=== 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: 6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a ################# 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/01 job id: 49040703 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/02 job id: 49040704 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/03 job id: 49040705 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/04 job id: 49040706 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/05 job id: 49040707 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/06 job id: 49040708 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/07 job id: 49040709 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/08 job id: 49040710 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/09 job id: 49040711 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/10 job id: 49040713 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/11 job id: 49040714 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/12 job id: 49040715 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/13 job id: 49040716 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/14 job id: 49040717 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/15 job id: 49040720 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/16 job id: 49040723 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/17 job id: 49040724 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/18 job id: 49040725 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/19 job id: 49040726 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/20 job id: 49040727 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/21 job id: 49040728 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/22 job id: 49040729 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/23 job id: 49040730 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/24 job id: 49040731 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/25 job id: 49040732 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/26 job id: 49040733 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/27 job id: 49040735 --- 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 0.018 0.036 133.271 133.272 farming_run 1 2.0 132.732 132.733 133.230 133.233 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.503896E+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 0.005 0.023 114.732 114.732 qs_energies 1 2.0 0.000 0.000 114.526 114.527 mp2_main 1 3.0 0.000 0.000 112.321 112.323 mp2_gpw_main 1 4.0 0.027 0.033 111.229 111.231 mp2_ri_gpw_compute_in 1 5.0 0.171 0.174 92.476 92.750 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.720 54.994 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.169 41.103 46.079 get_2c_integrals 1 6.0 0.008 0.009 36.948 37.582 integrate_v_rspace 273 8.0 0.441 0.454 24.696 29.554 pw_transfer 6555 10.6 0.374 0.387 26.828 27.238 grid_integrate_task_list 273 9.0 20.597 25.945 20.597 25.945 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 25.484 25.896 fft_wrap_pw1pw2_100 2178 12.4 1.118 1.278 23.047 23.441 compute_2c_integrals 1 7.0 0.003 0.003 18.972 18.973 rpa_ri_compute_en 1 5.0 0.027 0.030 18.635 18.960 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.555 18.689 mp2_eri_2c_integrate_gpw 1 9.0 2.384 2.416 18.552 18.687 cp_fm_cholesky_decompose 12 8.2 17.915 18.530 17.915 18.530 cholesky_decomp 1 7.0 0.000 0.000 16.806 17.426 fft3d_s 5443 13.4 16.124 16.435 16.146 16.458 ao_to_mo_and_store_B_mult_1 272 7.0 10.778 15.330 10.778 15.330 calculate_wavefunction 272 8.0 5.380 5.540 12.317 12.994 rpa_num_int 1 6.0 0.000 0.002 10.621 10.621 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.606 10.618 calc_mat_Q 8 8.0 0.000 0.000 9.453 9.538 contract_S_to_Q 8 9.0 0.000 0.000 8.873 8.965 calc_potential_gpw 544 9.5 0.005 0.005 8.224 8.650 parallel_gemm_fm 14 9.1 0.000 0.000 8.467 8.570 parallel_gemm_fm_cosma 14 10.1 8.467 8.570 8.467 8.570 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.052 8.269 potential_pw2rs 545 10.0 0.106 0.109 7.471 8.096 create_integ_mat 1 6.0 0.022 0.027 7.638 7.638 collocate_single_gaussian 272 10.0 0.040 0.042 7.311 7.570 array2fm 1 7.0 0.000 0.000 6.675 7.113 pw_scatter_s 2720 13.7 4.364 4.450 4.364 4.450 pw_gather_s 2722 13.2 3.484 3.788 3.484 3.788 array2fm_buffer_send 1 8.0 2.971 3.147 2.971 3.147 pw_poisson_solve 545 10.5 1.115 1.191 2.166 2.358 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.231051, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2821.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 0.038 0.061 420.504 420.507 farming_run 1 2.0 419.756 419.761 420.445 420.470 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.239327E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.032 217.308 217.310 qs_energies 1 2.0 0.000 0.000 217.092 217.099 scf_env_do_scf 1 3.0 0.000 0.000 115.276 115.277 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.376 114.383 rebuild_ks_matrix 4 6.0 0.000 0.000 114.375 114.381 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 114.375 114.381 hfx_ks_matrix 4 8.0 0.001 0.001 114.030 114.035 integrate_four_center 4 9.0 0.152 0.462 114.029 114.034 integrate_four_center_main 4 10.0 0.134 0.615 102.261 105.458 integrate_four_center_bin 264 11.0 102.127 104.882 102.127 104.882 mp2_main 1 3.0 0.000 0.000 101.517 101.524 mp2_gpw_main 1 4.0 0.048 0.067 100.678 100.690 init_scf_loop 1 4.0 0.000 0.000 97.034 97.035 mp2_ri_gpw_compute_in 1 5.0 0.063 0.064 74.091 75.158 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.769 54.834 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.160 41.359 46.516 integrate_v_rspace 95 8.0 0.398 0.575 27.826 32.816 pw_transfer 2240 10.6 0.143 0.182 29.529 30.120 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.530 29.105 grid_integrate_task_list 95 9.0 23.250 28.459 23.250 28.459 mp2_ri_gpw_compute_en 1 5.0 0.053 0.063 26.426 28.439 ao_to_mo_and_store_B_mult_1 91 7.0 10.735 27.920 10.735 27.920 fft_wrap_pw1pw2_100 730 12.4 1.310 1.560 26.290 26.889 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.833 1.892 24.712 24.723 get_2c_integrals 1 6.0 0.000 0.000 20.234 20.258 compute_2c_integrals 1 7.0 0.002 0.003 19.216 19.218 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.691 19.096 mp2_eri_2c_integrate_gpw 1 9.0 1.720 1.831 18.690 19.095 fft3d_s 1823 13.4 18.507 19.032 18.521 19.047 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 18.240 18.241 calculate_wavefunction 91 8.0 2.020 2.049 9.623 9.861 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.590 8.696 9.249 potential_pw2rs 186 10.0 0.033 0.035 8.461 9.002 local_gemm 172 8.0 8.139 8.664 8.139 8.664 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.212 8.594 mp2_ri_gpw_compute_en_comm 22 7.0 0.501 0.523 7.796 8.403 calc_potential_gpw 182 9.5 0.002 0.003 7.915 8.224 collocate_single_gaussian 91 10.0 0.017 0.023 7.778 8.201 mp_sync 37 10.5 3.748 7.164 3.748 7.164 integrate_four_center_load 4 10.0 0.000 0.000 6.757 6.761 hfx_load_balance 1 11.0 0.000 0.000 6.757 6.761 mp2_ri_gpw_compute_en_ener 172 7.0 6.342 6.427 6.342 6.427 mp_sendrecv_dm3 2068 8.0 5.826 6.414 5.826 6.414 pw_gather_s 912 13.2 4.464 4.999 4.464 4.999 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.687161, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1499.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 452.460544E+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 0.023 0.032 49.828 49.829 qs_mol_dyn_low 1 2.0 0.003 0.003 49.560 49.568 qs_forces 11 3.9 0.002 0.003 49.489 49.498 qs_energies 11 4.9 0.003 0.010 48.016 48.032 scf_env_do_scf 11 5.9 0.001 0.001 42.115 42.115 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 40.075 40.076 dbcsr_multiply_generic 2286 12.5 0.093 0.099 30.127 30.656 qs_scf_new_mos 108 7.5 0.000 0.001 29.948 30.218 qs_scf_loop_do_ot 108 8.5 0.000 0.001 29.948 30.217 ot_scf_mini 108 9.5 0.002 0.002 28.454 28.631 multiply_cannon 2286 13.5 0.195 0.203 24.845 26.507 multiply_cannon_loop 2286 14.5 1.771 1.868 24.138 25.797 velocity_verlet 10 3.0 0.001 0.002 23.601 23.601 ot_mini 108 10.5 0.001 0.001 16.817 17.082 qs_ot_get_derivative 108 11.5 0.001 0.001 14.052 14.240 mp_waitall_1 245248 16.5 6.675 13.167 6.675 13.167 multiply_cannon_metrocomm3 54864 15.5 0.075 0.081 4.860 11.855 multiply_cannon_multrec 54864 15.5 3.955 6.164 7.948 11.562 rebuild_ks_matrix 119 8.3 0.000 0.000 7.965 8.136 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.965 8.136 qs_ot_get_p 119 10.4 0.001 0.001 7.484 7.755 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.024 7.174 multiply_cannon_sync_h2d 54864 15.5 5.440 6.031 5.440 6.031 dbcsr_mm_accdrv_process 76910 16.1 1.702 2.693 3.906 5.677 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.885 5.339 mp_sum_l 7287 12.8 3.220 5.151 3.220 5.151 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.003 5.034 sum_up_and_integrate 119 10.3 0.001 0.002 4.605 4.611 integrate_v_rspace 119 11.3 0.002 0.003 4.594 4.602 init_scf_run 11 5.9 0.000 0.001 4.597 4.597 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.597 4.597 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.471 4.575 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.340 4.341 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.179 4.260 calculate_rho_elec 119 8.7 0.011 0.017 4.179 4.259 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.109 4.109 cp_fm_redistribute_end 50 14.0 2.097 4.077 2.103 4.080 cp_fm_diag_elpa_base 50 14.0 1.971 3.965 1.975 3.972 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.863 3.063 apply_single 119 13.6 0.000 0.000 2.863 3.063 jit_kernel_multiply 13 15.8 2.144 2.894 2.144 2.894 calculate_dm_sparse 119 9.5 0.000 0.000 2.715 2.823 acc_transpose_blocks 54864 15.5 0.215 0.240 2.114 2.637 ot_diis_step 108 11.5 0.006 0.006 2.595 2.596 calculate_first_density_matrix 1 7.0 0.000 0.000 2.445 2.446 density_rs2pw 119 9.7 0.004 0.004 2.310 2.409 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.197 2.199 grid_integrate_task_list 119 12.3 2.008 2.111 2.008 2.111 wfi_extrapolate 11 7.9 0.001 0.001 2.084 2.084 init_scf_loop 11 6.9 0.000 0.000 2.022 2.022 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.944 1.993 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.910 1.973 potential_pw2rs 119 12.3 0.004 0.004 1.949 1.968 multiply_cannon_metrocomm1 54864 15.5 0.057 0.062 1.048 1.867 pw_transfer 1439 11.6 0.051 0.057 1.775 1.856 make_m2s 4572 13.5 0.054 0.056 1.751 1.801 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.699 1.782 mp_sum_d 4135 12.0 1.065 1.732 1.065 1.732 make_images 4572 14.5 0.133 0.138 1.668 1.718 transfer_rs2pw 487 10.6 0.005 0.006 1.461 1.613 mp_alltoall_d11v 2130 13.8 1.279 1.521 1.279 1.521 fft3d_ps 1201 14.6 0.371 0.480 1.415 1.486 mp_waitany 12084 13.8 1.253 1.449 1.253 1.449 transfer_pw2rs 487 13.2 0.006 0.006 1.422 1.440 grid_collocate_task_list 119 9.7 1.346 1.430 1.346 1.430 fft_wrap_pw1pw2_140 487 13.2 0.140 0.155 1.318 1.404 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.384 1.397 acc_transpose_blocks_sync 164592 16.5 1.129 1.366 1.129 1.366 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.988 1.008 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=49.829000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 488.890368E+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 1395232. 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 0.010 0.026 49.164 49.165 qs_mol_dyn_low 1 2.0 0.003 0.003 48.778 48.813 qs_forces 11 3.9 0.004 0.022 48.700 48.702 qs_energies 11 4.9 0.003 0.012 46.793 46.804 scf_env_do_scf 11 5.9 0.001 0.001 41.013 41.013 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 37.432 37.433 dbcsr_multiply_generic 2286 12.5 0.104 0.108 25.803 26.750 qs_scf_new_mos 108 7.5 0.001 0.001 25.926 26.171 qs_scf_loop_do_ot 108 8.5 0.001 0.001 25.925 26.171 ot_scf_mini 108 9.5 0.002 0.003 24.832 25.001 velocity_verlet 10 3.0 0.001 0.002 23.349 23.350 multiply_cannon 2286 13.5 0.213 0.220 18.491 21.014 multiply_cannon_loop 2286 14.5 1.199 1.271 16.845 19.616 ot_mini 108 10.5 0.001 0.001 15.309 15.556 mp_waitall_1 200699 16.5 8.278 13.657 8.278 13.657 qs_ot_get_derivative 108 11.5 0.001 0.001 12.677 12.844 multiply_cannon_metrocomm3 27432 15.5 0.070 0.073 5.077 10.851 multiply_cannon_multrec 27432 15.5 1.800 4.143 6.263 9.412 rebuild_ks_matrix 119 8.3 0.000 0.000 8.492 8.782 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 8.492 8.782 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.511 7.772 dbcsr_mm_accdrv_process 47894 16.0 3.526 5.918 4.383 7.235 qs_ot_get_p 119 10.4 0.001 0.001 6.229 6.470 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.072 6.146 mp_sum_l 7287 12.8 3.137 5.513 3.137 5.513 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.188 5.265 calculate_rho_elec 119 8.7 0.021 0.024 5.188 5.265 sum_up_and_integrate 119 10.3 0.001 0.002 5.259 5.264 integrate_v_rspace 119 11.3 0.003 0.005 5.245 5.251 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.308 4.904 apply_single 119 13.6 0.000 0.000 3.308 4.904 init_scf_run 11 5.9 0.000 0.001 4.392 4.392 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.391 4.392 make_m2s 4572 13.5 0.053 0.054 3.734 4.132 make_images 4572 14.5 0.205 0.244 3.644 4.039 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.954 3.985 init_scf_loop 11 6.9 0.000 0.000 3.536 3.538 density_rs2pw 119 9.7 0.004 0.004 3.422 3.531 cp_dbcsr_syevd 50 12.0 0.003 0.003 3.493 3.494 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.994 3.082 potential_pw2rs 119 12.3 0.006 