=== 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: ee725ffc68fafd11fbf186e4af07b2db94632368 ################# 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 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SPGLIB 2.3.1, LIBGRPP 20231215, # SPFFT 1.0.6, SPLA 1.5.5 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 28.03.2024 # # \ 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; \ rm -rf build; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-hdf5 --with-libvdwxc --with-pexsi --with-plumed --with-sirius=no --with-spfft --with-spla; \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 #USE_DEEPMD := 2.2.7 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 #USE_SIRIUS := 7.5.2 USE_SPFFT := 1.0.6 USE_SPGLIB := 2.3.1 USE_SPLA := 1.5.5 # Only needed for SIRIUS #LIBVDWXC_VER := 0.4.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_DEEPMD),) USE_DEEPMD := $(strip $(USE_DEEPMD)) DEEPMD_INC := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/include DEEPMD_LIB := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/lib CFLAGS += -I$(DEEPMD_INC) DFLAGS += -D__DEEPMD LIBS += $(DEEPMD_LIB)/libdeepmd.so LIBS += $(DEEPMD_LIB)/libdeepmd_c.so LIBS += $(DEEPMD_LIB)/libdeepmd_cc.so LIBS += $(DEEPMD_LIB)/libdeepmd_dyn_cudart.so LIBS += $(DEEPMD_LIB)/libdeepmd_op.so LIBS += $(DEEPMD_LIB)/libdeepmd_op_cuda.so LIBS += $(DEEPMD_LIB)/libtensorflow_cc.so.2 LIBS += $(DEEPMD_LIB)/libtensorflow_framework.so.2 endif ifneq ($(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_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(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 ifeq ($(USE_ACC), yes) SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_SPFFT),) USE_SPFFT := $(strip $(USE_SPFFT)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/include ifeq ($(USE_ACC), yes) SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/lib endif CFLAGS += -I$(SPFFT_INC) DFLAGS += -D__SPFFT LIBS += $(SPFFT_LIB)/libspfft.a endif ifneq ($(USE_SPLA),) USE_SPLA := $(strip $(USE_SPLA)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPLA_LIB := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/lib/cuda else SPLA_LIB := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/lib endif CFLAGS += -I$(SPLA_INC) DFLAGS += -D__SPLA LIBS += $(SPLA_LIB)/libspla.a endif ifneq ($(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_f90cstub.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/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/01 job id: 53419078 --- 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/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/02 job id: 53419079 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/03 job id: 53419080 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/04 job id: 53419081 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/05 job id: 53419082 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/06 job id: 53419083 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/07 job id: 53419084 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/08 job id: 53419085 --- 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/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/09 job id: 53419086 --- 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/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/10 job id: 53419087 --- 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/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/11 job id: 53419089 --- 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/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/12 job id: 53419090 --- 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/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/13 job id: 53419091 --- 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/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/14 job id: 53419092 --- 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/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/15 job