.785 ot_diis_step 99 11.5 0.012 0.012 3.698 3.704 grid_integrate_task_list 110 12.3 3.387 3.566 3.387 3.566 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.542 3.549 fft_wrap_pw1pw2_140 451 13.1 0.777 0.788 3.352 3.386 density_rs2pw 110 9.6 0.004 0.004 3.083 3.274 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.187 3.188 hybrid_alltoall_any 4261 16.3 0.201 0.855 2.453 3.053 cp_fm_cholesky_invert 11 10.9 3.018 3.022 3.018 3.022 make_images_data 4110 15.4 0.041 0.047 2.527 2.993 wfi_extrapolate 11 7.9 0.001 0.001 2.851 2.851 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.740 2.741 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.682 2.684 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.611 2.619 cp_fm_diag_elpa_base 48 14.0 2.552 2.582 2.609 2.617 fft3d_ps 1111 14.6 1.138 1.177 2.577 2.596 calculate_dm_sparse 110 9.5 0.001 0.001 2.511 2.546 grid_collocate_task_list 110 9.6 2.366 2.529 2.366 2.529 multiply_cannon_sync_h2d 8220 15.4 2.384 2.464 2.384 2.464 calculate_first_density_matrix 1 7.0 0.000 0.001 2.450 2.452 potential_pw2rs 110 12.3 0.015 0.016 2.349 2.351 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.269 2.290 mp_alltoall_d11v 2046 13.8 2.114 2.289 2.114 2.289 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.132 2.146 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.056 2.095 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.787 2.002 cp_fm_cholesky_decompose 22 10.9 1.885 1.900 1.885 1.900 dbcsr_complete_redistribute 325 12.2 0.556 0.570 1.595 1.708 mp_allgather_i34 2055 14.4 0.494 1.704 0.494 1.704 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.562 1.680 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.650 1.657 acc_transpose_blocks 8220 15.4 0.040 0.041 1.445 1.502 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.437 1.447 transfer_rs2pw 451 10.6 0.005 0.005 1.219 1.431 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.983 1.405 qs_create_task_list 11 7.9 0.001 0.001 1.276 1.375 generate_qs_task_list 11 8.9 0.383 0.464 1.275 1.374 mp_waitany 9240 13.8 1.123 1.333 1.123 1.333 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.218 1.249 mp_alltoall_z22v 1111 16.6 1.205 1.242 1.205 1.242 transfer_pw2rs 451 13.1 0.004 0.005 1.212 1.214 jit_kernel_multiply 7 15.4 0.821 1.203 0.821 1.203 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.749000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=809.636364, yerr=10.806334 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_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 1.362465E+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 0.035 95.529 95.530 qs_mol_dyn_low 1 2.0 0.003 0.004 95.203 95.213 qs_forces 11 3.9 0.004 0.011 95.125 95.127 qs_energies 11 4.9 0.001 0.002 90.837 90.840 scf_env_do_scf 11 5.9 0.001 0.001 80.218 80.218 velocity_verlet 10 3.0 0.002 0.002 60.248 60.255 scf_env_do_scf_inner_loop 99 6.5 0.006 0.029 49.636 49.637 dbcsr_multiply_generic 2055 12.4 0.120 0.125 31.483 31.535 qs_scf_new_mos 99 7.5 0.001 0.001 30.456 30.512 qs_scf_loop_do_ot 99 8.5 0.001 0.001 30.455 30.511 init_scf_loop 11 6.9 0.000 0.000 30.505 30.506 ot_scf_mini 99 9.5 0.002 0.002 28.590 28.640 prepare_preconditioner 11 7.9 0.000 0.000 28.290 28.296 make_preconditioner 11 8.9 0.000 0.000 28.290 28.296 make_full_inverse_cholesky 11 9.9 0.000 0.000 22.446 27.733 multiply_cannon 2055 13.4 0.343 0.366 22.856 23.810 multiply_cannon_loop 2055 14.4 0.824 0.839 20.900 21.303 cp_fm_upper_to_full 70 14.2 12.749 18.330 12.749 18.330 ot_mini 99 10.5 0.001 0.001 15.340 15.385 rebuild_ks_matrix 110 8.3 0.000 0.001 14.809 14.828 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.809 14.828 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.424 13.443 dbcsr_complete_redistribute 325 12.2 1.010 1.042 7.705 10.914 qs_ot_get_derivative 99 11.5 0.001 0.001 10.655 10.707 mp_waitall_1 84994 16.7 8.964 10.044 8.964 10.044 multiply_cannon_multrec 8220 15.4 4.091 4.248 9.757 9.927 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.535 9.749 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.828 9.009 mp_alltoall_i22 605 13.7 5.455 8.676 5.455 8.676 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.157 8.188 calculate_rho_elec 110 8.6 0.225 0.225 8.156 8.187 qs_ot_get_p 110 10.4 0.001 0.001 7.566 7.620 sum_up_and_integrate 110 10.3 0.002 0.002 7.599 7.619 integrate_v_rspace 110 11.3 0.003 0.003 7.571 7.592 make_m2s 4110 13.4 0.045 0.046 6.776 7.296 make_images 4110 14.4 0.889 0.943 6.583 7.102 cp_fm_cholesky_invert 11 10.9 6.366 6.371 6.366 6.371 init_scf_run 11 5.9 0.000 0.001 6.236 6.236 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.236 6.236 pw_transfer 1331 11.6 0.074 0.075 6.041 6.046 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.924 5.929 dbcsr_mm_accdrv_process 11614 15.7 3.838 4.127 5.519 5.734 qs_ot_p2m_diag 48 11.0 0.151 0.156 5.726 5.732 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.968 5.359 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.805 5.250 apply_single 110 13.6 0.000 0.000 4.805 5.250 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.132 5.133 fft_wrap_pw1pw2_140 451 13.1 1.280 1.286 5.022 5.026 ot_diis_step 99 11.5 0.015 0.015 4.662 4.664 make_images_data 4110 15.4 0.046 0.049 3.712 4.642 hybrid_alltoall_any 4261 16.3 0.264 0.566 3.596 4.558 density_rs2pw 110 9.6 0.004 0.004 4.458 4.480 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.336 4.336 cp_fm_diag_elpa_base 48 14.0 3.735 3.990 4.333 4.333 fft3d_ps 1111 14.6 1.299 1.307 3.976 3.984 grid_integrate_task_list 110 12.3 3.695 3.770 3.695 3.770 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.311 3.747 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.708 3.711 wfi_extrapolate 11 7.9 0.001 0.001 3.694 3.694 calculate_dm_sparse 110 9.5 0.001 0.001 3.251 3.283 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.166 3.168 multiply_cannon_sync_h2d 8220 15.4 3.129 3.156 3.129 3.156 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.106 3.135 potential_pw2rs 110 12.3 0.021 0.021 2.927 2.932 cp_fm_cholesky_decompose 22 10.9 2.794 2.822 2.794 2.822 mp_alltoall_d11v 2046 13.8 2.704 2.762 2.704 2.762 grid_collocate_task_list 110 9.6 2.680 2.695 2.680 2.695 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.544 2.556 mp_alltoall_z22v 1111 16.6 2.424 2.436 2.424 2.436 calculate_first_density_matrix 1 7.0 0.000 0.000 2.410 2.410 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.219 2.269 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.123 2.227 copy_dbcsr_to_fm 151 11.3 0.003 0.003 2.016 2.045 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.987 1.989 qs_create_task_list 11 7.9 0.003 0.005 1.897 1.941 generate_qs_task_list 11 8.9 0.731 0.783 1.894 1.937 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=95.530000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1252.000000, yerr=46.820741 