0.006 2.934 2.974 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.929 2.930 cp_fm_redistribute_end 50 14.0 1.479 2.876 1.490 2.880 jit_kernel_multiply 11 16.5 0.799 2.877 0.799 2.877 cp_fm_diag_elpa_base 50 14.0 1.345 2.729 1.377 2.776 ot_diis_step 108 11.5 0.011 0.011 2.502 2.504 transfer_rs2pw 487 10.6 0.005 0.005 2.333 2.476 prepare_preconditioner 11 7.9 0.000 0.000 2.440 2.467 make_preconditioner 11 8.9 0.000 0.000 2.440 2.467 pw_transfer 1439 11.6 0.064 0.069 2.354 2.405 make_images_data 4572 15.5 0.047 0.054 1.623 2.386 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.319 2.383 calculate_first_density_matrix 1 7.0 0.000 0.001 2.357 2.359 calculate_dm_sparse 119 9.5 0.001 0.001 2.269 2.355 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.262 2.314 transfer_pw2rs 487 13.2 0.004 0.005 2.191 2.227 multiply_cannon_sync_h2d 27432 15.5 1.614 2.224 1.614 2.224 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.085 2.088 hybrid_alltoall_any 4725 16.4 0.054 0.118 1.402 2.052 acc_transpose_blocks 27432 15.5 0.112 0.118 1.622 1.955 fft3d_ps 1201 14.6 0.523 0.581 1.900 1.951 grid_integrate_task_list 119 12.3 1.821 1.951 1.821 1.951 wfi_extrapolate 11 7.9 0.001 0.001 1.947 1.947 mp_alltoall_d11v 2130 13.8 1.649 1.870 1.649 1.870 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.711 1.792 make_images_sizes 4572 15.5 0.005 0.005 1.241 1.674 mp_alltoall_i44 4572 16.5 1.237 1.669 1.237 1.669 mp_allgather_i34 2286 14.5 1.062 1.659 1.062 1.659 fft_wrap_pw1pw2_140 487 13.2 0.158 0.169 1.557 1.611 mp_sendrecv_dv 16779 12.7 1.499 1.539 1.499 1.539 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.461 1.503 grid_collocate_task_list 119 9.7 1.288 1.349 1.288 1.349 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.323 1.338 mp_sum_d 4135 12.0 0.809 1.286 0.809 1.286 cp_fm_cholesky_invert 11 10.9 1.199 1.206 1.199 1.206 transfer_rs2pw_50 119 11.7 0.184 0.193 1.016 1.137 mp_alltoall_z22v 1201 16.6 1.006 1.133 1.006 1.133 transfer_pw2rs_50 119 14.3 0.175 0.189 0.974 1.091 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.043 1.071 dbcsr_complete_redistribute 329 12.2 0.081 0.124 0.949 1.035 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=49.165000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.545455, yerr=2.270908 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 517.648384E+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 0.027 0.044 38.311 38.312 qs_mol_dyn_low 1 2.0 0.003 0.003 38.006 38.014 qs_forces 11 3.9 0.002 0.003 37.933 37.941 qs_energies 11 4.9 0.001 0.001 36.267 36.275 scf_env_do_scf 11 5.9 0.000 0.001 31.076 31.076 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 27.854 27.855 dbcsr_multiply_generic 2286 12.5 0.095 0.098 19.031 19.093 velocity_verlet 10 3.0 0.001 0.002 18.066 18.067 qs_scf_new_mos 108 7.5 0.001 0.001 17.976 17.988 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.975 17.987 ot_scf_mini 108 9.5 0.002 0.002 17.177 17.192 multiply_cannon 2286 13.5 0.196 0.201 14.547 15.351 multiply_cannon_loop 2286 14.5 0.858 0.892 13.618 14.274 ot_mini 108 10.5 0.001 0.001 10.634 10.652 qs_ot_get_derivative 108 11.5 0.001 0.001 9.100 9.116 multiply_cannon_multrec 18288 15.5 1.883 2.894 7.169 7.678 rebuild_ks_matrix 119 8.3 0.000 0.000 7.272 7.290 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.271 7.290 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.458 6.475 dbcsr_mm_accdrv_process 38222 16.0 5.117 6.060 5.192 6.125 mp_waitall_1 158411 16.6 4.382 5.573 4.382 5.573 sum_up_and_integrate 119 10.3 0.001 0.001 4.730 4.735 integrate_v_rspace 119 11.3 0.003 0.003 4.717 4.724 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.476 4.479 calculate_rho_elec 119 8.7 0.031 0.031 4.476 4.478 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.588 4.470 qs_ot_get_p 119 10.4 0.001 0.001 4.139 4.162 init_scf_run 11 5.9 0.000 0.001 3.920 3.920 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.920 3.920 init_scf_loop 11 6.9 0.000 0.000 3.196 3.197 density_rs2pw 119 9.7 0.004 0.004 3.006 3.120 make_m2s 4572 13.5 0.044 0.045 2.664 2.850 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.334 2.792 apply_single 119 13.6 0.000 0.000 2.334 2.792 make_images 4572 14.5 0.192 0.204 2.579 2.764 multiply_cannon_metrocomm3 18288 15.5 0.046 0.048 1.758 2.755 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.737 2.743 potential_pw2rs 119 12.3 0.007 0.009 2.459 2.486 calculate_first_density_matrix 1 7.0 0.000 0.000 2.402 2.402 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.401 2.402 prepare_preconditioner 11 7.9 0.000 0.000 2.313 2.317 make_preconditioner 11 8.9 0.000 0.000 2.313 2.316 pw_transfer 1439 11.6 0.065 0.069 2.253 2.287 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.139 2.239 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.160 2.194 transfer_rs2pw 487 10.6 0.005 0.005 1.950 2.075 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.063 2.074 cp_fm_diag_elpa_base 50 14.0 2.032 2.051 2.060 2.072 calculate_dm_sparse 119 9.5 0.000 0.001 2.003 2.010 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.003 2.010 grid_integrate_task_list 119 12.3 1.791 1.929 1.791 1.929 mp_sum_l 7287 12.8 1.432 1.898 1.432 1.898 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.828 1.830 transfer_pw2rs 487 13.2 0.004 0.004 1.767 1.793 acc_transpose_blocks 18288 15.5 0.079 0.080 1.740 1.783 fft3d_ps 1201 14.6 0.531 0.552 1.732 1.771 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.629 1.636 fft_wrap_pw1pw2_140 487 13.2 0.209 0.217 1.565 1.601 ot_diis_step 108 11.5 0.011 0.011 1.515 1.515 make_images_data 4572 15.5 0.047 0.052 1.180 1.507 wfi_extrapolate 11 7.9 0.001 0.001 1.456 1.456 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.314 1.317 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.980 1.304 grid_collocate_task_list 119 9.7 1.236 1.292 1.236 1.292 multiply_cannon_sync_h2d 18288 15.5 1.035 1.216 1.035 1.216 mp_sendrecv_dv 11067 12.7 1.184 1.216 1.184 1.216 mp_alltoall_d11v 2130 13.8 0.920 1.196 0.920 1.196 mp_alltoall_z22v 1201 16.6 1.007 1.076 1.007 1.076 make_images_sizes 4572 15.5 0.005 0.005 0.730 1.075 mp_alltoall_i44 4572 16.5 0.726 1.071 0.726 1.071 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.537 1.054 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.033 1.048 cp_fm_cholesky_invert 11 10.9 1.021 1.026 1.021 1.026 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.955 0.962 acc_transpose_blocks_kernels 18288 16.5 0.219 0.227 0.902 0.913 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.885 0.887 transfer_rs2pw_50 119 11.7 0.155 0.159 0.782 0.857 transfer_pw2rs_50 119 14.3 0.145 0.152 0.737 0.822 mp_allgather_i34 2286 14.5 0.378 0.780 0.378 0.780 acc_transpose_blocks_sync 54864 16.5 0.740 0.780 0.740 0.780 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=38.312000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=494.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 561.336320E+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 0.010 0.029 46.482 46.482 qs_mol_dyn_low 1 2.0 0.003 0.003 46.285 46.293 qs_forces 11 3.9 0.003 0.009 46.223 46.226 qs_energies 11 4.9 0.001 0.002 44.345 44.358 scf_env_do_scf 11 5.9 0.001 0.001 38.593 38.595 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 33.823 33.823 dbcsr_multiply_generic 2286 12.5 0.100 0.103 24.027 24.247 qs_scf_new_mos 108 7.5 0.001 0.001 23.530 23.595 qs_scf_loop_do_ot 108 8.5 0.001 0.001 23.529 23.594 velocity_verlet 10 3.0 0.001 0.002 22.857 22.859 ot_scf_mini 108 9.5 0.002 0.002 22.359 22.403 multiply_cannon 2286 13.5 0.220 0.229 17.946 19.373 multiply_cannon_loop 2286 14.5 1.504 1.605 16.555 17.913 ot_mini 108 10.5 0.001 0.001 13.756 13.826 qs_ot_get_derivative 108 11.5 0.001 0.001 11.742 11.789 multiply_cannon_multrec 27432 15.5 2.462 3.135 9.597 10.306 dbcsr_mm_accdrv_process 47916 15.9 6.563 8.986 7.032 9.048 rebuild_ks_matrix 119 8.3 0.000 0.000 7.638 7.686 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 7.637 7.686 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.787 6.832 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.760 5.739 qs_ot_get_p 119 10.4 0.001 0.001 5.411 5.494 mp_waitall_1 137007 16.6 4.184 5.461 4.184 5.461 init_scf_loop 11 6.9 0.000 0.000 4.735 4.736 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.614 4.664 calculate_rho_elec 119 8.7 0.040 0.046 4.613 4.663 sum_up_and_integrate 119 10.3 0.001 0.001 4.638 4.647 integrate_v_rspace 119 11.3 0.003 0.003 4.626 4.635 init_scf_run 11 5.9 0.000 0.001 4.258 4.259 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.258 4.258 make_m2s 4572 13.5 0.054 0.056 3.632 3.890 make_images 4572 14.5 0.272 0.333 3.523 3.776 prepare_preconditioner 11 7.9 0.000 0.000 3.736 3.744 make_preconditioner 11 8.9 0.000 0.000 3.736 3.744 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.300 3.652 qs_ot_p2m_diag 50 11.0 0.015 0.023 3.459 3.473 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.793 3.194 apply_single 119 13.6 0.000 0.000 2.793 3.194 density_rs2pw 119 9.7 0.004 0.004 3.027 3.140 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.979 2.980 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.882 2.907 mp_sum_l 7287 12.8 2.029 2.680 2.029 2.680 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 1.396 2.547 acc_transpose_blocks 27432 15.5 0.116 0.120 2.423 2.543 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.525 2.533 cp_fm_diag_elpa_base 50 14.0 2.480 2.501 2.522 2.530 pw_transfer 1439 11.6 0.065 0.069 2.409 2.445 calculate_dm_sparse 119 9.5 0.000 0.000 2.351 2.429 calculate_first_density_matrix 1 7.0 0.000 0.000 2.415 2.418 potential_pw2rs 119 12.3 0.009 0.009 2.381 2.389 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.317 2.356 jit_kernel_multiply 6 16.6 0.404 2.354 0.404 2.354 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.044 2.047 transfer_rs2pw 487 10.6 0.004 0.005 1.910 2.032 ot_diis_step 108 11.5 0.012 0.012 1.943 1.945 make_images_data 4572 15.5 0.047 0.051 1.566 1.936 grid_integrate_task_list 119 12.3 1.815 1.890 1.815 1.890 fft3d_ps 1201 14.6 0.555 0.606 1.838 1.864 wfi_extrapolate 11 7.9 0.001 0.001 1.773 1.773 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.755 1.769 fft_wrap_pw1pw2_140 487 13.2 0.244 0.257 1.726 1.766 transfer_pw2rs 487 13.2 0.004 0.004 1.652 1.664 hybrid_alltoall_any 4725 16.4 0.065 0.152 1.268 1.551 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.504 1.548 acc_transpose_blocks_sync 82296 16.5 1.420 1.538 1.420 1.538 cp_fm_cholesky_invert 11 10.9 1.443 1.450 1.443 1.450 dbcsr_complete_redistribute 329 12.2 0.127 0.155 1.099 1.385 multiply_cannon_metrocomm1 27432 15.5 0.034 0.035 0.373 1.380 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.333 1.349 grid_collocate_task_list 119 9.7 1.250 1.340 1.250 1.340 make_images_sizes 4572 15.5 0.005 0.005 0.934 1.310 mp_alltoall_i44 4572 16.5 0.929 1.305 0.929 1.305 cp_fm_upper_to_full 72 14.2 0.871 1.253 0.871 1.253 mp_alltoall_d11v 2130 13.8 1.086 1.212 1.086 1.212 mp_alltoall_z22v 1201 16.6 1.167 1.197 1.167 1.197 mp_sendrecv_dv 8211 12.7 1.158 1.175 1.158 1.175 qs_energies_init_hamiltonians 11 5.9 0.009 0.013 1.137 1.142 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.746 1.026 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.946 0.950 mp_allgather_i34 2286 14.5 0.665 0.937 0.665 0.937 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=46.482000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=532.181818, yerr=3.833303 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 621.748224E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.030 34.312 34.313 qs_mol_dyn_low 1 2.0 0.003 0.003 34.059 34.068 qs_forces 11 3.9 0.002 0.002 33.984 33.986 qs_energies 11 4.9 0.001 0.002 32.167 32.172 scf_env_do_scf 11 5.9 0.000 0.001 26.982 26.982 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 23.852 23.853 velocity_verlet 10 3.0 0.002 0.002 17.312 17.315 dbcsr_multiply_generic 2286 12.5 0.093 0.097 15.165 15.326 qs_scf_new_mos 108 7.5 0.001 0.001 14.497 14.538 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.496 14.537 ot_scf_mini 108 9.5 0.002 0.002 13.682 13.718 multiply_cannon 2286 13.5 0.226 0.232 11.529 12.283 multiply_cannon_loop 2286 14.5 0.646 0.665 10.461 10.870 ot_mini 108 10.5 0.001 0.001 7.867 7.903 rebuild_ks_matrix 119 8.3 0.000 0.000 6.678 6.732 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.678 6.732 multiply_cannon_multrec 9144 15.5 1.721 1.890 6.445 6.631 qs_ot_get_derivative 108 11.5 0.001 0.001 6.478 6.513 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.963 6.012 dbcsr_mm_accdrv_process 12550 15.8 3.816 4.631 4.612 4.678 sum_up_and_integrate 119 10.3 0.001 0.001 4.226 4.231 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.212 4.228 calculate_rho_elec 119 8.7 0.060 0.061 4.211 4.227 integrate_v_rspace 119 11.3 0.003 0.003 4.216 4.221 qs_ot_get_p 119 10.4 0.001 0.001 3.724 3.771 init_scf_run 11 5.9 0.000 0.001 3.611 3.612 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.611 3.612 mp_waitall_1 115863 16.7 2.655 3.217 2.655 3.217 init_scf_loop 11 6.9 0.000 0.000 3.104 3.106 make_m2s 4572 13.5 0.034 0.035 2.528 2.716 make_images 4572 14.5 0.270 0.301 2.438 2.626 density_rs2pw 119 9.7 0.003 0.004 2.502 2.585 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.489 2.512 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.397 2.401 pw_transfer 1439 11.6 0.065 0.067 2.316 2.339 prepare_preconditioner 11 7.9 0.000 0.000 2.281 2.286 make_preconditioner 11 8.9 0.000 0.000 2.281 2.286 