id: 53419093 --- 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/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/16 job id: 53419094 --- 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/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/17 job id: 53419095 --- 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/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/18 job id: 53419096 --- 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/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/19 job id: 53419097 --- 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/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/20 job id: 53419098 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/21 job id: 53419099 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/22 job id: 53419100 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/23 job id: 53419101 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/24 job id: 53419102 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/25 job id: 53419103 --- 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] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/26 job id: 53419104 --- 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] = 20 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/27 job id: 53419105 --- 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/ee725ffc68fafd11fbf186e4af07b2db94632368_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 172669. 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 0.029 0.041 128.170 128.171 farming_run 1 2.0 127.668 127.669 128.110 128.142 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.546519E+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 822089. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1131389. 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 0.023 110.075 110.075 qs_energies 1 2.0 0.001 0.001 109.889 109.891 mp2_main 1 3.0 0.000 0.000 107.990 107.992 mp2_gpw_main 1 4.0 0.032 0.064 107.026 107.028 mp2_ri_gpw_compute_in 1 5.0 0.174 0.176 88.398 88.480 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 57.108 57.190 mp2_eri_3c_integrate_gpw 272 7.0 0.150 0.159 43.540 48.318 get_2c_integrals 1 6.0 0.008 0.009 30.796 31.115 integrate_v_rspace 273 8.0 0.441 0.458 25.496 29.820 fft_wrap_pw1pw2 5465 10.4 0.068 0.070 27.994 28.614 fft_wrap_pw1pw2_100 2178 11.4 1.325 1.437 25.438 26.096 grid_integrate_task_list 273 9.0 20.746 25.986 20.746 25.986 compute_2c_integrals 1 7.0 0.002 0.002 21.451 21.453 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 20.786 21.147 mp2_eri_2c_integrate_gpw 1 9.0 2.391 2.432 20.784 21.144 rpa_ri_compute_en 1 5.0 0.019 0.021 18.503 18.584 fft3d_s 5443 12.4 16.764 17.126 16.785 17.148 ao_to_mo_and_store_B_mult_1 272 7.0 10.779 15.307 10.779 15.307 calculate_wavefunction 272 8.0 5.275 5.356 12.732 13.657 calc_potential_gpw 544 9.5 0.004 0.005 10.847 11.620 rpa_num_int 1 6.0 0.000 0.000 10.579 10.579 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.517 10.579 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 9.774 10.516 potential_pw2rs 545 10.0 0.106 0.108 8.442 10.359 cp_fm_cholesky_decompose 12 8.2 9.538 9.875 9.538 9.875 calc_mat_Q 8 8.0 0.000 0.000 9.470 9.598 contract_S_to_Q 8 9.0 0.000 0.000 8.892 9.019 cholesky_decomp 1 7.0 0.000 0.000 8.535 8.868 parallel_gemm_fm 14 9.1 0.000 0.000 8.516 8.664 parallel_gemm_fm_cosma 14 10.1 8.516 8.664 8.516 8.664 collocate_single_gaussian 272 10.0 0.040 0.041 7.810 8.605 create_integ_mat 1 6.0 0.007 0.009 7.752 7.752 array2fm 1 7.0 0.000 0.000 6.662 7.165 pw_gather_s 2722 12.2 3.940 4.930 3.940 4.930 pw_poisson_solve 545 10.5 0.010 0.010 4.645 4.872 pw_scatter_s 2720 12.7 4.479 4.689 4.479 4.689 pw_poisson_set 548 11.5 0.018 0.019 3.338 3.548 array2fm_buffer_send 1 8.0 2.959 3.164 2.959 3.164 pw_copy 4911 11.6 2.611 2.723 2.611 2.723 ao_to_mo_and_store_B_E_Ex_1 272 7.0 1.082 1.232 2.025 2.216 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=107.026493, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2806.