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 1.103377E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 12049536 0.0% 0.0% 100.0% average stack size 0.0 0.0 564.9 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 630.280192E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10439712 MPI messages size (bytes): total size 4.526313E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.566812E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3611304 96569262080 32768 < size <= 131072 1307392 74803314688 131072 < size <= 4194304 5190114 3200150479480 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57182. MP_Allreduce 11186 792. MP_Sync 88 MP_Alltoall 2242 2551199. MP_SendRecv 24510 18752. MP_ISendRecv 24510 18752. MP_Wait 42802 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.036 229.822 229.825 qs_mol_dyn_low 1 2.0 0.003 0.004 229.366 229.382 qs_forces 11 3.9 0.005 0.005 229.275 229.276 qs_energies 11 4.9 0.002 0.003 223.427 223.441 scf_env_do_scf 11 5.9 0.001 0.002 206.047 206.051 scf_env_do_scf_inner_loop 118 6.6 0.004 0.011 181.107 181.109 velocity_verlet 10 3.0 0.003 0.004 139.021 139.022 qs_scf_new_mos 118 7.6 0.001 0.001 137.276 137.538 qs_scf_loop_do_ot 118 8.6 0.001 0.001 137.275 137.537 ot_scf_mini 118 9.6 0.003 0.003 130.346 130.569 dbcsr_multiply_generic 2529 12.6 0.183 0.187 127.971 128.691 multiply_cannon 2529 13.6 0.246 0.254 103.179 104.876 multiply_cannon_loop 2529 14.6 2.341 2.456 100.378 102.266 ot_mini 118 10.6 0.001 0.001 67.893 68.086 multiply_cannon_multrec 60696 15.6 32.305 34.412 41.865 43.995 qs_ot_get_derivative 118 11.6 0.001 0.001 42.736 42.946 qs_ot_get_p 129 10.4 0.001 0.001 39.932 40.219 rebuild_ks_matrix 129 8.3 0.001 0.001 36.016 36.410 qs_ks_build_kohn_sham_matrix 129 9.3 0.015 0.017 36.015 36.409 mp_waitall_1 269446 16.5 31.109 34.108 31.109 34.108 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.436 32.776 qs_ot_p2m_diag 84 11.4 0.080 0.099 32.129 32.202 cp_dbcsr_syevd 84 12.4 0.005 0.005 29.428 29.432 multiply_cannon_sync_h2d 60696 15.6 27.080 29.222 27.080 29.222 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.752 25.747 apply_single 129 13.6 0.001 0.001 24.752 25.746 cp_fm_diag_elpa 84 13.4 0.000 0.000 25.537 25.573 cp_fm_diag_elpa_base 84 14.4 25.382 25.453 25.531 25.566 ot_diis_step 118 11.6 0.008 0.008 24.915 24.917 init_scf_loop 11 6.9 0.000 0.001 24.849 24.850 prepare_preconditioner 11 7.9 0.000 0.000 19.985 20.045 make_preconditioner 11 8.9 0.000 0.000 19.985 20.045 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 19.599 19.772 make_full_inverse_cholesky 11 9.9 0.000 0.000 19.152 19.378 multiply_cannon_metrocomm3 60696 15.6 0.123 0.128 15.539 17.596 make_m2s 5058 13.6 0.107 0.113 15.765 16.176 sum_up_and_integrate 129 10.3 0.002 0.005 16.073 16.089 integrate_v_rspace 129 11.3 0.003 0.004 16.014 16.031 make_images 5058 14.6 0.404 0.422 15.579 15.999 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.613 15.716 calculate_rho_elec 129 8.7 0.045 0.064 15.613 15.715 init_scf_run 11 5.9 0.000 0.001 13.009 13.010 scf_env_initial_rho_setup 11 6.9 0.000 0.001 13.009 13.010 cp_fm_cholesky_invert 11 10.9 11.415 11.424 11.415 11.424 dbcsr_mm_accdrv_process 125516 16.2 4.461 4.871 9.099 9.921 density_rs2pw 129 9.7 0.006 0.007 9.171 9.786 wfi_extrapolate 11 7.9 0.001 0.001 9.721 9.721 multiply_cannon_metrocomm1 60696 15.6 0.095 0.099 6.908 9.490 pw_transfer 1559 11.6 0.076 0.097 9.039 9.281 fft_wrap_pw1pw2 1301 12.7 0.010 0.013 8.835 9.079 mp_sum_l 8016 12.9 7.745 9.064 7.745 9.064 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.812 8.941 calculate_dm_sparse 129 9.5 0.001 0.001 8.779 8.897 make_images_data 5058 15.6 0.068 0.072 7.588 8.574 qs_ot_get_orbitals 118 10.6 0.001 0.001 8.179 8.258 grid_integrate_task_list 129 12.3 7.029 7.720 7.029 7.720 hybrid_alltoall_any 5245 16.5 0.299 2.286 6.501 7.471 fft3d_ps 1301 14.7 2.213 2.860 6.845 7.224 fft_wrap_pw1pw2_140 527 13.2 1.115 1.176 6.811 7.056 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.755 6.768 potential_pw2rs 129 12.3 0.009 0.010 6.370 6.449 mp_alltoall_d11v 2429 14.1 4.863 6.396 4.863 6.396 cp_fm_cholesky_decompose 22 10.9 6.364 6.381 6.364 6.381 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.846 5.968 grid_collocate_task_list 129 9.7 4.879 5.289 4.879 5.289 transfer_rs2pw 527 10.6 0.008 0.009 4.552 5.253 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=229.825000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.545455, yerr=5.929336 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 840.564736E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 58100. MP_Allreduce 11103 959. MP_Sync 87 MP_Alltoall 1969 5161471. 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 0.015 0.034 208.077 208.079 qs_mol_dyn_low 1 2.0 0.003 0.004 207.553 207.569 qs_forces 11 3.9 0.004 0.005 207.441 207.443 qs_energies 11 4.9 0.004 0.005 200.549 200.562 scf_env_do_scf 11 5.9 0.001 0.001 183.414 183.424 scf_env_do_scf_inner_loop 117 6.6 0.003 0.011 146.299 146.302 velocity_verlet 10 3.0 0.001 0.002 130.193 130.195 qs_scf_new_mos 117 7.6 0.001 0.001 104.621 105.144 qs_scf_loop_do_ot 117 8.6 0.001 0.001 104.620 105.144 dbcsr_multiply_generic 2507 12.6 0.190 0.196 100.183 101.417 ot_scf_mini 117 9.6 0.004 0.004 99.706 100.259 multiply_cannon 2507 13.6 0.508 0.569 78.161 82.721 multiply_cannon_loop 2507 14.6 1.577 1.641 74.409 77.081 ot_mini 117 10.6 0.001 0.001 52.126 52.651 mp_waitall_1 214728 16.6 25.636 40.393 25.636 40.393 multiply_cannon_multrec 30084 15.6 21.374 26.722 31.976 37.692 init_scf_loop 11 6.9 0.000 0.000 37.015 37.016 rebuild_ks_matrix 128 8.3 0.001 0.001 33.884 34.481 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 33.884 34.480 prepare_preconditioner 11 7.9 0.000 0.000 32.582 32.630 make_preconditioner 11 8.9 0.000 0.000 32.582 32.630 make_full_inverse_cholesky 11 9.9 0.000 0.000 31.223 31.815 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.540 31.076 qs_ot_get_derivative 117 11.6 0.001 0.002 30.075 30.630 qs_ot_get_p 128 10.4 0.001 0.001 29.115 29.770 multiply_cannon_metrocomm3 30084 15.6 0.097 0.102 15.244 28.900 qs_ot_p2m_diag 83 11.4 0.189 0.217 23.438 23.476 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.144 23.192 apply_single 128 13.6 0.001 0.001 22.144 23.192 cp_dbcsr_syevd 83 12.4 0.006 0.006 22.103 22.106 ot_diis_step 117 11.6 0.014 0.015 21.869 21.871 multiply_cannon_sync_h2d 30084 15.6 18.317 20.604 18.317 20.604 cp_fm_cholesky_invert 11 10.9 18.963 18.975 18.963 18.975 cp_fm_diag_elpa 83 13.4 0.000 0.000 18.223 18.256 cp_fm_diag_elpa_base 83 14.4 17.901 18.039 18.217 18.245 make_m2s 5014 13.6 0.090 0.096 15.909 17.557 make_images 5014 14.6 1.178 1.373 15.698 17.343 sum_up_and_integrate 128 10.3 0.002 0.004 15.356 15.390 integrate_v_rspace 128 11.3 0.003 