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.223 2.249 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.137 2.186 calculate_first_density_matrix 1 7.0 0.000 0.000 2.184 2.184 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.098 2.100 grid_integrate_task_list 119 12.3 1.935 1.996 1.935 1.996 calculate_dm_sparse 119 9.5 0.000 0.000 1.951 1.972 potential_pw2rs 119 12.3 0.010 0.011 1.910 1.923 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.778 1.795 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.754 1.761 cp_fm_diag_elpa_base 50 14.0 1.723 1.741 1.752 1.759 fft_wrap_pw1pw2_140 487 13.2 0.323 0.332 1.710 1.739 fft3d_ps 1201 14.6 0.561 0.574 1.638 1.654 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.608 1.610 acc_transpose_blocks 9144 15.5 0.041 0.043 1.499 1.526 transfer_rs2pw 487 10.6 0.004 0.004 1.420 1.504 make_images_data 4572 15.5 0.042 0.045 1.104 1.455 jit_kernel_multiply 6 16.1 0.757 1.446 0.757 1.446 grid_collocate_task_list 119 9.7 1.344 1.415 1.344 1.415 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.399 1.413 wfi_extrapolate 11 7.9 0.001 0.001 1.365 1.366 ot_diis_step 108 11.5 0.013 0.013 1.358 1.358 hybrid_alltoall_any 4725 16.4 0.065 0.174 0.995 1.350 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.279 1.323 apply_single 119 13.6 0.000 0.000 1.279 1.323 qs_energies_init_hamiltonians 11 5.9 0.013 0.014 1.265 1.268 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.256 1.263 transfer_pw2rs 487 13.2 0.003 0.004 1.203 1.216 cp_fm_cholesky_invert 11 10.9 1.163 1.166 1.163 1.166 mp_alltoall_d11v 2130 13.8 0.995 1.091 0.995 1.091 mp_alltoall_z22v 1201 16.6 0.950 0.989 0.950 0.989 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.733 0.956 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.948 0.955 make_images_sizes 4572 15.5 0.005 0.005 0.640 0.942 mp_alltoall_i44 4572 16.5 0.635 0.937 0.635 0.937 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.874 0.935 mp_allgather_i34 2286 14.5 0.378 0.918 0.378 0.918 mp_sum_l 7287 12.8 0.669 0.891 0.669 0.891 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.839 0.845 mp_sendrecv_dv 5355 12.7 0.771 0.791 0.771 0.791 acc_transpose_blocks_sync 27432 16.5 0.725 0.752 0.725 0.752 qs_env_update_s_mstruct 11 6.9 0.008 0.011 0.702 0.749 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.717 0.721 dbcsr_complete_redistribute 329 12.2 0.165 0.177 0.678 0.715 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=34.313000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=588.181818, yerr=6.925817 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 768.442368E+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 0.034 54.016 54.017 qs_mol_dyn_low 1 2.0 0.003 0.003 53.729 53.737 qs_forces 11 3.9 0.002 0.002 53.618 53.619 qs_energies 11 4.9 0.001 0.001 51.406 51.413 scf_env_do_scf 11 5.9 0.001 0.001 44.974 44.975 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 35.033 35.035 velocity_verlet 10 3.0 0.002 0.002 29.427 29.437 dbcsr_multiply_generic 2286 12.5 0.102 0.104 24.208 24.626 qs_scf_new_mos 108 7.5 0.001 0.001 23.488 23.594 qs_scf_loop_do_ot 108 8.5 0.001 0.001 23.488 23.594 ot_scf_mini 108 9.5 0.002 0.002 22.154 22.256 multiply_cannon 2286 13.5 0.302 0.314 17.768 19.597 multiply_cannon_loop 2286 14.5 0.862 0.876 16.176 18.358 ot_mini 108 10.5 0.001 0.001 13.865 14.033 qs_ot_get_derivative 108 11.5 0.001 0.001 11.565 11.665 multiply_cannon_multrec 9144 15.5 3.323 4.757 9.502 10.177 init_scf_loop 11 6.9 0.000 0.000 9.898 9.903 prepare_preconditioner 11 7.9 0.000 0.000 8.784 8.803 make_preconditioner 11 8.9 0.000 0.000 8.784 8.803 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.222 8.656 rebuild_ks_matrix 119 8.3 0.000 0.000 8.438 8.647 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.014 8.438 8.646 mp_waitall_1 94719 16.7 5.437 8.263 5.437 8.263 dbcsr_mm_accdrv_process 12550 15.8 5.043 7.137 6.040 7.825 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.594 7.783 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.911 5.894 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 2.988 5.242 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.164 5.190 calculate_rho_elec 119 8.7 0.118 0.121 5.163 5.189 qs_ot_get_p 119 10.4 0.001 0.001 4.816 4.997 sum_up_and_integrate 119 10.3 0.001 0.001 4.791 4.798 integrate_v_rspace 119 11.3 0.003 0.003 4.781 4.788 cp_fm_upper_to_full 72 14.2 3.233 4.722 3.233 4.722 mp_sum_l 7287 12.8 2.276 4.331 2.276 4.331 init_scf_run 11 5.9 0.000 0.001 4.274 4.274 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.273 4.274 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.907 4.011 apply_single 119 13.6 0.000 0.000 2.907 4.011 make_m2s 4572 13.5 0.038 0.038 3.629 3.994 make_images 4572 14.5 0.351 0.382 3.509 3.875 dbcsr_complete_redistribute 329 12.2 0.278 0.284 2.304 3.193 pw_transfer 1439 11.6 0.068 0.069 3.160 3.178 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.062 3.080 density_rs2pw 119 9.7 0.003 0.004 3.026 3.038 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.932 2.941 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.887 2.782 cp_fm_cholesky_invert 11 10.9 2.741 2.748 2.741 2.748 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.650 2.702 mp_alltoall_i22 627 13.8 1.666 2.570 1.666 2.570 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.525 2.527 calculate_dm_sparse 119 9.5 0.000 0.000 2.487 2.506 fft_wrap_pw1pw2_140 487 13.2 0.577 0.584 2.483 2.503 make_images_data 4572 15.5 0.046 0.048 1.875 2.464 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.558 2.429 hybrid_alltoall_any 4725 16.4 0.090 0.149 1.806 2.381 calculate_first_density_matrix 1 7.0 0.000 0.000 2.349 2.352 ot_diis_step 108 11.5 0.014 0.015 2.203 2.203 grid_integrate_task_list 119 12.3 2.172 2.181 2.172 2.181 fft3d_ps 1201 14.6 0.598 0.614 2.158 2.177 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.145 2.147 potential_pw2rs 119 12.3 0.014 0.014 2.127 2.139 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.093 2.093 cp_fm_diag_elpa_base 50 14.0 1.929 1.991 2.090 2.090 wfi_extrapolate 11 7.9 0.001 0.001 1.844 1.844 acc_transpose_blocks 9144 15.5 0.043 0.044 1.812 1.842 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.787 1.789 jit_kernel_multiply 5 15.8 0.968 1.774 0.968 1.774 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.748 1.762 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.660 1.722 grid_collocate_task_list 119 9.7 1.586 1.596 1.586 1.596 mp_alltoall_d11v 2130 13.8 1.544 1.573 1.544 1.573 mp_alltoall_z22v 1201 16.6 1.424 1.455 1.424 1.455 transfer_rs2pw 487 10.6 0.005 0.005 1.410 1.449 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.345 1.362 transfer_pw2rs 487 13.2 0.003 0.004 1.264 1.276 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.118 1.133 acc_transpose_blocks_sync 27432 16.5 1.097 1.121 1.097 1.121 make_images_sizes 4572 15.5 0.005 0.005 0.658 1.104 mp_alltoall_i44 4572 16.5 0.653 1.100 0.653 1.100 cp_fm_cholesky_decompose 22 10.9 1.089 1.099 1.089 1.099 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=54.017000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=725.727273, yerr=10.888617 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 501.522432E+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 2085352. 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 0.009 0.030 80.133 80.134 qs_mol_dyn_low 1 2.0 0.003 0.003 79.871 79.882 qs_forces 11 3.9 0.003 0.003 79.769 79.771 qs_energies 11 4.9 0.002 0.002 76.894 76.918 scf_env_do_scf 11 5.9 0.000 0.001 67.726 67.728 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 62.156 62.157 dbcsr_multiply_generic 2055 12.4 0.105 0.108 49.335 49.586 qs_scf_new_mos 99 7.5 0.000 0.001 45.050 45.153 qs_scf_loop_do_ot 99 8.5 0.000 0.001 45.050 45.153 multiply_cannon 2055 13.4 0.188 0.193 41.824 42.990 ot_scf_mini 99 9.5 0.002 0.002 42.777 42.870 multiply_cannon_loop 2055 14.4 1.770 1.807 40.776 41.918 velocity_verlet 10 3.0 0.002 0.002 41.454 41.455 ot_mini 99 10.5 0.001 0.001 24.898 24.983 multiply_cannon_multrec 49320 15.4 11.651 12.337 17.755 18.629 qs_ot_get_derivative 99 11.5 0.001 0.001 18.102 18.189 rebuild_ks_matrix 110 8.3 0.000 0.001 14.560 14.775 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.559 14.774 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.764 12.955 multiply_cannon_sync_h2d 49320 15.4 9.890 10.630 9.890 10.630 mp_waitall_1 220248 16.4 9.417 10.443 9.417 10.443 qs_ot_get_p 110 10.4 0.001 0.001 9.608 9.726 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.259 7.701 apply_single 110 13.6 0.000 0.001 7.258 7.701 init_scf_run 11 5.9 0.000 0.001 7.100 7.100 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.099 7.100 sum_up_and_integrate 110 10.3 0.002 0.003 7.036 7.050 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.488 7.040 integrate_v_rspace 110 11.3 0.003 0.003 7.011 7.030 multiply_cannon_metrocomm3 49320 15.4 0.084 0.088 5.564 6.944 ot_diis_step 99 11.5 0.005 0.006 6.633 6.633 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.526 6.541 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.361 6.476 calculate_rho_elec 110 8.6 0.020 0.024 6.361 6.476 dbcsr_mm_accdrv_process 87628 16.1 2.936 3.048 5.972 6.281 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.750 5.750 init_scf_loop 11 6.9 0.000 0.000 5.542 5.543 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.245 5.309 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.162 5.186 cp_fm_diag_elpa_base 48 14.0 5.148 5.171 5.160 5.184 make_m2s 4110 13.4 0.061 0.066 3.988 4.133 wfi_extrapolate 11 7.9 0.001 0.001 4.073 4.073 make_images 4110 14.4 0.176 0.189 3.892 4.041 calculate_dm_sparse 110 9.5 0.001 0.001 3.921 4.034 mp_sum_l 6594 12.7 3.066 3.900 3.066 3.900 density_rs2pw 110 9.6 0.004 0.005 3.412 3.573 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.537 3.542 grid_integrate_task_list 110 12.3 3.251 3.441 3.251 3.441 prepare_preconditioner 11 7.9 0.000 0.000 3.367 3.392 make_preconditioner 11 8.9 0.000 0.000 3.367 3.392 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.218 3.266 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.145 3.190 pw_transfer 1331 11.6 0.055 0.064 3.098 3.177 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.010 3.092 multiply_cannon_metrocomm1 49320 15.4 0.064 0.067 1.936 3.088 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.955 2.991 calculate_first_density_matrix 1 7.0 0.000 0.000 2.929 2.933 jit_kernel_multiply 13 15.9 2.760 2.807 2.760 2.807 potential_pw2rs 110 12.3 0.005 0.006 2.634 2.657 fft_wrap_pw1pw2_140 451 13.1 0.390 0.429 2.553 2.646 acc_transpose_blocks 49320 15.4 0.206 0.214 2.536 2.597 mp_alltoall_d11v 2046 13.8 2.064 2.554 2.064 2.554 fft3d_ps 1111 14.6 0.795 0.884 2.305 2.363 grid_collocate_task_list 110 9.6 2.159 2.291 2.159 2.291 transfer_rs2pw 451 10.6 0.005 0.006 1.944 2.126 mp_waitany 14300 13.8 1.819 2.119 1.819 2.119 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.927 1.967 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.926 1.948 make_images_data 4110 15.4 0.043 0.047 1.801 1.935 hybrid_alltoall_any 4261 16.3 0.085 0.479 1.565 1.856 cp_fm_cholesky_invert 11 10.9 1.842 1.846 1.842 1.846 transfer_pw2rs 451 13.1 0.006 0.007 1.794 1.806 mp_sum_d 3889 11.9 1.292 1.770 1.292 1.770 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=80.134000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.727273, yerr=2.489648 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 591.425536E+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 3098089. 