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 200775. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.036 425.623 425.624 farming_run 1 2.0 424.829 424.838 425.565 425.580 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.281446E+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 713945. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 480173. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 29913300. MP_SendRecv 2876 2171443. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 223.354 223.355 qs_energies 1 2.0 0.000 0.000 223.135 223.143 scf_env_do_scf 1 3.0 0.000 0.000 115.110 115.110 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.238 114.245 rebuild_ks_matrix 4 6.0 0.000 0.000 114.237 114.244 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 114.237 114.244 hfx_ks_matrix 4 8.0 0.001 0.001 113.891 113.894 integrate_four_center 4 9.0 0.153 0.464 113.890 113.893 mp2_main 1 3.0 0.000 0.000 107.738 107.745 mp2_gpw_main 1 4.0 0.048 0.072 106.902 106.917 integrate_four_center_main 4 10.0 0.132 0.615 101.957 105.697 integrate_four_center_bin 270 11.0 101.825 105.650 101.825 105.650 init_scf_loop 1 4.0 0.000 0.000 96.544 96.545 mp2_ri_gpw_compute_in 1 5.0 0.065 0.068 80.045 81.169 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 56.753 57.877 mp2_eri_3c_integrate_gpw 91 7.0 0.148 0.161 44.312 49.475 integrate_v_rspace 95 8.0 0.398 0.580 27.990 33.018 fft_wrap_pw1pw2 1868 10.4 0.030 0.036 30.537 31.071 fft_wrap_pw1pw2_100 730 11.4 0.651 0.808 28.246 28.802 grid_integrate_task_list 95 9.0 23.288 28.545 23.288 28.545 mp2_ri_gpw_compute_en 1 5.0 0.055 0.065 26.706 28.480 ao_to_mo_and_store_B_mult_1 91 7.0 10.761 27.738 10.761 27.738 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.842 1.887 24.928 24.938 get_2c_integrals 1 6.0 0.000 0.000 23.208 23.227 compute_2c_integrals 1 7.0 0.002 0.003 22.375 22.378 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 21.781 22.245 mp2_eri_2c_integrate_gpw 1 9.0 1.738 1.898 21.780 22.243 fft3d_s 1823 12.4 19.212 20.117 19.226 20.130 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 18.564 18.564 calc_potential_gpw 182 9.5 0.002 0.002 12.321 12.962 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 10.722 11.528 calculate_wavefunction 91 8.0 1.980 2.013 10.001 10.757 potential_pw2rs 186 10.0 0.033 0.035 9.092 10.401 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.582 8.806 9.303 collocate_single_gaussian 91 10.0 0.018 0.024 8.327 8.905 local_gemm 172 8.0 8.250 8.752 8.250 8.752 mp2_ri_gpw_compute_en_comm 22 7.0 0.503 0.529 7.900 8.594 mp_sync 37 10.5 3.897 7.077 3.897 7.077 integrate_four_center_load 4 10.0 0.000 0.000 6.770 6.776 hfx_load_balance 1 11.0 0.000 0.000 6.770 6.776 mp_sendrecv_dm3 2068 8.0 5.929 6.604 5.929 6.604 mp2_ri_gpw_compute_en_ener 172 7.0 6.335 6.427 6.335 6.427 pw_poisson_solve 186 10.4 0.005 0.005 5.540 6.024 pw_gather_s 912 12.2 4.850 5.583 4.850 5.583 pw_scatter_s 910 12.7 4.091 4.638 4.091 4.638 pw_poisson_set 189 11.4 0.009 0.010 4.094 4.595 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=106.893218, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1534.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 446.464000E+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 3690 72628. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 226540. 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 0.012 0.029 56.670 56.671 qs_mol_dyn_low 1 2.0 0.006 0.017 56.347 56.357 qs_forces 11 3.9 0.002 0.006 56.243 56.246 qs_energies 11 4.9 0.002 0.003 54.728 54.747 scf_env_do_scf 11 5.9 0.000 0.001 49.073 49.073 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 47.108 47.109 qs_scf_new_mos 108 7.5 0.000 0.000 36.849 37.252 qs_scf_loop_do_ot 108 8.5 0.000 0.000 36.848 37.252 dbcsr_multiply_generic 2286 12.5 0.096 0.100 36.276 36.806 ot_scf_mini 108 9.5 0.002 0.002 35.014 35.217 multiply_cannon 2286 13.5 0.178 0.187 27.475 29.319 multiply_cannon_loop 2286 14.5 1.827 1.927 26.775 