0.003 15.296 15.334 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.220 15.253 calculate_rho_elec 128 8.7 0.087 0.104 15.219 15.252 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 12.075 12.485 init_scf_run 11 5.9 0.000 0.001 12.161 12.163 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.161 12.162 make_images_data 5014 15.6 0.068 0.077 9.119 11.257 dbcsr_mm_accdrv_process 62242 16.2 5.452 6.420 10.049 10.627 hybrid_alltoall_any 5200 16.5 0.356 1.554 7.642 10.452 multiply_cannon_metrocomm4 27577 15.6 0.104 0.116 3.727 10.313 mp_irecv_dv 69486 16.3 3.527 9.924 3.527 9.924 density_rs2pw 128 9.7 0.006 0.007 8.810 9.344 pw_transfer 1547 11.6 0.085 0.098 9.220 9.307 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.996 9.087 wfi_extrapolate 11 7.9 0.001 0.001 8.886 8.886 cp_fm_cholesky_decompose 22 10.9 8.566 8.654 8.566 8.654 grid_integrate_task_list 128 12.3 7.144 7.638 7.144 7.638 fft_wrap_pw1pw2_140 523 13.2 1.210 1.235 7.503 7.618 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.618 7.429 calculate_dm_sparse 128 9.5 0.001 0.001 6.598 6.751 fft3d_ps 1291 14.7 2.842 3.032 6.559 6.643 mp_sum_l 7950 12.9 4.323 6.419 4.323 6.419 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.308 6.315 potential_pw2rs 128 12.3 0.015 0.017 5.783 5.819 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.609 5.685 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.313 5.473 mp_alltoall_d11v 2415 14.1 4.536 5.453 4.536 5.453 grid_collocate_task_list 128 9.7 5.000 5.417 5.000 5.417 mp_allgather_i34 2507 14.6 2.261 5.317 2.261 5.317 transfer_rs2pw 523 10.6 0.007 0.007 4.332 5.004 dbcsr_complete_redistribute 395 12.7 0.813 0.933 3.592 4.556 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=208.079000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=800.090909, yerr=4.316028 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 957.550592E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58351. MP_Allreduce 11057 1000. 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 0.040 0.088 202.546 202.548 qs_mol_dyn_low 1 2.0 0.003 0.005 201.972 202.005 qs_forces 11 3.9 0.004 0.005 201.859 201.863 qs_energies 11 4.9 0.002 0.008 195.038 195.055 scf_env_do_scf 11 5.9 0.001 0.003 178.177 178.178 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 136.314 136.315 velocity_verlet 10 3.0 0.003 0.004 129.086 129.088 qs_scf_new_mos 117 7.6 0.001 0.001 95.887 96.353 qs_scf_loop_do_ot 117 8.6 0.001 0.001 95.886 96.352 ot_scf_mini 117 9.6 0.003 0.004 91.445 91.930 dbcsr_multiply_generic 2507 12.6 0.183 0.191 86.982 88.766 multiply_cannon 2507 13.6 0.509 0.529 64.355 68.655 multiply_cannon_loop 2507 14.6 1.133 1.203 60.695 65.078 ot_mini 117 10.6 0.001 0.001 46.903 47.428 init_scf_loop 11 6.9 0.001 0.005 41.735 41.736 mp_waitall_1 170520 16.6 28.088 39.609 28.088 39.609 prepare_preconditioner 11 7.9 0.000 0.001 37.587 37.645 make_preconditioner 11 8.9 0.000 0.002 37.587 37.645 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.081 36.552 rebuild_ks_matrix 128 8.3 0.001 0.001 32.027 32.756 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 32.026 32.755 qs_ot_get_p 128 10.4 0.001 0.001 29.262 29.839 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.909 29.550 multiply_cannon_metrocomm3 20056 15.6 0.063 0.067 16.696 28.252 qs_ot_get_derivative 117 11.6 0.001 0.002 26.743 27.248 multiply_cannon_multrec 20056 15.6 12.975 15.959 22.290 25.359 qs_ot_p2m_diag 83 11.4 0.266 0.273 24.244 24.288 cp_dbcsr_syevd 83 12.4 0.005 0.006 22.805 22.807 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.027 21.318 apply_single 128 13.6 0.001 0.001 20.027 21.317 ot_diis_step 117 11.6 0.018 0.018 19.969 19.970 cp_fm_cholesky_invert 11 10.9 18.573 18.587 18.573 18.587 cp_fm_diag_elpa 83 13.4 0.000 0.000 18.531 18.555 cp_fm_diag_elpa_base 83 14.4 18.050 18.249 18.526 18.549 make_m2s 5014 13.6 0.080 0.084 16.052 16.995 make_images 5014 14.6 1.166 1.271 15.821 16.758 sum_up_and_integrate 128 10.3 0.002 0.003 15.618 15.646 integrate_v_rspace 128 11.3 0.003 0.004 15.558 15.581 multiply_cannon_sync_h2d 20056 15.6 13.386 15.493 13.386 15.493 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.419 15.493 calculate_rho_elec 128 8.7 0.130 0.145 15.418 15.492 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.133 11.516 init_scf_run 11 5.9 0.000 0.001 11.376 11.376 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.375 11.376 make_images_data 5014 15.6 0.063 0.072 9.361 10.800 hybrid_alltoall_any 5200 16.5 0.455 2.056 8.096 9.991 multiply_cannon_metrocomm4 17549 15.6 0.066 0.077 3.459 9.730 cp_fm_cholesky_decompose 22 10.9 9.708 9.729 9.708 9.729 pw_transfer 1547 11.6 0.086 0.103 9.419 9.540 mp_irecv_dv 50230 16.2 3.333 9.471 3.333 9.471 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.194 9.321 dbcsr_mm_accdrv_process 41502 16.2 5.779 5.956 8.812 8.943 density_rs2pw 128 9.7 0.006 0.006 8.460 8.760 wfi_extrapolate 11 7.9 0.001 0.001 8.311 8.311 grid_integrate_task_list 128 12.3 7.300 7.806 7.300 7.806 fft_wrap_pw1pw2_140 523 13.2 1.302 1.337 7.606 7.744 cp_fm_upper_to_full 105 14.8 5.918 7.557 5.918 7.557 dbcsr_complete_redistribute 395 12.7 1.171 1.200 5.249 7.179 mp_sum_l 7950 12.9 4.791 6.927 4.791 6.927 fft3d_ps 1291 14.7 2.762 2.980 6.572 6.665 calculate_dm_sparse 128 9.5 0.001 0.001 6.049 6.150 mp_alltoall_d11v 2415 14.1 5.286 6.018 5.286 6.018 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.268 5.961 potential_pw2rs 128 12.3 0.020 0.022 5.850 5.879 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.786 5.793 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.770 5.695 grid_collocate_task_list 128 9.7 5.162 5.583 5.162 5.583 mp_allgather_i34 2507 14.6 2.153 5.269 2.153 5.269 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.587 4.795 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.467 4.548 transfer_rs2pw 523 10.6 0.006 0.007 3.945 4.374 mp_alltoall_i22 716 14.1 2.256 4.337 2.256 4.337 transfer_fm_to_dbcsr 11 9.9 0.018 0.025 2.487 4.303 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=202.548000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=905.818182, yerr=14.357881 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.154310E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922227920 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4039 57682. MP_Allreduce 11174 1079. MP_Sync 88 MP_Alltoall 1724 12509605. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.033 218.057 218.057 qs_mol_dyn_low 1 2.0 0.003 0.005 217.623 217.637 qs_forces 11 3.9 0.004 0.004 217.492 217.500 qs_energies 11 4.9 0.003 0.014 210.220 210.231 scf_env_do_scf 11 5.9 0.001 0.002 192.394 192.404 velocity_verlet 10 3.0 0.003 0.005 143.850 143.854 scf_env_do_scf_inner_loop 118 6.6 0.003 0.010 137.032 137.037 qs_scf_new_mos 118 7.6 0.001 0.001 97.784 98.153 qs_scf_loop_do_ot 118 8.6 0.001 0.001 97.783 98.153 ot_scf_mini 118 9.6 0.003 0.004 92.950 93.279 dbcsr_multiply_generic 2529 12.6 0.192 0.198 85.552 86.422 multiply_cannon 2529 13.6 0.552 0.590 57.576 60.353 init_scf_loop 11 6.9 0.001 0.003 55.164 55.169 multiply_cannon_loop 2529 14.6 1.847 1.937 53.059 55.166 prepare_preconditioner 11 7.9 0.000 0.000 50.869 50.897 make_preconditioner 11 8.9 0.000 0.002 50.869 50.897 make_full_inverse_cholesky 11 9.9 0.012 0.022 44.239 49.532 ot_mini 118 10.6 0.001 0.001 46.783 47.125 qs_ot_get_p 129 10.4 0.001 0.001 31.609 31.983 multiply_cannon_multrec 30348 15.6 13.731 18.801 26.717 31.464 rebuild_ks_matrix 129 8.3 0.001 0.001 30.839 31.135 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 30.839 31.134 mp_waitall_1 149172 16.7 19.944 30.149 19.944 30.149 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.886 28.153 qs_ot_get_derivative 118 11.6 0.002 0.002 26.427 26.764 qs_ot_p2m_diag 84 11.4 0.347 0.395 26.455 26.518 cp_dbcsr_syevd 84 12.4 0.005 0.006 24.725 24.728 make_m2s 5058 13.6 0.098 0.102 22.783 23.839 make_images 5058 14.6 1.985 2.323 22.468 23.525 cp_fm_cholesky_invert 11 10.9 21.252 21.261 21.252 21.261 cp_fm_diag_elpa 84 13.4 0.000 0.000 20.542 20.556 cp_fm_diag_elpa_base 84 14.4 19.434 19.834 20.533 20.547 ot_diis_step 118 11.6 0.018 0.019 20.227 20.228 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.605 20.192 apply_single 129 13.6 0.001 0.001 19.604 20.192 cp_fm_upper_to_full 106 14.8 11.949 17.733 11.949 17.733 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.478 15.542 calculate_rho_elec 129 8.7 0.174 0.191 15.477 15.541 multiply_cannon_metrocomm3 30348 15.6 0.050 0.053 6.573 15.476 sum_up_and_integrate 129 10.3 0.002 0.003 15.199 15.227 integrate_v_rspace 129 11.3 0.003 0.004 15.139 15.169 make_images_data 5058 15.6 0.068 0.076 12.412 14.274 dbcsr_complete_redistribute 397 12.7 1.510 1.641 9.929 13.897 dbcsr_mm_accdrv_process 62780 16.2 8.494 9.385 12.554 13.065 hybrid_alltoall_any 5245 16.5 0.538 2.227 10.838 12.960 copy_fm_to_dbcsr 210 11.7 0.001 0.002 8.243 12.201 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 11.389 11.646 init_scf_run 11 5.9 0.001 0.006 11.589 11.590 scf_env_initial_rho_setup 11 6.9 0.000 0.002 11.588 11.590 multiply_cannon_sync_h2d 30348 15.6 10.541 11.337 10.541 11.337 cp_fm_cholesky_decompose 22 10.9 10.501 10.593 10.501 10.593 transfer_fm_to_dbcsr 11 9.9 0.001 0.005 6.608 10.450 mp_alltoall_i22 720 14.1 5.983 10.053 5.983 10.053 pw_transfer 1559 11.6 0.086 0.099 9.463 9.549 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.235 9.325 density_rs2pw 129 9.7 0.006 0.006 8.182 8.630 wfi_extrapolate 11 7.9 0.001 0.001 8.450 8.450 fft_wrap_pw1pw2_140 527 13.2 1.455 1.495 7.931 8.037 grid_integrate_task_list 129 12.3 7.606 7.930 7.606 7.930 multiply_cannon_metrocomm4 25290 15.6 0.085 0.096 2.832 7.103 mp_irecv_dv 76751 16.2 2.677 6.826 2.677 6.826 calculate_dm_sparse 129 9.5 0.001 0.001 6.574 6.655 mp_alltoall_d11v 2429 14.1 5.964 6.572 5.964 6.572 fft3d_ps 1301 14.7 2.869 2.950 6.393 6.442 grid_collocate_task_list 129 9.7 5.329 5.709 5.329 5.709 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.647 5.696 potential_pw2rs 129 12.3 0.024 0.024 5.244 5.260 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.053 5.147 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.622 4.668 mp_sum_l 8016 12.9 3.249 4.558 3.249 4.558 qs_energies_init_hamiltonians 11 5.9 0.003 0.009 4.546 4.549 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.408 4.506 mp_allgather_i34 2529 14.6 2.585 4.469 2.585 4.469 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=218.057000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1089.000000, yerr=19.221200 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.617355E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108638608 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58273. MP_Allreduce 11070 1167. MP_Sync 86 MP_Alltoall 1700 18828160. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.054 170.626 170.634 qs_mol_dyn_low 1 2.0 0.003 0.003 170.177 170.197 qs_forces 11 3.9 0.004 0.004 170.064 170.068 qs_energies 11 4.9 0.002 0.002 162.705 162.712 scf_env_do_scf 11 5.9 0.001 0.001 145.477 145.486 velocity_verlet 10 3.0 0.002 0.002 110.786 110.790 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 109.920 109.921 qs_scf_new_mos 116 7.6 0.001 0.001 73.462 73.577 qs_scf_loop_do_ot 116 8.6 0.001 0.001 73.461 73.576 dbcsr_multiply_generic 2485 12.5 0.179 0.191 72.155 72.573 ot_scf_mini 116 9.6 0.003 0.004 69.045 69.160 multiply_cannon 2485 13.5 0.556 0.578 53.510 56.702 multiply_cannon_loop 2485 14.5 0.800 0.845 50.394 51.433 ot_mini 116 10.6 0.001 0.001 37.822 37.932 init_scf_loop 11 6.9 0.000 0.000 35.406 35.407 prepare_preconditioner 11 7.9 0.000 0.000 31.579 31.610 make_preconditioner 11 8.9 0.000 0.000 31.579 31.610 mp_waitall_1 124680 16.7 25.050 31.556 25.050 31.556 make_full_inverse_cholesky 11 9.9 0.014 0.028 29.496 29.740 rebuild_ks_matrix 127 8.3 0.001 0.001 28.772 28.902 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 28.771 28.901 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.246 26.359 multiply_cannon_multrec 9940 15.5 10.257 15.476 17.807 21.626 qs_ot_get_derivative 116 11.6 0.001 0.002 20.606 20.718 multiply_cannon_metrocomm3 9940 15.5 0.025 0.027 12.984 20.543 qs_ot_get_p 127 10.4 0.001 0.001 18.531 18.660 cp_fm_cholesky_invert 11 10.9 18.062 18.067 18.062 18.067 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.267 17.431 apply_single 127 13.6 0.001 0.001 17.267 17.431 ot_diis_step 116 11.6 0.019 0.020 17.142 17.143 make_m2s 4970 13.5 0.066 0.068 14.744 15.614 make_images 4970 14.5 2.160 2.582 14.439 15.311 qs_ot_p2m_diag 82 11.4 0.490 0.495 14.638 14.656 sum_up_and_integrate 127 10.3 0.002 0.002 13.986 14.036 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.947 13.983 calculate_rho_elec 127 8.7 0.254 0.265 13.947 13.982 integrate_v_rspace 127 11.3 0.004 0.004 13.926 13.977 cp_dbcsr_syevd 82 12.4 0.005 0.006 13.460 13.461 multiply_cannon_sync_h2d 9940 15.5 10.627 10.933 10.627 10.933 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.562 10.573 cp_fm_diag_elpa_base 82 14.4 10.326 10.401 10.558 10.569 init_scf_run 11 5.9 0.000 0.001 10.389 10.390 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.389 10.389 make_images_data 4970 15.5 0.055 0.064 8.464 9.856 hybrid_alltoall_any 5155 16.4 0.836 3.742 8.313 9.466 pw_transfer 1535 11.6 0.084 0.095 9.224 9.258 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.003 9.044 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.090 8.168 grid_integrate_task_list 127 12.3 7.745 8.056 7.745 8.056 cp_fm_cholesky_decompose 22 10.9 7.891 8.029 7.891 8.029 fft_wrap_pw1pw2_140 519 13.2 1.789 1.827 7.966 8.011 dbcsr_mm_accdrv_process 20590 16.1 3.292 4.303 7.204 7.906 multiply_cannon_metrocomm1 