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 0.010 0.026 69.770 69.770 qs_mol_dyn_low 1 2.0 0.003 0.003 69.538 69.547 qs_forces 11 3.9 0.003 0.004 69.455 69.457 qs_energies 11 4.9 0.001 0.002 66.085 66.088 scf_env_do_scf 11 5.9 0.001 0.001 57.386 57.389 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.676 49.677 dbcsr_multiply_generic 2055 12.4 0.116 0.120 38.258 38.388 velocity_verlet 10 3.0 0.003 0.011 36.432 36.434 qs_scf_new_mos 99 7.5 0.001 0.001 33.673 33.794 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.672 33.793 multiply_cannon 2055 13.4 0.225 0.245 31.543 32.605 ot_scf_mini 99 9.5 0.003 0.003 31.998 32.097 multiply_cannon_loop 2055 14.4 1.165 1.190 30.276 31.188 ot_mini 99 10.5 0.001 0.001 18.560 18.673 multiply_cannon_multrec 24660 15.4 7.045 8.808 14.234 15.857 rebuild_ks_matrix 110 8.3 0.000 0.001 13.615 13.692 qs_ks_build_kohn_sham_matrix 110 9.3 0.019 0.036 13.615 13.691 qs_ot_get_derivative 99 11.5 0.001 0.001 12.783 12.891 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.943 12.009 mp_waitall_1 176588 16.5 7.603 9.948 7.603 9.948 multiply_cannon_metrocomm3 24660 15.4 0.070 0.073 5.116 7.996 init_scf_loop 11 6.9 0.000 0.000 7.673 7.673 multiply_cannon_sync_h2d 24660 15.4 6.364 7.344 6.364 7.344 dbcsr_mm_accdrv_process 52282 16.1 5.706 6.600 7.018 7.319 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.488 7.179 apply_single 110 13.6 0.000 0.001 6.488 7.179 qs_ot_get_p 110 10.4 0.001 0.001 6.767 6.915 sum_up_and_integrate 110 10.3 0.001 0.003 6.474 6.486 integrate_v_rspace 110 11.3 0.003 0.003 6.447 6.461 init_scf_run 11 5.9 0.000 0.001 6.315 6.316 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.315 6.316 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.025 6.032 calculate_rho_elec 110 8.6 0.039 0.047 6.025 6.032 ot_diis_step 99 11.5 0.010 0.010 5.729 5.729 prepare_preconditioner 11 7.9 0.000 0.000 5.647 5.663 make_preconditioner 11 8.9 0.000 0.000 5.647 5.663 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.832 5.584 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.215 5.373 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.818 4.838 make_m2s 4110 13.4 0.057 0.059 4.308 4.718 make_images 4110 14.4 0.406 0.465 4.198 4.605 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.350 4.350 pw_transfer 1331 11.6 0.066 0.074 3.643 3.779 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.757 3.771 cp_fm_diag_elpa_base 48 14.0 3.708 3.736 3.754 3.768 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.536 3.676 wfi_extrapolate 11 7.9 0.001 0.001 3.555 3.556 density_rs2pw 110 9.6 0.004 0.005 3.266 3.434 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.340 3.403 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.336 3.338 grid_integrate_task_list 110 12.3 3.158 3.323 3.158 3.323 fft_wrap_pw1pw2_140 451 13.1 0.460 0.475 2.977 3.120 calculate_dm_sparse 110 9.5 0.001 0.001 3.086 3.108 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.983 3.008 fft3d_ps 1111 14.6 1.111 1.337 2.634 2.787 hybrid_alltoall_any 4261 16.3 0.107 0.459 2.025 2.782 make_images_data 4110 15.4 0.048 0.052 2.330 2.761 calculate_first_density_matrix 1 7.0 0.000 0.000 2.661 2.662 cp_fm_cholesky_invert 11 10.9 2.606 2.613 2.606 2.613 mp_sum_l 6594 12.7 1.777 2.370 1.777 2.370 potential_pw2rs 110 12.3 0.008 0.009 2.340 2.353 grid_collocate_task_list 110 9.6 2.167 2.333 2.167 2.333 acc_transpose_blocks 24660 15.4 0.114 0.118 2.039 2.125 mp_alltoall_d11v 2046 13.8 1.821 2.105 1.821 2.105 jit_kernel_multiply 10 16.2 0.951 2.031 0.951 2.031 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.976 1.997 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.870 1.872 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.830 1.841 transfer_rs2pw 451 10.6 0.006 0.007 1.511 1.675 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.647 1.656 multiply_cannon_metrocomm4 22605 15.4 0.076 0.080 0.770 1.641 mp_allgather_i34 2055 14.4 0.574 1.580 0.574 1.580 mp_irecv_dv 57340 16.2 0.645 1.520 0.645 1.520 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.385 1.487 dbcsr_complete_redistribute 325 12.2 0.236 0.301 1.212 1.483 mp_waitany 10164 13.8 1.212 1.405 1.212 1.405 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.770000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=559.818182, yerr=7.259078 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 668.188672E+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 0.030 59.235 59.236 qs_mol_dyn_low 1 2.0 0.003 0.003 58.936 58.945 qs_forces 11 3.9 0.003 0.003 58.614 58.615 qs_energies 11 4.9 0.006 0.016 55.454 55.458 scf_env_do_scf 11 5.9 0.001 0.001 47.678 47.679 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.284 39.286 velocity_verlet 10 3.0 0.001 0.002 32.351 32.361 dbcsr_multiply_generic 2055 12.4 0.106 0.111 28.428 28.658 qs_scf_new_mos 99 7.5 0.001 0.001 24.826 24.932 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.826 24.932 ot_scf_mini 99 9.5 0.002 0.002 23.580 23.693 multiply_cannon 2055 13.4 0.212 0.220 22.367 23.470 multiply_cannon_loop 2055 14.4 0.817 0.848 21.272 22.216 ot_mini 99 10.5 0.001 0.001 13.625 13.738 rebuild_ks_matrix 110 8.3 0.000 0.000 12.005 12.139 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.005 12.138 multiply_cannon_multrec 16440 15.4 3.843 5.325 10.010 11.390 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.562 10.678 mp_waitall_1 139946 16.5 6.581 9.622 6.581 9.622 qs_ot_get_derivative 99 11.5 0.001 0.001 9.255 9.372 init_scf_loop 11 6.9 0.000 0.001 8.361 8.365 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.135 7.063 prepare_preconditioner 11 7.9 0.000 0.000 6.626 6.642 make_preconditioner 11 8.9 0.000 0.001 6.626 6.642 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.927 6.284 sum_up_and_integrate 110 10.3 0.001 0.002 6.220 6.237 integrate_v_rspace 110 11.3 0.003 0.003 6.194 6.211 dbcsr_mm_accdrv_process 34862 16.1 5.293 5.689 6.011 6.122 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.658 5.669 calculate_rho_elec 110 8.6 0.058 0.059 5.658 5.668 init_scf_run 11 5.9 0.000 0.002 5.376 5.377 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.376 5.376 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.887 5.341 apply_single 110 13.6 0.000 0.000 4.887 5.341 qs_ot_get_p 110 10.4 0.001 0.002 5.214 5.337 make_m2s 4110 13.4 0.050 0.052 4.041 4.346 ot_diis_step 99 11.5 0.010 0.011 4.340 4.340 make_images 4110 14.4 0.399 0.526 3.927 4.232 multiply_cannon_sync_h2d 16440 15.4 3.254 3.914 3.254 3.914 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.167 3.794 qs_ot_p2m_diag 48 11.0 0.041 0.044 3.610 3.614 pw_transfer 1331 11.6 0.065 0.072 3.528 3.542 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.421 3.438 grid_integrate_task_list 110 12.3 3.182 3.426 3.182 3.426 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.246 3.246 density_rs2pw 110 9.6 0.004 0.004 2.904 3.085 fft_wrap_pw1pw2_140 451 13.1 0.580 0.586 2.969 2.985 wfi_extrapolate 11 7.9 0.001 0.001 2.926 2.926 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.730 2.739 cp_fm_diag_elpa_base 48 14.0 2.666 2.694 2.729 2.738 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.723 2.724 make_images_data 4110 15.4 0.045 0.050 2.243 2.628 hybrid_alltoall_any 4261 16.3 0.109 0.383 2.016 2.593 cp_fm_cholesky_invert 11 10.9 2.565 2.571 2.565 2.571 calculate_dm_sparse 110 9.5 0.001 0.001 2.515 2.552 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.416 2.479 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.413 2.449 fft3d_ps 1111 14.6 1.091 1.100 2.363 2.374 grid_collocate_task_list 110 9.6 2.218 2.372 2.218 2.372 calculate_first_density_matrix 1 7.0 0.001 0.004 2.361 2.364 multiply_cannon_metrocomm4 14385 15.4 0.047 0.051 0.849 2.317 mp_irecv_dv 48980 15.7 0.776 2.185 0.776 2.185 mp_alltoall_d11v 2046 13.8 1.741 2.184 1.741 2.184 potential_pw2rs 110 12.3 0.011 0.012 2.074 2.081 mp_sum_l 6594 12.7 1.402 1.991 1.402 1.991 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.957 1.958 dbcsr_complete_redistribute 325 12.2 0.326 0.353 1.423 1.901 cp_fm_upper_to_full 70 14.2 1.406 1.788 1.406 1.788 acc_transpose_blocks 16440 15.4 0.076 0.078 1.587 1.683 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.636 1.654 cp_fm_cholesky_decompose 22 10.9 1.534 1.555 1.534 1.555 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.453 1.468 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.376 1.453 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.980 1.439 transfer_rs2pw 451 10.6 0.005 0.006 1.223 1.425 mp_allgather_i34 2055 14.4 0.409 1.387 0.409 1.387 rs_gather_matrices 110 12.3 0.138 0.150 0.888 1.315 mp_waitany 17072 13.8 1.120 1.292 1.120 1.292 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.263 1.270 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.236000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=633.545455, yerr=7.866657 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 732.807168E+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 0.053 0.069 76.757 76.758 qs_mol_dyn_low 1 2.0 0.003 0.003 76.387 76.397 qs_forces 11 3.9 0.003 0.003 76.316 76.316 qs_energies 11 4.9 0.002 0.005 72.763 72.768 scf_env_do_scf 11 5.9 0.001 0.002 63.776 63.779 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 49.000 49.002 velocity_verlet 10 3.0 0.002 0.002 43.085 43.087 dbcsr_multiply_generic 2055 12.4 0.116 0.122 33.062 33.460 qs_scf_new_mos 99 7.5 0.001 0.001 32.741 32.846 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.740 32.845 ot_scf_mini 99 9.5 0.003 0.004 30.934 31.027 multiply_cannon 2055 13.4 0.237 0.252 24.171 25.682 multiply_cannon_loop 2055 14.4 1.417 1.504 22.371 23.528 ot_mini 99 10.5 0.001 0.001 16.722 16.850 init_scf_loop 11 6.9 0.001 0.002 14.721 14.722 multiply_cannon_multrec 24660 15.4 4.087 6.696 13.040 14.298 rebuild_ks_matrix 110 8.3 0.000 0.000 12.990 13.103 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.990 13.102 prepare_preconditioner 11 7.9 0.000 0.000 12.852 12.865 make_preconditioner 11 8.9 0.000 0.001 12.852 12.865 make_full_inverse_cholesky 11 9.9 0.000 0.000 11.013 12.512 qs_ot_get_derivative 99 11.5 0.001 0.001 12.384 12.480 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.485 11.587 dbcsr_mm_accdrv_process 52304 16.0 7.738 9.090 8.797 9.743 qs_ot_get_p 110 10.4 0.001 0.001 8.702 8.853 mp_waitall_1 121746 16.5 5.768 8.841 5.768 8.841 make_m2s 4110 13.4 0.060 0.062 6.756 7.125 sum_up_and_integrate 110 10.3 0.001 0.002 7.095 7.109 integrate_v_rspace 110 11.3 0.003 0.003 7.069 7.082 make_images 4110 14.4 0.575 0.696 6.611 6.982 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.750 6.765 calculate_rho_elec 110 8.6 0.078 0.081 6.750 6.764 qs_ot_p2m_diag 48 11.0 0.055 0.064 6.322 6.343 init_scf_run 11 5.9 0.000 0.001 6.069 6.069 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.068 6.069 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.649 5.650 cp_fm_upper_to_full 70 14.2 3.474 5.026 3.474 5.026 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.782 4.793 cp_fm_diag_elpa_base 48 14.0 4.608 4.669 4.779 4.790 cp_fm_cholesky_invert 11 10.9 4.549 4.561 4.549 4.561 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.501 4.378 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.150 4.290 apply_single 110 13.6 0.000 0.000 4.149 4.290 ot_diis_step 99 11.5 0.011 0.011 4.262 4.262 pw_transfer 1331 11.6 0.066 0.076 4.104 4.162 dbcsr_complete_redistribute 325 12.2 0.399 0.440 2.971 4.144 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 4.011 4.065 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.997 4.058 make_images_data 4110 15.4 0.049 0.053 3.484 4.047 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.974 4.024 density_rs2pw 110 9.6 0.004 0.004 3.818 3.952 hybrid_alltoall_any 4261 16.3 0.123 0.462 2.831 3.657 wfi_extrapolate 11 7.9 0.001 0.001 3.498 3.498 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.291 3.448 grid_integrate_task_list 110 12.3 3.291 3.428 3.291 3.428 fft_wrap_pw1pw2_140 451 13.1 0.609 0.635 3.246 3.312 calculate_dm_sparse 110 9.5 0.001 0.001 3.128 3.169 mp_alltoall_i22 605 13.7 1.775 2.983 1.775 2.983 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.830 2.982 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.918 2.920 fft3d_ps 1111 14.6 1.088 1.117 2.881 2.917 potential_pw2rs 110 12.3 0.012 0.013 2.845 2.864 cp_fm_cholesky_decompose 22 10.9 2.606 2.669 2.606 2.669 acc_transpose_blocks 24660 15.4 0.112 0.115 2.431 2.559 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.493 2.543 calculate_first_density_matrix 1 7.0 0.000 0.004 2.460 2.464 multiply_cannon_sync_h2d 24660 15.4 2.346 2.461 2.346 2.461 grid_collocate_task_list 110 9.6 2.264 2.416 2.264 2.416 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.301 2.302 mp_alltoall_d11v 2046 13.8 2.117 2.265 2.117 2.265 mp_allgather_i34 2055 14.4 0.886 2.136 0.886 2.136 mp_sum_l 6594 12.7 1.454 2.116 1.454 2.116 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.049 2.074 transfer_rs2pw 451 10.6 0.005 0.006 1.866 2.031 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.750 1.759 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.611 1.712 transfer_pw2rs 451 13.1 0.004 0.005 1.681 1.709 mp_alltoall_z22v 1111 16.6 1.575 1.654 1.575 1.654 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.584 1.600 multiply_cannon_metrocomm4 20550 15.4 0.060 0.065 0.852 1.555 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=76.758000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=694.727273, yerr=8.634928 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 854.720512E+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 0.010 0.031 62.675 62.676 qs_mol_dyn_low 1 2.0 0.003 0.003 62.433 62.443 qs_forces 11 3.9 0.003 0.003 62.360 62.367 qs_energies 11 4.9 0.001 0.002 58.633 58.645 scf_env_do_scf 11 5.9 0.000 0.001 49.596 49.596 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 39.858 39.859 velocity_verlet 10 3.0 0.002 0.002 35.253 35.256 dbcsr_multiply_generic 2055 12.4 0.107 0.111 25.240 25.438 qs_scf_new_mos 99 7.5 0.001 0.001 24.138 24.207 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.137 24.206 ot_scf_mini 99 9.5 0.002 0.002 22.769 22.821 multiply_cannon 2055 13.4 0.237 0.249 18.440 20.048 multiply_cannon_loop 2055 14.4 0.608 0.636 17.069 17.539 rebuild_ks_matrix 110 8.3 0.000 0.000 12.299 12.316 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.299 12.315 ot_mini 99 10.5 0.001 0.001 12.060 12.095 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.950 10.968 init_scf_loop 11 6.9 0.000 0.000 9.683 9.685 mp_waitall_1 103326 16.6 7.330 9.123 7.330 9.123 multiply_cannon_multrec 8220 15.4 3.187 4.658 7.624 8.666 qs_ot_get_derivative 99 11.5 0.001 0.001 8.284 8.331 prepare_preconditioner 11 7.9 0.000 0.000 7.932 7.937 make_preconditioner 11 8.9 0.000 0.000 7.932 7.937 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.435 7.554 sum_up_and_integrate 110 10.3 0.001 0.002 6.813 6.824 integrate_v_rspace 110 11.3 0.003 0.003 6.786 6.798 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.680 6.709 calculate_rho_elec 110 8.6 0.114 0.114 6.679 6.708 qs_ot_get_p 110 10.4 0.001 0.001 6.545 6.559 init_scf_run 11 5.9 0.000 0.001 5.766 5.766 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.766 5.766 make_m2s 4110 13.4 0.038 0.040 5.115 5.402 make_images 4110 14.4 0.648 0.715 4.983 5.271 dbcsr_mm_accdrv_process 17442 15.9 3.094 4.184 4.299 5.262 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.790 4.802 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.357 