28.669 velocity_verlet 10 3.0 0.006 0.023 27.882 27.884 ot_mini 108 10.5 0.001 0.001 21.546 21.805 qs_ot_get_derivative 108 11.5 0.001 0.001 18.393 18.629 mp_waitall_1 245248 16.5 10.062 16.279 10.062 16.279 multiply_cannon_metrocomm3 54864 15.5 0.071 0.076 7.009 14.017 multiply_cannon_multrec 54864 15.5 3.538 5.478 7.490 10.777 qs_ot_get_p 119 10.4 0.010 0.025 8.511 8.825 mp_sum_l 7287 12.8 6.725 8.731 6.725 8.731 rebuild_ks_matrix 119 8.3 0.000 0.000 8.010 8.157 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.009 8.157 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.061 7.194 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.526 6.974 multiply_cannon_sync_h2d 54864 15.5 4.980 6.371 4.980 6.371 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.001 6.138 dbcsr_mm_accdrv_process 76910 16.1 1.836 2.888 3.868 5.559 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.396 5.476 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.517 4.517 sum_up_and_integrate 119 10.3 0.001 0.002 4.463 4.470 integrate_v_rspace 119 11.3 0.002 0.003 4.452 4.460 init_scf_run 11 5.9 0.000 0.001 4.426 4.426 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.426 4.426 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.182 4.322 calculate_rho_elec 119 8.7 0.011 0.016 4.181 4.321 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.316 4.316 cp_fm_redistribute_end 50 14.0 2.202 4.292 2.209 4.295 cp_fm_diag_elpa_base 50 14.0 2.080 4.185 2.084 4.193 multiply_cannon_metrocomm1 54864 15.5 0.056 0.061 2.337 3.833 calculate_dm_sparse 119 9.5 0.000 0.000 3.059 3.280 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.967 3.182 apply_single 119 13.6 0.000 0.000 2.967 3.182 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.752 2.818 ot_diis_step 108 11.5 0.006 0.006 2.816 2.817 acc_transpose_blocks 54864 15.5 0.222 0.241 2.212 2.764 jit_kernel_multiply 13 15.8 1.969 2.577 1.969 2.577 calculate_first_density_matrix 1 7.0 0.000 0.001 2.479 2.485 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.360 2.376 density_rs2pw 119 9.7 0.003 0.004 2.115 2.197 mp_sum_d 4137 12.0 1.542 2.185 1.542 2.185 grid_integrate_task_list 119 12.3 2.021 2.099 2.021 2.099 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.038 2.091 init_scf_loop 11 6.9 0.000 0.000 1.949 1.949 wfi_extrapolate 11 7.9 0.001 0.001 1.877 1.877 potential_pw2rs 119 12.3 0.004 0.004 1.843 1.852 make_m2s 4572 13.5 0.054 0.056 1.766 1.812 make_images 4572 14.5 0.133 0.139 1.683 1.729 fft_wrap_pw1pw2 1201 11.6 0.011 0.013 1.576 1.651 mp_alltoall_d11v 2130 13.8 1.416 1.537 1.416 1.537 transfer_rs2pw 487 10.6 0.006 0.007 1.429 1.530 mp_waitany 12084 13.8 1.299 1.484 1.299 1.484 grid_collocate_task_list 119 9.7 1.351 1.440 1.351 1.440 acc_transpose_blocks_sync 164592 16.5 1.198 1.434 1.198 1.434 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.394 1.420 fft3d_ps 1201 13.6 0.359 0.457 1.351 1.418 transfer_pw2rs 487 13.2 0.005 0.006 1.368 1.376 fft_wrap_pw1pw2_140 487 12.2 0.049 0.054 1.232 1.307 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.843 1.256 mp_alltoall_i22 627 13.8 0.866 1.202 0.866 1.202 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=56.671000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=425.545455, yerr=1.157084 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 486.383616E+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 3679 72937. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1279738. 