9940 15.5 0.030 0.030 4.332 7.654 wfi_extrapolate 11 7.9 0.001 0.001 7.349 7.349 density_rs2pw 127 9.7 0.006 0.006 6.683 6.925 calculate_dm_sparse 127 9.5 0.001 0.001 6.066 6.139 grid_collocate_task_list 127 9.7 5.537 5.862 5.537 5.862 fft3d_ps 1281 14.7 2.748 2.821 5.759 5.790 dbcsr_complete_redistribute 393 12.7 2.096 2.153 5.237 5.629 multiply_cannon_metrocomm4 7455 15.5 0.026 0.030 1.862 5.458 mp_irecv_dv 28618 15.9 1.823 5.373 1.823 5.373 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.255 5.255 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.166 5.188 mp_alltoall_d11v 2401 14.1 4.637 5.052 4.637 5.052 mp_allgather_i34 2485 14.5 1.309 4.660 1.309 4.660 potential_pw2rs 127 12.3 0.026 0.026 4.209 4.214 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.053 4.099 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.585 3.886 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.630 3.714 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.385 3.701 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.650 3.697 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.464 3.561 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.425 3.444 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.634000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1513.272727, yerr=38.026524 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456868864 0.0% 0.0% 100.0% flops 32 x 32 x 32 1971389988864 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992007581696 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753954734080 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 11613074706432 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176888320 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176888320 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911117062144 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.241596E+12 0.0% 0.0% 100.0% flops max/rank 11.787141E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806515136 0.0% 0.0% 100.0% number of processed stacks 1981760 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.6 marketing flops 145.659521E+12 ------------------------------------------------------------------------------- # multiplications 2533 max memory usage/rank 3.216749E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101320 MPI messages size (bytes): total size 1.145104E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.301851E+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 45808 35567697920 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592330608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4065 58253. MP_Allreduce 11240 1499. MP_Sync 86 MP_Alltoall 1724 36993608. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.052 290.748 290.755 qs_mol_dyn_low 1 2.0 0.003 0.004 290.209 290.221 qs_forces 11 3.9 0.005 0.005 290.103 290.108 qs_energies 11 4.9 0.003 0.006 281.235 281.252 scf_env_do_scf 11 5.9 0.001 0.001 258.610 258.620 velocity_verlet 10 3.0 0.002 0.002 210.831 210.840 scf_env_do_scf_inner_loop 118 6.6 0.004 0.010 135.072 135.074 init_scf_loop 11 6.9 0.000 0.000 123.266 123.268 prepare_preconditioner 11 7.9 0.000 0.000 118.443 118.460 make_preconditioner 11 8.9 0.000 0.000 118.443 118.460 make_full_inverse_cholesky 11 9.9 0.037 0.039 95.113 115.629 qs_scf_new_mos 118 7.6 0.001 0.001 90.776 90.824 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.775 90.823 ot_scf_mini 118 9.6 0.004 0.004 85.856 85.862 dbcsr_multiply_generic 2533 12.6 0.214 0.226 82.735 83.176 cp_fm_upper_to_full 104 14.8 51.907 74.222 51.907 74.222 multiply_cannon 2533 13.6 0.684 0.746 59.255 60.957 multiply_cannon_loop 2533 14.6 1.055 1.095 55.140 56.284 ot_mini 118 10.6 0.001 0.001 44.578 44.586 dbcsr_complete_redistribute 393 12.7 4.003 4.048 29.356 41.843 copy_fm_to_dbcsr 208 11.6 0.001 0.001 25.803 38.315 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.288 35.696 rebuild_ks_matrix 129 8.3 0.001 0.001 34.410 34.424 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 34.409 34.424 cp_fm_cholesky_invert 11 10.9 33.750 33.757 33.750 33.757 mp_alltoall_i22 716 14.1 21.047 33.514 21.047 33.514 mp_waitall_1 104740 16.8 28.198 32.729 28.198 32.729 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.939 31.965 qs_ot_get_p 129 10.4 0.001 0.001 25.936 25.977 qs_ot_get_derivative 118 11.6 0.002 0.002 25.058 25.062 qs_ot_p2m_diag 82 11.4 0.869 0.874 21.797 21.827 make_m2s 5066 13.6 0.078 0.080 18.860 20.381 cp_dbcsr_syevd 82 12.4 0.006 0.006 20.032 20.033 multiply_cannon_metrocomm3 10132 15.6 0.025 0.026 18.821 19.976 make_images 5066 14.6 3.105 3.295 18.375 19.897 ot_diis_step 118 11.6 0.022 0.022 19.494 19.495 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 18.787 18.868 apply_single 129 13.6 0.001 0.001 18.786 18.868 multiply_cannon_multrec 10132 15.6 10.396 12.200 18.244 18.330 qs_rho_update_rho_low 129 7.7 0.001 0.001 16.983 16.997 calculate_rho_elec 129 8.7 0.485 0.486 16.982 16.997 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.851 16.851 cp_fm_diag_elpa_base 82 14.4 12.489 14.030 16.847 16.847 sum_up_and_integrate 129 10.3 0.002 0.002 16.339 16.431 integrate_v_rspace 129 11.3 0.005 0.005 16.276 16.368 multiply_cannon_sync_h2d 10132 15.6 14.420 14.473 14.420 14.473 hybrid_alltoall_any 5251 16.5 1.317 3.073 10.859 13.214 make_images_data 5066 15.6 0.066 0.071 10.628 12.893 init_scf_run 11 5.9 0.000 0.001 12.263 12.266 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.263 12.265 pw_transfer 1559 11.6 0.096 0.097 11.666 11.670 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 11.426 11.431 fft_wrap_pw1pw2_140 527 13.2 3.071 3.167 10.152 10.159 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.641 9.643 dbcsr_mm_accdrv_process 20950 16.1 4.187 6.224 7.597 9.508 cp_fm_cholesky_decompose 22 10.9 9.233 9.291 9.233 9.291 wfi_extrapolate 11 7.9 0.001 0.001 9.132 9.133 grid_integrate_task_list 129 12.3 8.615 8.820 8.615 8.820 qs_energies_init_hamiltonians 11 5.9 0.004 0.005 8.153 8.154 density_rs2pw 129 9.7 0.005 0.005 7.848 7.894 mp_alltoall_d11v 2417 14.1 7.178 7.320 7.178 7.320 calculate_dm_sparse 129 9.5 0.001 0.001 6.790 6.855 fft3d_ps 1301 14.7 2.879 2.892 6.630 6.699 grid_collocate_task_list 129 9.7 6.464 6.543 6.464 6.543 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.469 6.542 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.219 6.314 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=290.755000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2930.909091, yerr=152.610291 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_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 1.259876E+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 6956154. 