4.358 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 2.896 4.343 pw_transfer 1331 11.6 0.066 0.071 4.314 4.329 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.206 4.223 cp_fm_cholesky_invert 11 10.9 3.953 3.958 3.953 3.958 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.791 3.841 apply_single 110 13.6 0.000 0.000 3.791 3.841 density_rs2pw 110 9.6 0.004 0.004 3.616 3.761 ot_diis_step 99 11.5 0.012 0.012 3.729 3.730 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.688 3.696 cp_fm_diag_elpa_base 48 14.0 3.614 3.652 3.686 3.693 fft_wrap_pw1pw2_140 451 13.1 0.775 0.789 3.507 3.536 grid_integrate_task_list 110 12.3 3.354 3.509 3.354 3.509 make_images_data 4110 15.4 0.041 0.046 2.724 3.204 hybrid_alltoall_any 4261 16.3 0.201 0.859 2.628 3.168 wfi_extrapolate 11 7.9 0.001 0.001 3.124 3.124 calculate_dm_sparse 110 9.5 0.001 0.001 2.883 2.924 fft3d_ps 1111 14.6 1.141 1.184 2.873 2.894 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.711 2.719 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.588 2.591 multiply_cannon_sync_h2d 8220 15.4 2.353 2.543 2.353 2.543 calculate_first_density_matrix 1 7.0 0.000 0.000 2.533 2.536 potential_pw2rs 110 12.3 0.015 0.015 2.521 2.534 grid_collocate_task_list 110 9.6 2.362 2.495 2.362 2.495 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.452 2.495 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.463 2.493 cp_fm_cholesky_decompose 22 10.9 2.393 2.410 2.393 2.410 mp_alltoall_d11v 2046 13.8 2.065 2.352 2.065 2.352 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.137 2.146 multiply_cannon_metrocomm1 8220 15.4 0.022 0.022 1.523 2.095 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.785 1.990 dbcsr_complete_redistribute 325 12.2 0.548 0.564 1.733 1.870 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.744 1.760 transfer_rs2pw 451 10.6 0.005 0.005 1.587 1.739 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.680 1.684 mp_allgather_i34 2055 14.4 0.538 1.682 0.538 1.682 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.542 1.656 mp_alltoall_z22v 1111 16.6 1.493 1.516 1.493 1.516 make_images_sizes 4110 15.4 0.005 0.005 0.786 1.475 mp_alltoall_i44 4110 16.4 0.781 1.470 0.781 1.470 acc_transpose_blocks 8220 15.4 0.039 0.040 1.375 1.417 mp_waitany 9240 13.8 1.192 1.367 1.192 1.367 qs_create_task_list 11 7.9 0.001 0.001 1.247 1.343 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.323 1.343 generate_qs_task_list 11 8.9 0.393 0.464 1.247 1.343 transfer_pw2rs 451 13.1 0.004 0.004 1.326 1.341 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=62.676000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=808.272727, yerr=11.021392 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 1.384063E+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 0.038 0.054 87.117 87.118 qs_mol_dyn_low 1 2.0 0.003 0.003 86.735 86.744 qs_forces 11 3.9 0.004 0.007 86.654 86.657 qs_energies 11 4.9 0.001 0.001 82.576 82.580 scf_env_do_scf 11 5.9 0.001 0.001 72.539 72.539 velocity_verlet 10 3.0 0.002 0.003 55.572 55.578 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 44.207 44.209 dbcsr_multiply_generic 2055 12.4 0.120 0.125 29.429 29.520 init_scf_loop 11 6.9 0.000 0.000 28.258 28.262 qs_scf_new_mos 99 7.5 0.001 0.001 26.965 27.011 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.964 27.011 prepare_preconditioner 11 7.9 0.000 0.000 26.271 26.282 make_preconditioner 11 8.9 0.000 0.000 26.271 26.282 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.456 25.752 ot_scf_mini 99 9.5 0.002 0.002 25.166 25.209 multiply_cannon 2055 13.4 0.335 0.369 22.365 23.183 multiply_cannon_loop 2055 14.4 0.824 0.835 20.596 20.897 cp_fm_upper_to_full 70 14.2 12.680 18.233 12.680 18.233 ot_mini 99 10.5 0.001 0.001 14.088 14.130 rebuild_ks_matrix 110 8.3 0.000 0.001 13.536 13.570 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.536 13.569 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.287 12.317 dbcsr_complete_redistribute 325 12.2 1.011 1.054 7.394 10.577 multiply_cannon_multrec 8220 15.4 4.035 4.183 9.716 9.812 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.369 9.558 qs_ot_get_derivative 99 11.5 0.001 0.001 9.507 9.551 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.799 8.957 mp_alltoall_i22 605 13.7 5.425 8.635 5.425 8.635 mp_waitall_1 84994 16.7 7.479 8.254 7.479 8.254 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.088 7.123 calculate_rho_elec 110 8.6 0.225 0.225 7.088 7.122 sum_up_and_integrate 110 10.3 0.002 0.002 6.720 6.735 integrate_v_rspace 110 11.3 0.004 0.004 6.692 6.706 qs_ot_get_p 110 10.4 0.001 0.001 5.786 5.845 make_m2s 4110 13.4 0.044 0.045 5.334 5.811 dbcsr_mm_accdrv_process 11614 15.7 3.877 4.106 5.533 5.737 init_scf_run 11 5.9 0.000 0.001 5.699 5.699 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.699 5.699 make_images 4110 14.4 0.882 0.933 5.144 5.621 cp_fm_cholesky_invert 11 10.9 5.277 5.281 5.277 5.281 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.849 5.216 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.718 5.179 apply_single 110 13.6 0.000 0.000 4.717 5.178 pw_transfer 1331 11.6 0.074 0.074 4.923 4.934 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.807 4.818 ot_diis_step 99 11.5 0.015 0.016 4.563 4.563 fft_wrap_pw1pw2_140 451 13.1 1.280 1.285 4.192 4.202 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.186 4.195 grid_integrate_task_list 110 12.3 3.679 3.747 3.679 3.747 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.728 3.730 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.665 3.666 density_rs2pw 110 9.6 0.004 0.004 3.540 3.556 hybrid_alltoall_any 4261 16.3 0.263 0.561 2.795 3.496 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.016 3.473 make_images_data 4110 15.4 0.045 0.048 2.787 3.392 wfi_extrapolate 11 7.9 0.001 0.001 3.313 3.313 calculate_dm_sparse 110 9.5 0.001 0.001 3.164 3.194 multiply_cannon_sync_h2d 8220 15.4 3.126 3.151 3.126 3.151 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.071 3.071 cp_fm_diag_elpa_base 48 14.0 2.522 2.729 3.069 3.069 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.903 2.908 fft3d_ps 1111 14.6 1.301 1.313 2.859 2.866 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.667 2.693 grid_collocate_task_list 110 9.6 2.669 2.689 2.669 2.689 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.401 2.413 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.240 2.283 calculate_first_density_matrix 1 7.0 0.000 0.000 2.271 2.273 potential_pw2rs 110 12.3 0.021 0.021 2.228 2.232 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.092 2.172 mp_alltoall_d11v 2046 13.8 2.059 2.124 2.059 2.124 cp_fm_cholesky_decompose 22 10.9 2.003 2.019 2.003 2.019 qs_create_task_list 11 7.9 0.010 0.021 1.941 1.987 generate_qs_task_list 11 8.9 0.771 0.827 1.931 1.967 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.851 1.854 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.740 1.774 jit_kernel_multiply 10 15.1 1.463 1.759 1.463 1.759 acc_transpose_blocks 8220 15.4 0.040 0.041 1.727 1.747 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.118000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1240.636364, yerr=55.113557 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 630.521856E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 1883142. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.029 212.718 212.720 qs_mol_dyn_low 1 2.0 0.003 0.004 212.276 212.290 qs_forces 11 3.9 0.005 0.005 212.152 212.154 qs_energies 11 4.9 0.001 0.002 206.387 206.398 scf_env_do_scf 11 5.9 0.001 0.002 189.296 189.300 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 166.330 166.332 velocity_verlet 10 3.0 0.001 0.002 128.526 128.527 qs_scf_new_mos 117 7.6 0.001 0.001 123.964 124.230 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.963 124.230 dbcsr_multiply_generic 2507 12.6 0.180 0.186 122.017 122.618 ot_scf_mini 117 9.6 0.003 0.003 117.396 117.660 multiply_cannon 2507 13.6 0.245 0.258 100.494 102.151 multiply_cannon_loop 2507 14.6 2.289 2.364 98.067 99.733 ot_mini 117 10.6 0.001 0.001 63.624 63.895 multiply_cannon_multrec 60168 15.6 32.682 35.123 42.100 44.765 qs_ot_get_derivative 117 11.6 0.001 0.001 38.573 38.845 rebuild_ks_matrix 128 8.3 0.001 0.001 35.381 35.648 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.018 35.381 35.647 qs_ot_get_p 128 10.4 0.001 0.001 32.013 32.497 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.797 32.051 mp_waitall_1 267128 16.5 27.738 31.408 27.738 31.408 multiply_cannon_sync_h2d 60168 15.6 27.729 29.851 27.729 29.851 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.636 25.373 apply_single 128 13.6 0.001 0.001 24.636 25.372 qs_ot_p2m_diag 83 11.4 0.079 0.091 25.023 25.073 ot_diis_step 117 11.6 0.007 0.008 24.878 24.879 cp_dbcsr_syevd 83 12.4 0.005 0.005 22.895 22.897 init_scf_loop 11 6.9 0.000 0.001 22.882 22.884 cp_fm_diag_elpa 83 13.4 0.000 0.000 19.421 19.453 cp_fm_diag_elpa_base 83 14.4 19.308 19.348 19.416 19.450 prepare_preconditioner 11 7.9 0.000 0.000 18.098 18.164 make_preconditioner 11 8.9 0.000 0.000 18.098 18.164 make_full_inverse_cholesky 11 9.9 0.000 0.000 17.310 17.490 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 16.853 17.060 multiply_cannon_metrocomm3 60168 15.6 0.122 0.126 13.985 16.068 sum_up_and_integrate 128 10.3 0.002 0.004 15.593 15.607 integrate_v_rspace 128 11.3 0.004 0.004 15.534 15.552 make_m2s 5014 13.6 0.104 0.112 14.956 15.291 make_images 5014 14.6 0.398 0.417 14.773 15.116 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.821 14.875 calculate_rho_elec 128 8.7 0.045 0.064 14.821 14.874 init_scf_run 11 5.9 0.000 0.001 12.798 12.799 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.798 12.799 cp_fm_cholesky_invert 11 10.9 10.408 10.416 10.408 10.416 wfi_extrapolate 11 7.9 0.001 0.001 9.558 9.558 dbcsr_mm_accdrv_process 124484 16.2 4.324 4.572 8.924 9.542 density_rs2pw 128 9.7 0.006 0.007 8.543 9.096 pw_transfer 1547 11.6 0.075 0.091 8.467 8.752 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.264 8.552 calculate_dm_sparse 128 9.5 0.001 0.001 8.420 8.507 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.213 8.355 make_images_data 5014 15.6 0.066 0.071 7.331 8.277 multiply_cannon_metrocomm1 60168 15.6 0.095 0.099 5.532 7.916 grid_integrate_task_list 128 12.3 6.967 7.748 6.967 7.748 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.509 7.596 hybrid_alltoall_any 5200 16.5 0.296 2.268 6.360 7.555 fft_wrap_pw1pw2_140 523 13.2 1.109 1.171 6.812 7.100 mp_sum_l 7950 12.9 5.277 6.846 5.277 6.846 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.729 6.743 fft3d_ps 1291 14.7 2.192 2.838 6.288 6.727 potential_pw2rs 128 12.3 0.009 0.010 5.987 6.088 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.821 5.935 mp_alltoall_d11v 2415 14.1 4.628 5.911 4.628 5.911 cp_fm_cholesky_decompose 22 10.9 5.439 5.451 5.439 5.451 grid_collocate_task_list 128 9.7 4.841 5.317 4.841 5.317 transfer_rs2pw 523 10.6 0.007 0.009 4.348 4.985 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=212.720000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.909091, yerr=6.287861 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 841.961472E+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 4002 58214. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 5092893. 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 0.010 0.033 212.804 212.806 qs_mol_dyn_low 1 2.0 0.003 0.004 212.426 212.445 qs_forces 11 3.9 0.004 0.005 212.330 212.331 qs_energies 11 4.9 0.001 0.002 205.346 205.360 scf_env_do_scf 11 5.9 0.001 0.001 188.210 188.220 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 149.178 149.181 velocity_verlet 10 3.0 0.001 0.002 133.348 133.350 qs_scf_new_mos 117 7.6 0.001 0.001 107.409 107.958 qs_scf_loop_do_ot 117 8.6 0.001 0.001 107.408 107.957 dbcsr_multiply_generic 2507 12.6 0.193 0.199 101.662 103.598 ot_scf_mini 117 9.6 0.004 0.004 102.498 103.099 multiply_cannon 2507 13.6 0.505 0.568 78.380 85.936 multiply_cannon_loop 2507 14.6 1.579 1.656 74.520 79.806 ot_mini 117 10.6 0.001 0.001 53.957 54.583 mp_waitall_1 214728 16.6 27.551 42.455 27.551 42.455 init_scf_loop 11 6.9 0.000 0.000 38.927 38.929 multiply_cannon_multrec 30084 15.6 20.561 25.234 31.071 36.191 rebuild_ks_matrix 128 8.3 0.001 0.001 34.116 34.797 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 34.115 34.797 prepare_preconditioner 11 7.9 0.000 0.000 34.408 34.495 make_preconditioner 11 8.9 0.000 0.000 34.408 34.495 multiply_cannon_metrocomm3 30084 15.6 0.096 0.106 17.055 33.800 make_full_inverse_cholesky 11 9.9 0.000 0.000 32.986 33.579 qs_ot_get_derivative 117 11.6 0.001 0.001 31.591 32.188 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.711 31.344 qs_ot_get_p 128 10.4 0.001 0.001 30.041 30.787 qs_ot_p2m_diag 83 11.4 0.188 0.217 24.385 24.507 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.382 23.888 apply_single 128 13.6 0.001 0.001 22.382 23.888 cp_dbcsr_syevd 83 12.4 0.005 0.006 22.913 22.916 ot_diis_step 117 11.6 0.014 0.015 22.150 22.153 cp_fm_cholesky_invert 11 10.9 20.326 20.341 20.326 20.341 multiply_cannon_sync_h2d 30084 15.6 17.555 20.032 17.555 20.032 cp_fm_diag_elpa 83 13.4 0.000 0.000 19.012 19.052 cp_fm_diag_elpa_base 83 14.4 18.702 18.822 19.005 19.040 make_m2s 5014 13.6 0.089 0.096 15.791 17.520 make_images 5014 14.6 1.170 1.387 15.581 17.305 sum_up_and_integrate 128 10.3 0.002 0.004 15.553 15.581 integrate_v_rspace 128 11.3 0.003 0.004 15.493 15.525 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.132 15.162 calculate_rho_elec 128 