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 0.068 0.093 39.731 39.732 qs_mol_dyn_low 1 2.0 0.017 0.037 39.278 39.287 qs_forces 11 3.9 0.026 0.043 38.977 38.992 qs_energies 11 4.9 0.002 0.002 37.281 37.323 scf_env_do_scf 11 5.9 0.000 0.001 31.933 31.934 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.646 29.647 dbcsr_multiply_generic 2286 12.5 0.104 0.108 22.210 22.555 qs_scf_new_mos 108 7.5 0.001 0.001 20.897 21.127 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.896 21.127 ot_scf_mini 108 9.5 0.002 0.003 19.980 20.145 velocity_verlet 10 3.0 0.001 0.002 18.574 18.580 multiply_cannon 2286 13.5 0.207 0.214 16.927 18.323 multiply_cannon_loop 2286 14.5 1.187 1.240 15.748 17.275 ot_mini 108 10.5 0.001 0.001 12.282 12.515 mp_waitall_1 200699 16.5 5.604 10.826 5.604 10.826 qs_ot_get_derivative 108 11.5 0.001 0.001 9.919 10.086 multiply_cannon_metrocomm3 27432 15.5 0.071 0.073 4.027 9.493 multiply_cannon_multrec 27432 15.5 1.820 4.128 6.179 8.916 rebuild_ks_matrix 119 8.3 0.000 0.000 6.856 6.999 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.856 6.999 dbcsr_mm_accdrv_process 47894 16.0 3.456 5.827 4.280 6.376 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.036 6.167 qs_ot_get_p 119 10.4 0.001 0.001 4.773 4.990 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.697 4.608 mp_sum_l 7287 12.8 2.168 4.185 2.168 4.185 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.057 4.069 apply_single 119 13.6 0.000 0.000 3.056 4.069 init_scf_run 11 5.9 0.000 0.001 4.002 4.003 scf_env_initial_rho_setup 11 6.9 0.026 0.035 4.002 4.002 sum_up_and_integrate 119 10.3 0.001 0.001 3.714 3.721 integrate_v_rspace 119 11.3 0.002 0.003 3.701 3.707 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.673 3.706 calculate_rho_elec 119 8.7 0.021 0.024 3.672 3.705 qs_ot_p2m_diag 50 11.0 0.008 0.012 3.078 3.094 make_m2s 4572 13.5 0.053 0.054 2.699 3.025 make_images 4572 14.5 0.205 0.241 2.608 2.937 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.678 2.678 calculate_first_density_matrix 1 7.0 0.000 0.000 2.458 2.489 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.372 2.461 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.337 2.338 ot_diis_step 108 11.5 0.011 0.011 2.314 2.315 cp_fm_redistribute_end 50 14.0 1.181 2.303 1.187 2.306 init_scf_loop 11 6.9 0.000 0.000 2.265 2.266 cp_fm_diag_elpa_base 50 14.0 1.086 2.203 1.115 2.242 multiply_cannon_sync_h2d 27432 15.5 1.699 2.226 1.699 2.226 calculate_dm_sparse 119 9.5 0.000 0.001 2.068 2.149 density_rs2pw 119 9.7 0.003 0.004 1.968 2.066 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.029 2.031 acc_transpose_blocks 27432 15.5 0.115 0.121 1.629 1.985 grid_integrate_task_list 119 12.3 1.836 1.956 1.836 1.956 jit_kernel_multiply 10 16.3 0.765 1.925 0.765 1.925 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.741 1.782 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.681 1.713 make_images_data 4572 15.5 0.047 0.053 1.255 1.593 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.123 1.491 wfi_extrapolate 11 7.9 0.001 0.001 1.457 1.457 potential_pw2rs 119 12.3 0.006 0.006 1.451 1.457 fft3d_ps 1201 13.6 0.508 0.559 1.401 1.431 mp_alltoall_d11v 2130 13.8 1.275 1.395 1.275 1.395 prepare_preconditioner 11 7.9 0.000 0.000 1.363 1.389 make_preconditioner 11 8.9 0.000 0.000 1.363 1.389 transfer_rs2pw 487 10.6 0.005 0.006 1.220 1.361 grid_collocate_task_list 119 9.7 1.290 1.355 1.290 1.355 fft_wrap_pw1pw2_140 487 12.2 0.047 0.051 1.316 1.349 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.260 1.305 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.259 1.274 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.208 1.254 qs_energies_init_hamiltonians 11 5.9 0.039 0.073 1.038 1.078 mp_sum_d 4137 12.0 0.576 1.032 0.576 1.032 mp_allgather_i34 2286 14.5 0.603 1.018 0.603 1.018 transfer_pw2rs 487 13.2 0.004 0.005 0.968 0.974 acc_transpose_blocks_sync 82296 16.5 0.814 0.936 0.814 0.936 acc_transpose_blocks_kernels 27432 16.5 0.190 0.278 0.675 0.909 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.875 0.889 make_images_sizes 4572 15.5 0.005 0.005 0.579 0.833 mp_alltoall_i44 4572 16.5 0.574 0.829 0.574 0.829 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.732000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=461.909091, yerr=1.781447 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 514.752512E+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 3629 73909. 