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 0.014 0.042 85.391 85.392 qs_energies 1 2.0 0.000 0.000 84.861 84.873 ls_scf 1 3.0 0.000 0.000 83.930 83.943 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.754 72.890 multiply_cannon 111 7.7 0.018 0.020 55.980 57.170 multiply_cannon_loop 111 8.7 0.218 0.239 52.497 53.707 ls_scf_main 1 4.0 0.000 0.000 52.456 52.456 density_matrix_trs4 2 5.0 0.002 0.003 46.899 46.990 ls_scf_init_scf 1 4.0 0.000 0.000 28.439 28.441 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.313 27.361 mp_waitall_1 11031 10.9 22.575 25.360 22.575 25.360 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.182 25.201 multiply_cannon_multrec 2664 9.7 8.169 8.906 15.638 17.420 multiply_cannon_sync_h2d 2664 9.7 13.447 14.825 13.447 14.825 make_m2s 222 7.7 0.009 0.011 13.199 13.732 make_images 222 8.7 0.099 0.110 13.177 13.711 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.708 12.921 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.503 8.721 make_images_data 222 9.7 0.004 0.005 7.698 8.258 hybrid_alltoall_any 227 10.6 0.215 1.830 6.612 8.107 dbcsr_mm_accdrv_process 4760 10.4 0.585 0.706 7.089 8.084 dbcsr_mm_accdrv_process_sort 4760 11.4 6.305 7.208 6.305 7.208 calculate_norms 4752 9.8 5.518 6.151 5.518 6.151 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.017 5.165 mp_sum_l 887 5.1 3.043 4.652 3.043 4.652 make_images_sizes 222 9.7 0.000 0.000 0.774 3.794 mp_alltoall_i44 222 10.7 0.773 3.794 0.773 3.794 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.256 3.650 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.049 3.605 mp_irecv_dv 6231 10.9 2.034 3.579 2.034 3.579 arnoldi_extremal 4 6.8 0.000 0.000 3.299 3.319 arnoldi_normal_ev 4 7.8 0.001 0.003 3.299 3.319 build_subspace 16 8.4 0.009 0.012 3.208 3.210 ls_scf_post 1 4.0 0.000 0.000 3.035 3.048 ls_scf_store_result 1 5.0 0.000 0.000 2.856 2.896 dbcsr_special_finalize 555 9.7 0.005 0.006 2.381 2.861 dbcsr_merge_single_wm 555 10.7 0.479 0.630 2.372 2.852 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.363 2.621 make_images_pack 222 9.7 2.199 2.616 2.201 2.617 dbcsr_sort_data 658 11.4 2.151 2.568 2.151 2.568 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.477 2.071 2.479 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.298 2.356 buffer_matrices_ensure_size 222 8.7 1.796 2.179 1.796 2.179 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.765 1.766 rebuild_ks_matrix 3 7.3 0.000 0.000 1.755 1.756 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.755 1.756 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.392000, 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/dfc565023ea3187a044d14b931fefe7b58c72354_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 2.146120E+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 0.025 0.043 90.980 90.981 qs_energies 1 2.0 0.000 0.001 90.532 90.536 ls_scf 1 3.0 0.000 0.001 89.192 89.196 dbcsr_multiply_generic 111 6.7 0.015 0.017 75.180 75.472 multiply_cannon 111 7.7 0.029 0.043 53.015 58.114 ls_scf_main 1 4.0 0.000 0.001 55.351 55.355 multiply_cannon_loop 111 8.7 0.135 0.148 50.128 53.766 density_matrix_trs4 2 5.0 0.002 0.005 49.586 49.830 mp_waitall_1 9105 10.9 21.112 30.457 21.112 30.457 ls_scf_init_scf 1 4.0 0.000 0.002 30.104 30.105 ls_scf_init_matrix_S 1 5.0 0.001 0.013 28.883 28.994 multiply_cannon_multrec 1332 9.7 13.314 16.693 22.688 27.176 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.442 26.456 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.602 21.721 make_m2s 222 7.7 0.006 0.007 15.389 16.017 make_images 222 8.7 1.371 1.696 15.358 15.987 dbcsr_mm_accdrv_process 4041 10.4 0.359 0.513 8.973 10.508 make_images_data 222 9.7 0.004 0.005 8.973 9.989 dbcsr_mm_accdrv_process_sort 4041 11.4 8.491 9.988 8.491 9.988 hybrid_alltoall_any 227 10.6 0.542 2.541 8.285 9.832 mp_sum_l 887 5.1 5.314 8.811 5.314 8.811 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.221 7.810 mp_irecv_dv 3311 11.0 3.201 7.742 3.201 7.742 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.041 7.008 calculate_norms 2376 9.8 6.053 6.713 6.053 6.713 multiply_cannon_sync_h2d 1332 9.7 4.892 6.384 4.892 6.384 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.974 5.167 arnoldi_extremal 4 6.8 0.000 0.000 4.697 4.714 arnoldi_normal_ev 4 7.8 0.002 0.005 4.696 4.713 build_subspace 16 8.4 0.014 0.021 4.443 4.445 ls_scf_post 1 4.0 0.000 0.001 3.737 3.741 ls_scf_store_result 1 5.0 0.000 0.000 3.435 3.560 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.192 3.424 dbcsr_matrix_vector_mult_local 304 10.0 2.780 3.264 2.782 3.265 mp_allgather_i34 111 8.7 0.878 3.166 0.878 3.166 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.179 2.874 ls_scf_dm_to_ks 2 5.0 0.000 0.005 2.627 2.722 dbcsr_data_new 4174 10.1 2.107 2.391 2.107 2.391 make_images_pack 222 9.7 1.818 2.128 1.820 2.130 dbcsr_sort_data 436 11.2 1.841 2.107 1.841 2.107 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.883 1.885 rebuild_ks_matrix 3 7.3 0.000 0.000 1.870 1.872 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.870 1.872 make_images_sizes 222 9.7 0.000 0.000 0.650 1.859 mp_alltoall_i44 222 10.7 0.650 1.858 0.650 1.858 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.981000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1776.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_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 2.848731E+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 0.056 91.845 91.846 qs_energies 1 2.0 0.000 0.001 91.207 91.211 ls_scf 1 3.0 0.000 0.001 89.738 89.742 dbcsr_multiply_generic 111 6.7 0.016 0.016 74.440 74.683 ls_scf_main 1 4.0 0.000 0.001 55.981 55.984 multiply_cannon 111 7.7 0.035 0.067 51.596 55.966 multiply_cannon_loop 111 8.7 0.116 0.130 48.811 52.259 density_matrix_trs4 2 5.0 0.002 0.006 50.061 50.240 mp_waitall_1 7281 11.0 23.008 32.612 23.008 32.612 ls_scf_init_scf 1 4.0 0.000 0.002 30.061 30.063 ls_scf_init_matrix_S 1 5.0 0.000 0.002 28.882 28.944 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.558 26.571 multiply_cannon_multrec 888 9.7 12.713 15.304 21.387 24.592 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.534 21.836 make_m2s 222 7.7 0.006 0.007 16.383 17.125 make_images 222 8.7 1.580 1.855 16.345 17.088 make_images_data 222 9.7 0.004 0.004 9.549 10.520 hybrid_alltoall_any 227 10.6 0.640 2.956 8.986 10.128 dbcsr_mm_accdrv_process 3754 10.4 0.293 0.495 8.197 9.478 dbcsr_mm_accdrv_process_sort 3754 11.4 7.761 8.983 7.761 8.983 mp_sum_l 887 5.1 4.776 8.048 4.776 8.048 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.566 7.194 multiply_cannon_sync_h2d 888 9.7 6.079 7.111 6.079 7.111 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.449 7.010 mp_irecv_dv 2335 11.1 2.433 6.954 2.433 6.954 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.605 6.339 arnoldi_extremal 4 6.8 0.000 0.000 5.154 5.169 arnoldi_normal_ev 4 7.8 0.001 0.005 5.154 5.169 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.760 4.943 build_subspace 16 8.4 0.014 0.020 4.850 4.855 calculate_norms 1584 9.8 4.332 4.744 4.332 4.744 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.498 3.832 mp_allgather_i34 111 8.7 0.892 3.779 0.892 3.779 ls_scf_post 1 4.0 0.000 0.001 3.697 3.702 dbcsr_matrix_vector_mult_local 304 10.0 3.071 3.644 3.073 3.646 ls_scf_store_result 1 5.0 0.000 0.000 3.431 3.520 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.848 2.946 dbcsr_data_new 4116 9.9 2.112 2.447 2.112 2.447 make_images_sizes 222 9.7 0.000 0.000 1.013 2.329 mp_alltoall_i44 222 10.7 1.012 2.328 1.012 2.328 dbcsr_sort_data 325 11.1 1.862 2.180 1.862 2.180 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.894 1.895 rebuild_ks_matrix 3 7.3 0.000 0.000 1.875 1.877 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.875 1.877 make_images_pack 222 9.7 1.623 1.874 1.626 1.877 dbcsr_finalize 304 7.8 0.026 0.032 1.599 1.860 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=91.846000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2211.