8.7 0.087 0.104 15.132 15.161 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 12.908 13.366 init_scf_run 11 5.9 0.000 0.001 12.171 12.173 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.171 12.173 multiply_cannon_metrocomm4 27577 15.6 0.102 0.117 3.925 11.474 make_images_data 5014 15.6 0.067 0.076 9.223 11.412 mp_irecv_dv 69486 16.3 3.724 11.061 3.724 11.061 hybrid_alltoall_any 5200 16.5 0.354 1.556 7.731 10.684 dbcsr_mm_accdrv_process 62242 16.2 5.484 6.090 9.995 10.567 pw_transfer 1547 11.6 0.085 0.098 9.282 9.365 density_rs2pw 128 9.7 0.006 0.007 8.767 9.302 cp_fm_cholesky_decompose 22 10.9 9.085 9.190 9.085 9.190 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.058 9.145 wfi_extrapolate 11 7.9 0.001 0.001 9.014 9.014 mp_sum_l 7950 12.9 5.697 7.744 5.697 7.744 fft_wrap_pw1pw2_140 523 13.2 1.210 1.234 7.641 7.735 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.834 7.666 grid_integrate_task_list 128 12.3 7.180 7.518 7.180 7.518 calculate_dm_sparse 128 9.5 0.001 0.001 6.586 6.724 fft3d_ps 1291 14.7 2.839 3.014 6.622 6.675 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.236 6.246 potential_pw2rs 128 12.3 0.016 0.018 5.978 6.022 mp_allgather_i34 2507 14.6 2.381 5.837 2.381 5.837 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.716 5.823 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.258 5.472 grid_collocate_task_list 128 9.7 4.999 5.377 4.999 5.377 mp_alltoall_d11v 2415 14.1 4.625 5.204 4.625 5.204 transfer_rs2pw 523 10.6 0.007 0.008 4.241 4.741 dbcsr_complete_redistribute 395 12.7 0.770 0.845 3.613 4.555 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=212.806000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=801.000000, yerr=4.880387 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/17/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 2.906045E+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 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 984.678400E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+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 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172905552 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58278. MP_Allreduce 11070 1000. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.062 200.351 200.359 qs_mol_dyn_low 1 2.0 0.003 0.004 199.894 199.917 qs_forces 11 3.9 0.004 0.004 199.578 199.580 qs_energies 11 4.9 0.001 0.002 192.826 192.839 scf_env_do_scf 11 5.9 0.001 0.002 176.227 176.227 scf_env_do_scf_inner_loop 116 6.6 0.003 0.011 133.429 133.430 velocity_verlet 10 3.0 0.002 0.002 127.197 127.206 qs_scf_new_mos 116 7.6 0.001 0.001 94.060 94.497 qs_scf_loop_do_ot 116 8.6 0.001 0.001 94.060 94.496 ot_scf_mini 116 9.6 0.003 0.004 89.687 90.161 dbcsr_multiply_generic 2485 12.5 0.182 0.188 86.384 87.924 multiply_cannon 2485 13.5 0.503 0.528 64.043 68.393 multiply_cannon_loop 2485 14.5 1.124 1.185 60.371 64.138 ot_mini 116 10.6 0.001 0.001 46.006 46.514 init_scf_loop 11 6.9 0.000 0.000 42.683 42.684 mp_waitall_1 169034 16.6 28.261 40.022 28.261 40.022 prepare_preconditioner 11 7.9 0.000 0.000 38.491 38.550 make_preconditioner 11 8.9 0.000 0.000 38.491 38.550 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.961 37.510 rebuild_ks_matrix 127 8.3 0.001 0.001 31.562 32.160 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 31.562 32.160 qs_ot_get_p 127 10.4 0.001 0.001 28.507 29.093 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.436 28.988 multiply_cannon_metrocomm3 19880 15.5 0.062 0.067 17.030 27.999 qs_ot_get_derivative 116 11.6 0.001 0.002 26.198 26.675 multiply_cannon_multrec 19880 15.5 12.625 15.431 21.829 24.626 qs_ot_p2m_diag 82 11.4 0.262 0.269 23.474 23.546 cp_dbcsr_syevd 82 12.4 0.005 0.005 22.076 22.080 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.949 21.131 apply_single 127 13.6 0.001 0.001 19.949 21.131 ot_diis_step 116 11.6 0.017 0.018 19.664 19.664 cp_fm_cholesky_invert 11 10.9 19.100 19.117 19.100 19.117 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.826 17.847 cp_fm_diag_elpa_base 82 14.4 17.341 17.574 17.821 17.843 make_m2s 4970 13.5 0.079 0.082 16.052 17.510 make_images 4970 14.5 1.145 1.238 15.822 17.280 sum_up_and_integrate 127 10.3 0.002 0.003 15.188 15.212 multiply_cannon_sync_h2d 19880 15.5 13.262 15.181 13.262 15.181 integrate_v_rspace 127 11.3 0.003 0.004 15.128 15.150 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.859 14.939 calculate_rho_elec 127 8.7 0.129 0.143 14.858 14.938 make_images_data 4970 15.5 0.062 0.071 9.562 11.635 init_scf_run 11 5.9 0.000 0.001 11.413 11.414 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.413 11.413 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.757 11.100 hybrid_alltoall_any 5155 16.4 0.449 2.035 8.203 10.296 cp_fm_cholesky_decompose 22 10.9 9.929 9.978 9.929 9.978 multiply_cannon_metrocomm4 17395 15.5 0.064 0.076 3.570 9.840 mp_irecv_dv 49801 16.2 3.445 9.572 3.445 9.572 pw_transfer 1535 11.6 0.085 0.102 9.207 9.332 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.984 9.113 dbcsr_mm_accdrv_process 41158 16.2 5.610 5.891 8.711 8.819 density_rs2pw 127 9.7 0.006 0.006 8.218 8.597 wfi_extrapolate 11 7.9 0.001 0.001 8.122 8.122 fft_wrap_pw1pw2_140 519 13.2 1.292 1.328 7.649 7.785 cp_fm_upper_to_full 104 14.8 6.027 7.701 6.027 7.701 grid_integrate_task_list 127 12.3 7.261 7.679 7.261 7.679 dbcsr_complete_redistribute 393 12.7 1.176 1.219 5.094 7.040 mp_sum_l 7884 12.9 4.498 6.829 4.498 6.829 fft3d_ps 1281 14.7 2.736 2.964 6.381 6.477 calculate_dm_sparse 127 9.5 0.001 0.001 6.209 6.298 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.142 5.825 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.762 5.710 potential_pw2rs 127 12.3 0.020 0.022 5.616 5.657 mp_alltoall_d11v 2401 14.1 4.689 5.654 4.689 5.654 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.598 5.608 mp_allgather_i34 2485 14.5 2.181 5.458 2.181 5.458 grid_collocate_task_list 127 9.7 5.119 5.419 5.119 5.419 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.593 4.780 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.372 4.407 transfer_fm_to_dbcsr 11 9.9 0.019 0.024 2.511 4.372 transfer_rs2pw 519 10.6 0.006 0.007 3.765 4.225 mp_alltoall_i22 712 14.1 2.210 4.184 2.210 4.184 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=200.359000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=928.090909, yerr=18.213155 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.145565E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+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 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58132. MP_Allreduce 11095 1083. MP_Sync 86 MP_Alltoall 1700 12496371. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.040 207.789 207.790 qs_mol_dyn_low 1 2.0 0.003 0.003 207.367 207.381 qs_forces 11 3.9 0.004 0.004 207.106 207.122 qs_energies 11 4.9 0.001 0.002 199.872 199.885 scf_env_do_scf 11 5.9 0.001 0.002 182.428 182.439 velocity_verlet 10 3.0 0.001 0.002 135.715 135.725 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 129.754 129.756 qs_scf_new_mos 116 7.6 0.001 0.001 91.973 92.310 qs_scf_loop_do_ot 116 8.6 0.001 0.001 91.972 92.309 ot_scf_mini 116 9.6 0.003 0.003 87.390 87.717 dbcsr_multiply_generic 2485 12.5 0.187 0.193 82.435 83.279 multiply_cannon 2485 13.5 0.541 0.582 56.115 59.999 multiply_cannon_loop 2485 14.5 1.788 1.868 52.139 54.580 init_scf_loop 11 6.9 0.000 0.000 52.539 52.541 prepare_preconditioner 11 7.9 0.000 0.000 48.337 48.358 make_preconditioner 11 8.9 0.000 0.000 48.337 48.358 make_full_inverse_cholesky 11 9.9 0.012 0.023 41.925 46.908 ot_mini 116 10.6 0.001 0.001 44.724 45.055 multiply_cannon_multrec 29820 15.5 13.435 19.097 26.445 31.784 rebuild_ks_matrix 127 8.3 0.001 0.001 30.134 30.547 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.020 30.133 30.546 qs_ot_get_p 127 10.4 0.001 0.001 28.567 28.909 mp_waitall_1 146592 16.7 18.997 28.075 18.997 28.075 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.197 27.571 qs_ot_get_derivative 116 11.6 0.001 0.002 25.123 25.448 qs_ot_p2m_diag 82 11.4 0.338 0.386 23.809 23.877 make_m2s 4970 13.5 0.096 0.101 21.507 22.411 cp_dbcsr_syevd 82 12.4 0.005 0.006 22.281 22.283 make_images 4970 14.5 1.934 2.236 21.196 22.100 cp_fm_cholesky_invert 11 10.9 20.185 20.198 20.185 20.198 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.994 19.596 apply_single 127 13.6 0.001 0.001 18.993 19.596 ot_diis_step 116 11.6 0.017 0.020 19.469 19.471 cp_fm_diag_elpa 82 13.4 0.000 0.000 18.247 18.263 cp_fm_diag_elpa_base 82 14.4 17.220 17.543 18.240 18.255 cp_fm_upper_to_full 104 14.8 11.572 16.986 11.572 16.986 multiply_cannon_metrocomm3 29820 15.5 0.049 0.052 6.499 15.460 sum_up_and_integrate 127 10.3 0.002 0.003 14.892 14.916 integrate_v_rspace 127 11.3 0.003 0.004 14.832 14.858 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.683 14.722 calculate_rho_elec 127 8.7 0.171 0.188 14.682 14.721 make_images_data 4970 15.5 0.066 0.072 11.695 13.355 dbcsr_mm_accdrv_process 61748 16.2 8.476 9.483 12.581 13.103 dbcsr_complete_redistribute 393 12.7 1.495 1.604 9.365 13.086 hybrid_alltoall_any 5155 16.4 0.528 2.205 10.267 12.145 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.847 11.567 init_scf_run 11 5.9 0.000 0.001 11.452 11.453 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.452 11.453 multiply_cannon_sync_h2d 29820 15.5 10.330 11.410 10.330 11.410 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.628 10.860 transfer_fm_to_dbcsr 11 9.9 0.002 0.007 6.390 10.021 cp_fm_cholesky_decompose 22 10.9 9.538 9.631 9.538 9.631 mp_alltoall_i22 712 14.1 5.634 9.372 5.634 9.372 pw_transfer 1535 11.6 0.085 0.099 9.220 9.328 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.996 9.110 density_rs2pw 127 9.7 0.006 0.006 7.740 8.132 wfi_extrapolate 11 7.9 0.001 0.001 8.076 8.076 grid_integrate_task_list 127 12.3 7.506 7.977 7.506 7.977 fft_wrap_pw1pw2_140 519 13.2 1.433 1.476 7.800 7.928 multiply_cannon_metrocomm4 24850 15.5 0.081 0.094 2.836 7.807 mp_irecv_dv 75445 16.2 2.686 7.519 2.686 7.519 calculate_dm_sparse 127 9.5 0.001 0.001 6.507 6.600 fft3d_ps 1281 14.7 2.824 2.905 6.197 6.269 mp_alltoall_d11v 2401 14.1 5.265 5.845 5.265 5.845 grid_collocate_task_list 127 9.7 5.240 5.550 5.240 5.550 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.466 5.521 potential_pw2rs 127 12.3 0.022 0.023 5.171 5.215 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.819 4.905 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.445 4.516 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.414 4.418 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.346 4.417 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=207.790000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1083.818182, yerr=18.139952 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 5.909858E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.549111E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+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 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591807936 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4040 57667. MP_Allreduce 11177 1163. MP_Sync 87 MP_Alltoall 1724 18848449. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.032 192.458 192.458 qs_mol_dyn_low 1 2.0 0.003 0.003 191.956 191.969 qs_forces 11 3.9 0.013 0.016 191.756 191.758 qs_energies 11 4.9 0.001 0.002 184.237 184.254 scf_env_do_scf 11 5.9 0.001 0.001 166.243 166.254 velocity_verlet 10 3.0 0.002 0.002 125.572 125.582 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 124.038 124.040 qs_scf_new_mos 118 7.6 0.001 0.001 84.996 85.102 qs_scf_loop_do_ot 118 8.6 0.001 0.001 84.995 85.102 ot_scf_mini 118 9.6 0.003 0.004 80.396 80.488 dbcsr_multiply_generic 2527 12.6 0.185 0.195 75.355 75.728 multiply_cannon 2527 13.6 0.568 0.600 54.754 59.411 multiply_cannon_loop 2527 14.6 0.821 0.860 51.425 52.528 init_scf_loop 11 6.9 0.000 0.000 42.043 42.045 ot_mini 118 10.6 0.001 0.001 40.009 40.053 prepare_preconditioner 11 7.9 0.000 0.000 38.118 38.140 make_preconditioner 11 8.9 0.000 0.000 38.118 38.140 make_full_inverse_cholesky 11 9.9 0.015 0.025 35.983 36.336 mp_waitall_1 126780 16.7 26.830 33.146 26.830 33.146 rebuild_ks_matrix 129 8.3 0.001 0.001 30.326 30.395 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 30.325 30.395 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.689 27.752 qs_ot_get_p 129 10.4 0.001 0.001 27.177 27.223 cp_fm_cholesky_invert 11 10.9 22.887 22.898 22.887 22.898 qs_ot_p2m_diag 83 11.4 0.495 0.501 22.803 22.828 qs_ot_get_derivative 118 11.6 0.001 0.002 22.715 22.807 cp_dbcsr_syevd 83 12.4 0.005 0.006 21.443 21.445 multiply_cannon_multrec 10108 15.6 10.472 14.478 18.048 20.898 multiply_cannon_metrocomm3 10108 15.6 0.025 0.027 12.633 19.513 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.638 17.644 cp_fm_diag_elpa_base 83 14.4 17.337 17.472 17.634 17.639 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.208 17.560 apply_single 129 13.6 0.001 0.001 17.208 17.560 ot_diis_step 118 11.6 0.020 0.022 17.158 17.158 make_m2s 5054 13.6 0.068 0.072 16.077 16.966 make_images 5054 14.6 2.187 2.631 15.767 16.653 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.229 15.252 calculate_rho_elec 129 8.7 0.258 0.269 15.229 15.251 sum_up_and_integrate 129 10.3 0.002 0.002 15.171 15.211 integrate_v_rspace 129 11.3 0.004 0.004 15.110 15.152 multiply_cannon_sync_h2d 10108 15.6 10.899 11.195 10.899 11.195 init_scf_run 11 5.9 0.000 0.001 11.043 11.044 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.043 