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_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 3.389960E+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 0.028 0.047 97.931 97.932 qs_energies 1 2.0 0.000 0.000 97.336 97.343 ls_scf 1 3.0 0.000 0.000 95.676 95.686 dbcsr_multiply_generic 111 6.7 0.017 0.018 79.022 79.305 ls_scf_main 1 4.0 0.000 0.000 59.517 59.518 multiply_cannon 111 7.7 0.046 0.083 52.070 56.933 density_matrix_trs4 2 5.0 0.002 0.003 53.354 53.457 multiply_cannon_loop 111 8.7 0.152 0.167 46.932 50.803 ls_scf_init_scf 1 4.0 0.000 0.000 32.924 32.927 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.710 31.781 mp_waitall_1 6369 11.0 22.736 30.181 22.736 30.181 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.202 29.215 multiply_cannon_multrec 1332 9.7 14.162 17.508 22.147 24.816 make_m2s 222 7.7 0.007 0.008 20.938 22.448 make_images 222 8.7 3.149 3.611 20.888 22.401 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.345 17.111 make_images_data 222 9.7 0.004 0.004 11.627 13.260 hybrid_alltoall_any 227 10.6 0.796 3.760 10.880 12.890 dbcsr_mm_accdrv_process 3641 10.4 0.294 0.484 7.630 9.166 dbcsr_mm_accdrv_process_sort 3641 11.4 7.190 8.680 7.190 8.680 mp_sum_l 887 5.1 4.432 8.004 4.432 8.004 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.434 6.329 multiply_cannon_sync_h2d 1332 9.7 5.450 6.261 5.450 6.261 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.098 6.148 mp_irecv_dv 3229 10.9 2.075 6.082 2.075 6.082 arnoldi_extremal 4 6.8 0.000 0.000 5.583 5.599 arnoldi_normal_ev 4 7.8 0.001 0.005 5.583 5.599 build_subspace 16 8.4 0.014 0.021 5.240 5.247 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.590 4.968 mp_allgather_i34 111 8.7 2.249 4.860 2.249 4.860 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.563 4.769 calculate_norms 2376 9.8 4.208 4.607 4.208 4.607 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.772 4.046 dbcsr_matrix_vector_mult_local 304 10.0 3.251 3.744 3.253 3.746 dbcsr_sort_data 658 11.4 3.026 3.414 3.026 3.414 ls_scf_post 1 4.0 0.000 0.000 3.236 3.242 dbcsr_special_finalize 555 9.7 0.006 0.008 2.771 3.224 dbcsr_merge_single_wm 555 10.7 0.536 0.667 2.762 3.216 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.019 3.088 ls_scf_store_result 1 5.0 0.000 0.000 2.974 3.039 dbcsr_data_release 10477 10.7 1.586 2.386 1.586 2.386 dbcsr_finalize 304 7.8 0.049 0.061 1.807 1.979 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.932000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2755.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_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 4.780904E+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 0.037 0.056 92.117 92.118 qs_energies 1 2.0 0.000 0.000 91.375 91.378 ls_scf 1 3.0 0.000 0.000 89.427 89.430 dbcsr_multiply_generic 111 6.7 0.017 0.019 70.805 70.967 ls_scf_main 1 4.0 0.000 0.000 56.674 56.675 multiply_cannon 111 7.7 0.086 0.141 52.334 56.009 multiply_cannon_loop 111 8.7 0.088 0.096 49.745 51.223 density_matrix_trs4 2 5.0 0.002 0.003 49.585 49.643 ls_scf_init_scf 1 4.0 0.000 0.000 29.357 29.360 mp_waitall_1 5436 11.0 24.286 28.546 24.286 28.546 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.122 28.145 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.003 26.015 multiply_cannon_multrec 444 9.7 13.760 16.199 20.933 22.996 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.629 15.021 make_m2s 222 7.7 0.005 0.005 13.789 14.743 make_images 222 8.7 2.042 2.484 13.722 14.674 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 5.995 13.302 hybrid_alltoall_any 227 10.6 0.803 3.844 8.250 9.968 make_images_data 222 9.7 0.003 0.004 8.478 9.831 dbcsr_mm_accdrv_process 3003 10.4 0.339 0.398 6.860 7.991 multiply_cannon_sync_h2d 444 9.7 6.696 7.873 6.696 7.873 dbcsr_mm_accdrv_process_sort 3003 11.4 6.498 7.594 6.498 7.594 arnoldi_extremal 4 6.8 0.000 0.000 6.033 6.046 arnoldi_normal_ev 4 7.8 0.002 0.005 6.033 6.046 build_subspace 16 8.4 0.015 0.019 5.631 5.640 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.562 4.736 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.271 4.501 mp_sum_l 887 5.1 2.687 4.443 2.687 4.443 dbcsr_matrix_vector_mult_local 304 10.0 3.772 4.229 3.774 4.231 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.546 3.901 mp_irecv_dv 1241 11.2 1.532 3.873 1.532 3.873 calculate_norms 792 9.8 3.607 3.759 3.607 3.759 mp_allgather_i34 111 8.7 1.162 3.722 1.162 3.722 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.529 3.593 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.843 3.460 make_images_sizes 222 9.7 0.000 0.000 0.924 3.451 mp_alltoall_i44 222 10.7 0.924 3.451 0.924 3.451 ls_scf_post 1 4.0 0.000 0.000 3.395 3.398 ls_scf_store_result 1 5.0 0.000 0.000 3.186 3.217 dbcsr_finalize 304 7.8 0.062 0.077 2.194 2.306 dbcsr_data_new 4608 9.7 1.780 2.241 1.780 2.241 dbcsr_merge_all 275 8.9 0.480 0.528 2.053 2.147 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.021 2.022 rebuild_ks_matrix 3 7.3 0.000 0.000 1.988 1.989 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.988 1.989 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 1.932 1.932 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.118000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3712.