11.044 make_images_data 5054 15.6 0.056 0.065 9.130 10.822 hybrid_alltoall_any 5240 16.5 0.852 3.806 8.743 10.537 pw_transfer 1559 11.6 0.086 0.095 9.781 9.859 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.556 9.642 cp_fm_cholesky_decompose 22 10.9 9.495 9.637 9.495 9.637 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.101 9.160 multiply_cannon_metrocomm1 10108 15.6 0.029 0.031 5.198 8.358 fft_wrap_pw1pw2_140 527 13.2 1.813 1.852 8.266 8.353 grid_integrate_task_list 129 12.3 7.849 8.233 7.849 8.233 density_rs2pw 129 9.7 0.005 0.006 7.760 7.951 wfi_extrapolate 11 7.9 0.001 0.001 7.916 7.916 dbcsr_mm_accdrv_process 20926 16.1 3.429 4.468 7.226 7.901 calculate_dm_sparse 129 9.5 0.001 0.001 6.262 6.349 fft3d_ps 1301 14.7 2.786 2.856 6.266 6.317 dbcsr_complete_redistribute 395 12.7 2.090 2.133 5.522 5.962 grid_collocate_task_list 129 9.7 5.619 5.908 5.619 5.908 mp_alltoall_d11v 2423 14.1 4.995 5.846 4.995 5.846 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.327 5.336 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.247 5.257 potential_pw2rs 129 12.3 0.026 0.026 5.161 5.199 mp_allgather_i34 2527 14.6 1.503 5.116 1.503 5.116 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.173 4.215 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.989 4.076 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.914 4.005 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.522 3.879 multiply_cannon_metrocomm4 7581 15.6 0.026 0.028 1.802 3.862 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.582 3.858 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=192.458000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1477.090909, yerr=1.083307 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 1967095021568 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 1992004743168 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 15239170400256 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239170400256 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911117062144 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.237285E+12 0.0% 0.0% 100.0% flops max/rank 11.786603E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806448352 0.0% 0.0% 100.0% number of processed stacks 1981024 0.0% 0.0% 100.0% average stack size 0.0 0.0 3435.8 marketing flops 145.655226E+12 ------------------------------------------------------------------------------- # multiplications 2531 max memory usage/rank 3.203342E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101240 MPI messages size (bytes): total size 1.145036E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.310119E+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 45728 35500589056 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592178544 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4080 58046. MP_Allreduce 11277 1498. MP_Sync 87 MP_Alltoall 1724 36993574. 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 0.031 0.073 290.836 290.844 qs_mol_dyn_low 1 2.0 0.003 0.004 290.240 290.261 qs_forces 11 3.9 0.004 0.005 290.119 290.122 qs_energies 11 4.9 0.002 0.002 281.272 281.279 scf_env_do_scf 11 5.9 0.001 0.001 258.714 258.721 velocity_verlet 10 3.0 0.002 0.002 210.153 210.164 scf_env_do_scf_inner_loop 118 6.6 0.004 0.010 135.051 135.054 init_scf_loop 11 6.9 0.000 0.000 123.390 123.395 prepare_preconditioner 11 7.9 0.000 0.000 118.594 118.617 make_preconditioner 11 8.9 0.000 0.000 118.594 118.617 make_full_inverse_cholesky 11 9.9 0.037 0.039 94.806 115.745 qs_scf_new_mos 118 7.6 0.001 0.001 91.297 91.371 qs_scf_loop_do_ot 118 8.6 0.001 0.001 91.296 91.370 ot_scf_mini 118 9.6 0.004 0.004 86.372 86.412 dbcsr_multiply_generic 2531 12.6 0.215 0.228 83.244 83.805 cp_fm_upper_to_full 105 14.8 52.306 75.071 52.306 75.071 multiply_cannon 2531 13.6 0.690 0.730 59.611 60.962 multiply_cannon_loop 2531 14.6 1.051 1.070 55.464 56.989 ot_mini 118 10.6 0.001 0.001 45.354 45.392 dbcsr_complete_redistribute 395 12.7 3.993 4.038 29.783 42.465 copy_fm_to_dbcsr 209 11.7 0.001 0.001 26.273 38.869 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.746 36.187 rebuild_ks_matrix 129 8.3 0.001 0.001 34.147 34.207 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 34.146 34.206 mp_alltoall_i22 718 14.1 21.518 33.969 21.518 33.969 cp_fm_cholesky_invert 11 10.9 33.383 33.389 33.383 33.389 mp_waitall_1 104660 16.8 28.421 32.780 28.421 32.780 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.724 31.767 qs_ot_get_p 129 10.4 0.001 0.001 26.005 26.077 qs_ot_get_derivative 118 11.6 0.002 0.002 25.124 25.169 qs_ot_p2m_diag 83 11.4 0.879 0.886 21.830 21.862 make_m2s 5062 13.6 0.078 0.080 18.969 20.535 multiply_cannon_metrocomm3 10124 15.6 0.025 0.027 18.939 20.405 ot_diis_step 118 11.6 0.022 0.023 20.195 20.195 make_images 5062 14.6 3.087 3.287 18.483 20.046 cp_dbcsr_syevd 83 12.4 0.006 0.006 20.028 20.029 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.502 19.651 apply_single 129 13.6 0.001 0.001 19.502 19.650 multiply_cannon_multrec 10124 15.6 10.415 12.334 18.442 18.514 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.831 16.832 cp_fm_diag_elpa_base 83 14.4 12.437 14.040 16.827 16.828 qs_rho_update_rho_low 129 7.7 0.001 0.001 16.617 16.640 calculate_rho_elec 129 8.7 0.485 0.486 16.616 16.640 sum_up_and_integrate 129 10.3 0.002 0.002 16.130 16.220 integrate_v_rspace 129 11.3 0.004 0.005 16.067 16.157 multiply_cannon_sync_h2d 10124 15.6 14.423 14.440 14.423 14.440 hybrid_alltoall_any 5248 16.5 1.322 3.094 10.942 13.143 make_images_data 5062 15.6 0.065 0.070 10.791 13.059 init_scf_run 11 5.9 0.000 0.001 12.221 12.222 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.221 12.222 pw_transfer 1559 11.6 0.095 0.095 11.416 11.426 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 11.177 11.186 fft_wrap_pw1pw2_140 527 13.2 3.069 3.148 9.969 9.979 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.789 9.820 dbcsr_mm_accdrv_process 20942 16.1 4.282 6.314 7.777 9.730 cp_fm_cholesky_decompose 22 10.9 8.931 8.963 8.931 8.963 wfi_extrapolate 11 7.9 0.001 0.001 8.914 8.914 grid_integrate_task_list 129 12.3 8.621 8.816 8.621 8.816 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.164 8.165 density_rs2pw 129 9.7 0.005 0.006 7.557 7.614 mp_alltoall_d11v 2423 14.1 6.978 7.143 6.978 7.143 calculate_dm_sparse 129 9.5 0.001 0.001 6.866 6.913 grid_collocate_task_list 129 9.7 6.463 6.539 6.463 6.539 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.439 6.495 fft3d_ps 1301 14.7 2.875 2.886 6.389 6.431 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.147 6.227 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=290.844000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2889.454545, yerr=163.463848 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 1.259942E+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 22542442. 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 0.012 0.033 85.141 85.141 qs_energies 1 2.0 0.000 0.000 84.677 84.693 ls_scf 1 3.0 0.000 0.000 83.771 83.788 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.767 72.931 multiply_cannon 111 7.7 0.017 0.020 56.054 57.243 multiply_cannon_loop 111 8.7 0.216 0.241 52.533 53.830 ls_scf_main 1 4.0 0.000 0.000 52.286 52.295 density_matrix_trs4 2 5.0 0.002 0.003 46.771 46.848 ls_scf_init_scf 1 4.0 0.000 0.001 28.444 28.448 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.320 27.374 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.163 25.187 mp_waitall_1 11031 10.9 22.391 25.147 22.391 25.147 multiply_cannon_multrec 2664 9.7 8.161 8.917 15.556 17.342 multiply_cannon_sync_h2d 2664 9.7 13.725 15.002 13.725 15.002 make_m2s 222 7.7 0.009 0.012 13.132 13.670 make_images 222 8.7 0.098 0.109 13.110 13.651 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.587 12.977 make_images_data 222 9.7 0.004 0.005 7.714 8.290 dbcsr_mm_accdrv_process 4760 10.4 0.587 0.732 7.011 8.014 hybrid_alltoall_any 227 10.6 0.216 1.834 6.605 7.753 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.423 7.577 dbcsr_mm_accdrv_process_sort 4760 11.4 6.223 7.117 6.223 7.117 calculate_norms 4752 9.8 5.541 6.231 5.541 6.231 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.017 5.177 mp_sum_l 887 5.1 3.037 4.285 3.037 4.285 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.070 3.692 mp_irecv_dv 6231 10.9 2.053 3.659 2.053 3.659 make_images_sizes 222 9.7 0.000 0.000 0.768 3.464 mp_alltoall_i44 222 10.7 0.767 3.464 0.767 3.464 arnoldi_extremal 4 6.8 0.000 0.000 3.234 3.264 arnoldi_normal_ev 4 7.8 0.001 0.003 3.234 3.264 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.255 3.234 build_subspace 16 8.4 0.009 0.012 3.131 3.133 ls_scf_post 1 4.0 0.000 0.000 3.041 3.058 ls_scf_store_result 1 5.0 0.000 0.000 2.852 2.898 dbcsr_special_finalize 555 9.7 0.005 0.006 2.298 2.790 dbcsr_merge_single_wm 555 10.7 0.486 0.631 2.289 2.781 make_images_pack 222 9.7 2.205 2.617 2.207 2.619 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.316 2.557 dbcsr_sort_data 658 11.4 2.063 2.479 2.063 2.479 dbcsr_matrix_vector_mult_local 304 10.0 2.065 2.457 2.067 2.459 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.300 2.393 buffer_matrices_ensure_size 222 8.7 1.834 2.221 1.834 2.221 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.765 1.769 rebuild_ks_matrix 3 7.3 0.000 0.000 1.756 1.759 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.006 1.756 1.759 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.141000, 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 2.192060E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.045 89.407 89.408 qs_energies 1 2.0 0.000 0.000 88.839 88.842 ls_scf 1 3.0 0.000 0.000 87.496 87.500 dbcsr_multiply_generic 111 6.7 0.015 0.019 73.502 73.867 multiply_cannon 111 7.7 0.029 0.037 52.626 56.499 ls_scf_main 1 4.0 0.000 0.000 53.985 53.991 multiply_cannon_loop 111 8.7 0.135 0.146 49.798 52.837 density_matrix_trs4 2 5.0 0.002 0.003 48.250 48.446 mp_waitall_1 9105 10.9 20.293 30.020 20.293 30.020 ls_scf_init_scf 1 4.0 0.001 0.003 29.843 29.845 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.636 28.740 multiply_cannon_multrec 1332 9.7 13.391 17.179 22.684 27.577 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.269 26.282 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.322 20.741 make_m2s 222 7.7 0.006 0.008 14.740 15.361 make_images 222 8.7 1.367 1.695 14.709 15.331 dbcsr_mm_accdrv_process 4041 10.4 0.330 0.538 8.887 10.480 dbcsr_mm_accdrv_process_sort 4041 11.4 8.405 9.942 8.405 9.942 make_images_data 222 9.7 0.004 0.005 8.415 9.292 hybrid_alltoall_any 227 10.6 0.542 2.548 7.799 9.273 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.171 7.750 mp_irecv_dv 3311 11.0 3.151 7.697 3.151 7.697 mp_sum_l 887 5.1 4.660 7.471 4.660 7.471 calculate_norms 2376 9.8 6.072 6.769 6.072 6.769 multiply_cannon_sync_h2d 1332 9.7 4.868 6.119 4.868 6.119 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.473 5.857 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.953 5.181 arnoldi_extremal 4 6.8 0.000 0.000 4.811 4.829 arnoldi_normal_ev 4 7.8 0.001 0.004 4.811 4.829 build_subspace 16 8.4 0.014 0.021 4.554 4.560 ls_scf_post 1 4.0 0.000 0.000 3.668 3.672 ls_scf_store_result 1 5.0 0.000 0.000 3.402 3.491 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.215 3.447 dbcsr_matrix_vector_mult_local 304 10.0 2.780 3.239 2.782 3.241 mp_allgather_i34 111 8.7 0.825 2.914 0.825 2.914 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.641 2.732 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.187 2.700 dbcsr_data_new 4174 10.1 2.109 2.403 2.109 2.403 make_images_pack 222 9.7 1.816 2.139 1.818 2.141 dbcsr_sort_data 436 11.2 1.815 2.072 1.815 2.072 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.870 1.872 rebuild_ks_matrix 3 7.3 0.000 0.000 1.857 1.859 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.857 1.859 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.408000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1775.