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_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 8.878170E+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 0.073 0.090 105.272 105.273 qs_energies 1 2.0 0.000 0.000 103.938 103.949 ls_scf 1 3.0 0.000 0.000 101.004 101.015 dbcsr_multiply_generic 111 6.7 0.024 0.028 74.653 74.789 ls_scf_main 1 4.0 0.000 0.000 63.705 63.706 density_matrix_trs4 2 5.0 0.002 0.003 54.764 54.825 multiply_cannon 111 7.7 0.136 0.233 48.423 50.805 multiply_cannon_loop 111 8.7 0.099 0.101 45.323 45.864 ls_scf_init_scf 1 4.0 0.000 0.000 33.538 33.539 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.038 32.065 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.351 29.356 mp_waitall_1 4527 11.1 21.747 25.570 21.747 25.570 make_m2s 222 7.7 0.005 0.005 22.640 23.828 make_images 222 8.7 3.573 3.868 22.533 23.718 multiply_cannon_multrec 444 9.7 17.788 18.389 22.473 23.076 hybrid_alltoall_any 227 10.6 1.652 3.616 12.795 15.636 make_images_data 222 9.7 0.003 0.004 13.023 15.214 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.107 10.357 multiply_cannon_sync_h2d 444 9.7 8.790 8.840 8.790 8.840 arnoldi_extremal 4 6.8 0.000 0.000 7.519 7.526 arnoldi_normal_ev 4 7.8 0.003 0.009 7.519 7.526 build_subspace 16 8.4 0.026 0.036 6.966 6.974 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.583 5.734 dbcsr_matrix_vector_mult_local 304 10.0 5.170 5.467 5.173 5.469 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.067 5.163 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.835 5.102 dbcsr_mm_accdrv_process 1814 10.4 0.266 0.352 4.513 4.647 dbcsr_mm_accdrv_process_sort 1814 11.4 4.177 4.313 4.177 4.313 mp_allgather_i34 111 8.7 1.209 3.874 1.209 3.874 ls_scf_post 1 4.0 0.000 0.000 3.760 3.771 make_images_sizes 222 9.7 0.000 0.000 1.480 3.620 mp_alltoall_i44 222 10.7 1.480 3.620 1.480 3.620 ls_scf_store_result 1 5.0 0.000 0.000 3.481 3.517 calculate_norms 792 9.8 3.229 3.273 3.229 3.273 dbcsr_finalize 304 7.8 0.082 0.089 3.070 3.173 dbcsr_merge_all 275 8.9 0.882 0.920 2.854 2.950 qs_energies_init_hamiltonians 1 3.0 0.003 0.006 2.904 2.904 dbcsr_complete_redistribute 5 7.6 1.427 1.465 2.733 2.851 matrix_ls_to_qs 2 6.0 0.000 0.000 2.383 2.510 dbcsr_sort_data 325 11.1 2.438 2.499 2.438 2.499 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.384 2.386 dbcsr_data_new 6591 9.6 1.882 2.382 1.882 2.382 rebuild_ks_matrix 3 7.3 0.000 0.000 2.318 2.320 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.318 2.320 dbcsr_new_transposed 4 7.5 0.241 0.252 2.294 2.305 dbcsr_frobenius_norm 74 6.6 2.056 2.129 2.192 2.226 dbcsr_add_d 103 6.2 0.000 0.000 2.124 2.206 dbcsr_add_anytype 103 7.2 0.858 0.890 2.124 2.206 dbcsr_data_release 12724 10.6 1.981 2.180 1.981 2.180 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.273000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7037.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/dfc565023ea3187a044d14b931fefe7b58c72354_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 588.836864E+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 4806644. 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 0.029 0.144 289.687 289.688 qs_mol_dyn_low 1 2.0 0.003 0.026 288.423 288.483 qs_forces 5 3.8 0.005 0.024 288.306 288.313 qs_energies 5 4.8 0.005 0.079 284.937 284.973 scf_env_do_scf 5 5.8 0.000 0.010 268.513 268.515 scf_env_do_scf_inner_loop 105 6.6 0.002 0.012 233.770 233.773 qs_scf_new_mos 105 7.6 0.000 0.001 183.624 183.816 qs_scf_loop_do_ot 105 8.6 0.001 0.001 183.623 183.815 ot_scf_mini 105 9.6 0.003 0.005 173.349 173.548 dbcsr_multiply_generic 1445 12.2 0.127 0.139 140.468 141.222 velocity_verlet 4 3.0 0.001 0.007 121.189 121.190 multiply_cannon 1445 13.2 0.278 0.288 118.163 120.812 multiply_cannon_loop 1445 14.2 2.760 2.977 115.689 117.893 qs_ot_get_p 112 10.4 0.001 0.001 78.348 78.709 qs_ot_p2m_diag 40 11.0 0.020 0.031 65.411 65.500 ot_mini 105 10.6 0.001 0.001 63.970 64.126 cp_dbcsr_syevd 40 12.0 0.002 0.002 61.836 61.839 cp_fm_syevd 40 13.0 0.000 0.000 55.533 55.695 cp_fm_redistribute_end 40 14.0 25.057 49.999 25.067 50.004 cp_fm_syevd_base 40 14.0 24.915 49.857 24.915 49.857 mp_waitall_1 488190 16.1 40.251 47.712 40.251 47.712 multiply_cannon_multrec 69360 15.2 29.745 34.587 39.271 44.146 qs_ot_get_derivative 55 11.6 0.001 0.001 42.009 42.225 multiply_cannon_metrocomm3 69360 15.2 0.208 0.219 27.485 36.877 init_scf_loop 7 6.6 0.000 0.007 34.685 34.686 multiply_cannon_sync_h2d 69360 15.2 29.101 33.606 29.101 33.606 rebuild_ks_matrix 110 8.4 0.000 0.000 33.081 33.279 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.020 33.080 33.279 qs_ks_update_qs_env 112 7.6 0.001 0.001 30.433 30.616 prepare_preconditioner 7 7.6 0.000 0.004 29.132 29.168 make_preconditioner 7 8.6 0.000 0.003 29.132 29.168 qs_rho_update_rho_low 110 7.6 0.001 0.001 23.667 24.065 calculate_rho_elec 110 8.6 0.029 0.032 23.667 24.064 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.197 23.478 apply_single 62 13.6 0.000 0.000 23.196 23.478 ot_new_cg_direction 55 11.6 0.001 0.001 21.200 21.202 make_full_inverse_cholesky 7 9.6 0.000 0.000 20.530 20.596 density_rs2pw 110 9.6 0.005 0.007 17.288 17.647 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.573 15.757 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 15.227 15.372 mp_sum_l 4764 12.2 13.211 14.262 13.211 14.262 transfer_rs2pw 445 10.6 0.007 0.008 13.604 14.048 init_scf_run 5 5.8 0.000 0.002 13.539 13.540 scf_env_initial_rho_setup 5 6.8 0.000 0.001 13.539 13.540 cp_fm_cholesky_invert 7 10.6 13.406 13.428 13.406 13.428 pw_transfer 1645 12.4 0.080 0.102 12.936 13.163 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 12.797 13.021 calculate_dm_sparse 110 9.5 0.000 0.001 11.892 12.095 sum_up_and_integrate 60 10.3 0.001 0.002 11.736 11.747 integrate_v_rspace 60 11.3 0.001 0.002 11.719 11.730 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 11.240 11.319 fft_wrap_pw1pw2_240 915 15.0 1.158 1.248 11.071 11.304 qs_vxc_create 110 10.4 0.002 0.003 11.112 11.169 check_diag 80 13.5 8.598 8.923 10.948 11.152 dbcsr_mm_accdrv_process 154766 15.8 5.878 6.357 9.395 10.476 fft3d_pb 915 16.0 2.395 2.698 9.241 9.557 make_m2s 2890 13.2 0.078 0.084 8.683 9.196 acc_transpose_blocks 69360 15.2 0.337 0.367 8.585 9.131 make_images 2890 14.2 0.242 0.261 8.577 9.090 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.004 8.712 8.734 multiply_cannon_metrocomm1 69360 15.2 0.095 0.102 5.325 8.555 make_full_single_inverse 7 9.6 0.001 0.011 8.252 8.289 calculate_first_density_matrix 1 7.0 0.000 0.004 8.226 8.237 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.001 8.109 8.171 xc_rho_set_and_dset_create 110 12.4 0.078 0.102 7.669 7.940 transfer_rs2pw_30 110 11.6 1.252 1.313 7.217 7.800 potential_pw2rs 60 12.3 0.003 0.003 7.608 7.656 mp_sendrecv_dv 168740 12.6 7.493 7.650 7.493 7.650 xc_vxc_pw_create 60 11.3 0.039 0.054 7.547 7.604 xc_pw_derive 510 13.4 0.005 0.007 6.778 6.842 mp_alltoall_z22v 2340 17.7 6.307 6.714 6.307 6.714 cp_fm_cholesky_decompose 14 10.2 6.514 6.523 6.514 6.523 transfer_pw2rs 245 13.2 0.003 0.004 6.354 6.395 acc_transpose_blocks_kernels 69360 16.2 0.814 0.888 5.975 6.278 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=289.688000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=560.600000, yerr=1.959592 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: dfc565023ea3187a044d14b931fefe7b58c72354 Summary: empty Status: OK