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 2.850611E+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 0.033 0.052 90.957 90.958 qs_energies 1 2.0 0.000 0.000 90.282 90.285 ls_scf 1 3.0 0.000 0.001 88.877 88.881 dbcsr_multiply_generic 111 6.7 0.016 0.017 73.068 73.415 ls_scf_main 1 4.0 0.000 0.001 54.944 54.948 multiply_cannon 111 7.7 0.036 0.083 50.837 54.844 multiply_cannon_loop 111 8.7 0.116 0.127 48.094 51.978 density_matrix_trs4 2 5.0 0.002 0.005 49.089 49.293 mp_waitall_1 7281 11.0 21.452 31.187 21.452 31.187 ls_scf_init_scf 1 4.0 0.000 0.003 30.233 30.236 ls_scf_init_matrix_S 1 5.0 0.000 0.002 29.034 29.109 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.709 26.722 multiply_cannon_multrec 888 9.7 12.798 15.754 21.394 24.949 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 9.906 21.591 make_m2s 222 7.7 0.006 0.007 15.693 16.384 make_images 222 8.7 1.582 1.864 15.654 16.346 make_images_data 222 9.7 0.004 0.004 8.747 9.871 hybrid_alltoall_any 227 10.6 0.642 2.957 8.270 9.481 dbcsr_mm_accdrv_process 3754 10.4 0.319 0.498 8.113 9.374 dbcsr_mm_accdrv_process_sort 3754 11.4 7.676 8.876 7.676 8.876 mp_sum_l 887 5.1 4.856 7.891 4.856 7.891 multiply_cannon_sync_h2d 888 9.7 6.088 7.452 6.088 7.452 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.458 7.254 mp_irecv_dv 2335 11.1 2.443 7.190 2.443 7.190 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.437 6.356 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.613 6.029 arnoldi_extremal 4 6.8 0.000 0.000 5.659 5.681 arnoldi_normal_ev 4 7.8 0.001 0.005 5.659 5.681 build_subspace 16 8.4 0.014 0.020 5.344 5.351 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.737 5.006 calculate_norms 1584 9.8 4.350 4.743 4.350 4.743 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.694 4.043 ls_scf_post 1 4.0 0.000 0.000 3.699 3.703 dbcsr_matrix_vector_mult_local 304 10.0 3.065 3.625 3.068 3.627 ls_scf_store_result 1 5.0 0.000 0.000 3.431 3.513 mp_allgather_i34 111 8.7 0.850 3.450 0.850 3.450 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.838 2.951 dbcsr_data_new 4116 9.9 2.106 2.466 2.106 2.466 make_images_sizes 222 9.7 0.000 0.000 1.160 2.185 mp_alltoall_i44 222 10.7 1.160 2.184 1.160 2.184 dbcsr_sort_data 325 11.1 1.835 2.161 1.835 2.161 dbcsr_finalize 304 7.8 0.026 0.032 1.614 1.908 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.892 1.894 make_images_pack 222 9.7 1.621 1.882 1.624 1.885 rebuild_ks_matrix 3 7.3 0.000 0.000 1.874 1.876 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.874 1.876 dbcsr_merge_all 275 8.9 0.246 0.317 1.533 1.822 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=90.958000, 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/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 3.394404E+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 96.775 96.775 qs_energies 1 2.0 0.000 0.000 95.998 96.004 ls_scf 1 3.0 0.000 0.000 94.344 94.350 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.055 78.324 ls_scf_main 1 4.0 0.000 0.000 58.535 58.537 multiply_cannon 111 7.7 0.050 0.106 51.433 55.956 density_matrix_trs4 2 5.0 0.002 0.003 52.433 52.554 multiply_cannon_loop 111 8.7 0.152 0.170 46.330 48.603 ls_scf_init_scf 1 4.0 0.000 0.001 32.572 32.574 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.379 31.437 mp_waitall_1 6369 11.0 22.298 29.276 22.298 29.276 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.892 28.905 multiply_cannon_multrec 1332 9.7 14.186 16.972 22.101 24.914 make_m2s 222 7.7 0.007 0.008 21.000 22.443 make_images 222 8.7 3.140 3.598 20.950 22.396 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 8.949 16.853 make_images_data 222 9.7 0.004 0.004 11.715 13.330 hybrid_alltoall_any 227 10.6 0.800 3.772 11.095 12.928 dbcsr_mm_accdrv_process 3641 10.4 0.288 0.490 7.552 9.077 dbcsr_mm_accdrv_process_sort 3641 11.4 7.108 8.591 7.108 8.591 mp_sum_l 887 5.1 4.034 7.622 4.034 7.622 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.063 6.117 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.113 6.083 mp_irecv_dv 3229 10.9 2.037 6.031 2.037 6.031 multiply_cannon_sync_h2d 1332 9.7 5.453 5.943 5.453 5.943 arnoldi_extremal 4 6.8 0.000 0.000 5.298 5.310 arnoldi_normal_ev 4 7.8 0.001 0.005 5.297 5.310 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.469 5.233 build_subspace 16 8.4 0.014 0.021 4.957 4.965 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.537 4.701 mp_allgather_i34 111 8.7 2.224 4.663 2.224 4.663 calculate_norms 2376 9.8 4.195 4.531 4.195 4.531 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.655 3.952 dbcsr_matrix_vector_mult_local 304 10.0 3.254 3.724 3.256 3.726 dbcsr_sort_data 658 11.4 3.087 3.433 3.087 3.433 ls_scf_post 1 4.0 0.000 0.000 3.237 3.242 dbcsr_special_finalize 555 9.7 0.006 0.007 2.826 3.212 dbcsr_merge_single_wm 555 10.7 0.541 0.672 2.817 3.203 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.000 3.046 ls_scf_store_result 1 5.0 0.000 0.000 2.970 3.037 dbcsr_data_release 10477 10.7 1.586 2.431 1.586 2.431 dbcsr_finalize 304 7.8 0.049 0.061 1.811 1.982 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.775000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2712.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 4.752744E+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 0.054 92.591 92.592 qs_energies 1 2.0 0.000 0.000 91.797 91.803 ls_scf 1 3.0 0.000 0.000 89.862 89.867 dbcsr_multiply_generic 111 6.7 0.017 0.019 71.333 71.588 ls_scf_main 1 4.0 0.000 0.000 56.963 56.968 multiply_cannon 111 7.7 0.079 0.155 52.739 55.741 multiply_cannon_loop 111 8.7 0.088 0.095 50.140 51.749 density_matrix_trs4 2 5.0 0.002 0.003 50.055 50.138 ls_scf_init_scf 1 4.0 0.000 0.001 29.503 29.507 mp_waitall_1 5436 11.0 24.880 28.795 24.880 28.795 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.193 28.234 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.098 26.115 multiply_cannon_multrec 444 9.7 13.669 16.171 20.753 22.174 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.222 15.772 make_m2s 222 7.7 0.005 0.005 13.767 14.741 make_images 222 8.7 2.039 2.479 13.700 14.672 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 5.942 13.977 hybrid_alltoall_any 227 10.6 0.803 3.844 8.311 10.054 make_images_data 222 9.7 0.003 0.004 8.534 9.903 multiply_cannon_sync_h2d 444 9.7 6.734 8.216 6.734 8.216 dbcsr_mm_accdrv_process 3003 10.4 0.364 0.401 6.772 7.956 dbcsr_mm_accdrv_process_sort 3003 11.4 6.408 7.555 6.408 7.555 arnoldi_extremal 4 6.8 0.000 0.000 5.880 5.898 arnoldi_normal_ev 4 7.8 0.002 0.005 5.879 5.897 build_subspace 16 8.4 0.015 0.020 5.469 5.479 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.415 4.648 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.224 4.422 dbcsr_matrix_vector_mult_local 304 10.0 3.757 4.203 3.759 4.205 mp_sum_l 887 5.1 2.829 4.165 2.829 4.165 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.539 3.843 mp_allgather_i34 111 8.7 1.189 3.827 1.189 3.827 mp_irecv_dv 1241 11.2 1.525 3.802 1.525 3.802 calculate_norms 792 9.8 3.610 3.716 3.610 3.716 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.479 3.534 ls_scf_post 1 4.0 0.000 0.000 3.396 3.401 make_images_sizes 222 9.7 0.000 0.000 0.850 3.356 mp_alltoall_i44 222 10.7 0.849 3.355 0.849 3.355 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.000 3.289 ls_scf_store_result 1 5.0 0.000 0.000 3.183 3.221 dbcsr_finalize 304 7.8 0.062 0.077 2.197 2.306 dbcsr_data_new 4608 9.7 1.775 2.249 1.775 2.249 dbcsr_merge_all 275 8.9 0.479 0.528 2.055 2.144 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.074 2.076 rebuild_ks_matrix 3 7.3 0.000 0.000 2.041 2.043 qs_ks_build_kohn_sham_matrix 3 8.3 0.007 0.011 2.041 2.043 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 1.920 1.920 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.592000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3673.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 8.885199E+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 0.071 0.092 105.065 105.065 qs_energies 1 2.0 0.000 0.000 103.651 103.656 ls_scf 1 3.0 0.000 0.000 100.690 100.694 dbcsr_multiply_generic 111 6.7 0.024 0.027 74.348 74.493 ls_scf_main 1 4.0 0.000 0.000 63.320 63.321 density_matrix_trs4 2 5.0 0.002 0.003 54.355 54.417 multiply_cannon 111 7.7 0.158 0.263 48.432 50.446 multiply_cannon_loop 111 8.7 0.098 0.100 45.430 45.972 ls_scf_init_scf 1 4.0 0.001 0.001 33.581 33.581 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.083 32.103 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.382 29.391 mp_waitall_1 4527 11.1 21.476 25.126 21.476 25.126 make_m2s 222 7.7 0.005 0.005 22.316 23.402 make_images 222 8.7 3.576 3.880 22.209 23.294 multiply_cannon_multrec 444 9.7 17.849 18.486 22.489 23.018 hybrid_alltoall_any 227 10.6 1.658 3.620 12.517 15.265 make_images_data 222 9.7 0.003 0.004 12.722 14.855 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.176 10.426 multiply_cannon_sync_h2d 444 9.7 8.795 8.846 8.795 8.846 arnoldi_extremal 4 6.8 0.000 0.000 7.527 7.540 arnoldi_normal_ev 4 7.8 0.003 0.009 7.526 7.540 build_subspace 16 8.4 0.026 0.036 6.962 6.973 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.608 5.753 dbcsr_matrix_vector_mult_local 304 10.0 5.162 5.455 5.164 5.458 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.081 5.172 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.886 5.152 dbcsr_mm_accdrv_process 1814 10.4 0.265 0.345 4.452 4.585 dbcsr_mm_accdrv_process_sort 1814 11.4 4.117 4.254 4.117 4.254 ls_scf_post 1 4.0 0.000 0.000 3.789 3.794 make_images_sizes 222 9.7 0.000 0.000 1.466 3.609 mp_alltoall_i44 222 10.7 1.465 3.609 1.465 3.609 ls_scf_store_result 1 5.0 0.000 0.000 3.510 3.544 mp_allgather_i34 111 8.7 1.122 3.479 1.122 3.479 calculate_norms 792 9.8 3.227 3.266 3.227 3.266 dbcsr_finalize 304 7.8 0.082 0.089 3.075 3.175 dbcsr_merge_all 275 8.9 0.885 0.916 2.860 2.955 qs_energies_init_hamiltonians 1 3.0 0.008 0.017 2.930 2.930 dbcsr_complete_redistribute 5 7.6 1.427 1.459 2.734 2.854 matrix_ls_to_qs 2 6.0 0.000 0.000 2.406 2.533 dbcsr_sort_data 325 11.1 2.443 2.508 2.443 2.508 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.381 2.383 rebuild_ks_matrix 3 7.3 0.000 0.000 2.316 2.317 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.002 2.316 2.317 dbcsr_new_transposed 4 7.5 0.241 0.249 2.296 2.312 dbcsr_data_new 6591 9.6 1.836 2.294 1.836 2.294 dbcsr_frobenius_norm 74 6.6 2.058 2.127 2.183 2.210 dbcsr_add_d 103 6.2 0.000 0.000 2.124 2.203 dbcsr_add_anytype 103 7.2 0.858 0.890 2.124 2.202 dbcsr_data_release 12724 10.6 1.975 2.169 1.975 2.169 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.065000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6985.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a_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 593.399808E+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 557854006. 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 0.035 0.143 230.811 230.813 qs_mol_dyn_low 1 2.0 0.005 0.038 229.589 229.664 qs_forces 5 3.8 0.005 0.024 229.420 229.424 qs_energies 5 4.8 0.003 0.030 226.393 226.431 scf_env_do_scf 5 5.8 0.001 0.026 210.277 210.280 scf_env_do_scf_inner_loop 105 6.6 0.008 0.033 183.224 183.234 qs_scf_new_mos 105 7.6 0.000 0.001 142.687 142.900 qs_scf_loop_do_ot 105 8.6 0.001 0.001 142.687 142.900 dbcsr_multiply_generic 1445 12.2 0.127 0.135 136.513 136.952 ot_scf_mini 105 9.6 0.003 0.004 132.585 132.718 multiply_cannon 1445 13.2 0.276 0.290 116.953 119.233 multiply_cannon_loop 1445 14.2 2.733 2.968 115.237 116.459 velocity_verlet 4 3.0 0.004 0.020 110.088 110.089 ot_mini 105 10.6 0.001 0.002 61.136 61.252 mp_waitall_1 488190 16.1 35.952 45.136 35.952 45.136 multiply_cannon_multrec 69360 15.2 30.005 35.126 39.720 44.776 qs_ot_get_p 112 10.4 0.001 0.001 41.224 41.568 qs_ot_get_derivative 55 11.6 0.001 0.001 39.248 39.386 multiply_cannon_metrocomm3 69360 15.2 0.205 0.215 26.637 36.123 multiply_cannon_sync_h2d 69360 15.2 29.516 33.566 29.516 33.566 qs_ot_p2m_diag 40 11.0 0.020 0.031 29.582 29.702 rebuild_ks_matrix 110 8.4 0.000 0.000 29.289 29.466 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.016 29.289 29.466 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.963 27.116 init_scf_loop 7 6.6 0.003 0.080 27.018 27.020 cp_dbcsr_syevd 40 12.0 0.002 0.002 26.370 26.371 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.158 23.405 apply_single 62 13.6 0.000 0.000 23.158 23.405 prepare_preconditioner 7 7.6 0.000 0.000 22.045 22.080 make_preconditioner 7 8.6 0.000 0.002 22.045 22.080 cp_fm_syevd 40 13.0 0.000 0.000 21.313 21.489 ot_new_cg_direction 55 11.6 0.001 0.002 21.112 21.112 qs_rho_update_rho_low 110 7.6 0.001 0.001 16.977 17.441 calculate_rho_elec 110 8.6 0.029 0.032 16.976 17.441 cp_fm_redistribute_end 40 14.0 8.337 16.630 8.341 16.632 cp_fm_syevd_base 40 14.0 8.285 16.583 8.285 16.583 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.949 15.208 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.815 14.881 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.653 14.754 mp_sum_l 4764 12.2 12.673 13.540 12.673 13.540 init_scf_run 5 5.8 0.000 0.001 13.431 13.432 scf_env_initial_rho_setup 5 6.8 0.000 0.002 13.431 13.432 pw_transfer 1645 12.4 0.079 0.100 12.196 12.436 fft_wrap_pw1pw2 1425 13.5 0.012 0.016 12.058 12.303 calculate_dm_sparse 110 9.5 0.000 0.001 11.702 11.922 density_rs2pw 110 9.6 0.005 0.007 11.077 11.499 fft_wrap_pw1pw2_240 915 15.0 1.134 1.238 10.662 10.895 qs_vxc_create 110 10.4 0.002 0.004 10.757 10.805 dbcsr_mm_accdrv_process 154766 15.8 5.791 6.300 9.581 10.757 cp_fm_cholesky_invert 7 10.6 10.373 10.381 10.373 10.381 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.037 10.097 check_diag 80 13.5 8.598 8.907 9.404 9.555 acc_transpose_blocks 69360 15.2 0.325 0.362 8.619 9.219 fft3d_pb 915 16.0 2.381 2.661 8.888 9.135 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 8.950 8.968 calculate_first_density_matrix 1 7.0 0.000 0.003 8.561 8.584 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 8.484 8.544 sum_up_and_integrate 60 10.3 0.001 0.003 8.447 8.459 integrate_v_rspace 60 11.3 0.002 0.003 8.430 8.442 multiply_cannon_metrocomm1 69360 15.2 0.094 0.102 4.877 8.428 transfer_rs2pw 445 10.6 0.007 0.008 7.663 8.125 xc_rho_set_and_dset_create 110 12.4 0.076 0.099 7.507 7.768 xc_vxc_pw_create 60 11.3 0.039 0.049 7.222 7.269 make_m2s 2890 13.2 0.078 0.087 6.443 7.023 make_full_single_inverse 7 9.6 0.001 0.004 6.953 6.986 make_images 2890 14.2 0.239 0.260 6.336 6.917 xc_pw_derive 510 13.4 0.005 0.007 6.489 6.576 acc_transpose_blocks_kernels 69360 16.2 0.805 0.882 6.052 6.403 mp_alltoall_z22v 2340 17.7 5.943 6.203 5.943 6.203 jit_kernel_transpose 5 15.0 5.247 5.537 5.247 5.537 mp_waitany 7680 13.5 4.419 4.965 4.419 4.965 multiply_cannon_metrocomm4 67915 15.2 0.186 0.204 2.024 4.880 potential_pw2rs 60 12.3 0.003 0.003 4.656 4.678 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=230.813000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.800000, yerr=2.993326 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 6a2264c2e1f43c021cf3d2fd277b1d3a087d8d1a Summary: empty Status: OK