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 0.033 0.049 40.403 40.404 qs_mol_dyn_low 1 2.0 0.005 0.005 40.136 40.145 qs_forces 11 3.9 0.003 0.003 40.057 40.057 qs_energies 11 4.9 0.002 0.002 38.286 38.293 scf_env_do_scf 11 5.9 0.000 0.001 32.889 32.889 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.922 29.923 dbcsr_multiply_generic 2286 12.5 0.100 0.102 20.354 20.936 qs_scf_new_mos 108 7.5 0.001 0.001 19.923 19.943 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.923 19.942 velocity_verlet 10 3.0 0.008 0.024 19.459 19.460 ot_scf_mini 108 9.5 0.002 0.002 18.986 19.009 multiply_cannon 2286 13.5 0.195 0.200 15.116 16.871 multiply_cannon_loop 2286 14.5 0.862 0.916 13.889 16.012 ot_mini 108 10.5 0.001 0.001 11.375 11.408 qs_ot_get_derivative 108 11.5 0.001 0.001 9.551 9.575 rebuild_ks_matrix 119 8.3 0.000 0.000 7.552 7.687 qs_ks_build_kohn_sham_matrix 119 9.3 0.018 0.047 7.552 7.686 mp_waitall_1 158411 16.6 4.672 7.470 4.672 7.470 multiply_cannon_multrec 18288 15.5 1.851 2.873 7.126 7.402 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.650 6.776 dbcsr_mm_accdrv_process 38222 16.0 5.105 6.055 5.184 6.121 qs_ot_get_p 119 10.4 0.012 0.023 4.934 4.973 sum_up_and_integrate 119 10.3 0.001 0.001 4.876 4.887 integrate_v_rspace 119 11.3 0.003 0.003 4.863 4.874 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.492 4.517 calculate_rho_elec 119 8.7 0.031 0.031 4.491 4.517 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 2.028 4.207 init_scf_run 11 5.9 0.000 0.001 4.153 4.153 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.153 4.153 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.500 4.123 qs_ot_p2m_diag 50 11.0 0.012 0.012 3.139 3.160 density_rs2pw 119 9.7 0.003 0.003 2.965 3.055 make_m2s 4572 13.5 0.044 0.045 2.625 3.021 init_scf_loop 11 6.9 0.000 0.000 2.945 2.947 make_images 4572 14.5 0.194 0.207 2.539 2.935 mp_sum_l 7287 12.8 2.218 2.848 2.218 2.848 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.756 2.757 potential_pw2rs 119 12.3 0.007 0.007 2.579 2.626 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.163 2.600 apply_single 119 13.6 0.000 0.000 2.163 2.600 calculate_first_density_matrix 1 7.0 0.000 0.000 2.555 2.566 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.386 2.399 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.365 2.374 cp_fm_diag_elpa_base 50 14.0 2.339 2.354 2.364 2.372 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 2.231 2.280 calculate_dm_sparse 119 9.5 0.000 0.000 2.123 2.135 transfer_rs2pw 487 10.6 0.005 0.005 2.017 2.080 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.056 2.058 fft3d_ps 1201 13.6 0.535 0.557 1.917 1.971 prepare_preconditioner 11 7.9 0.000 0.000 1.943 1.956 make_preconditioner 11 8.9 0.000 0.000 1.943 1.956 grid_integrate_task_list 119 12.3 1.782 1.889 1.782 1.889 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.802 1.859 transfer_pw2rs 487 13.2 0.004 0.004 1.779 1.825 acc_transpose_blocks 18288 15.5 0.080 0.083 1.727 1.793 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.729 1.768 ot_diis_step 108 11.5 0.011 0.012 1.714 1.714 make_images_data 4572 15.5 0.047 0.051 1.106 1.706 hybrid_alltoall_any 4725 16.4 0.058 0.117 0.914 1.575 fft_wrap_pw1pw2_140 487 12.2 0.061 0.063 1.513 1.553 wfi_extrapolate 11 7.9 0.001 0.001 1.486 1.486 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.310 1.350 grid_collocate_task_list 119 9.7 1.233 1.326 1.233 1.326 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.293 1.316 mp_alltoall_z22v 1201 15.6 1.185 1.306 1.185 1.306 mp_sendrecv_dv 11067 12.7 1.218 1.267 1.218 1.267 mp_alltoall_d11v 2130 13.8 1.046 1.225 1.046 1.225 mp_allgather_i34 2286 14.5 0.672 1.207 0.672 1.207 multiply_cannon_sync_h2d 18288 15.5 0.975 1.106 0.975 1.106 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.636 1.053 make_images_sizes 4572 15.5 0.005 0.005 0.764 0.986 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.972 0.981 mp_alltoall_i44 4572 16.5 0.759 0.981 0.759 0.981 qs_energies_init_hamiltonians 11 5.9 0.004 0.005 0.969 0.969 acc_transpose_blocks_kernels 18288 16.5 0.219 0.227 0.891 0.941 cp_fm_cholesky_invert 11 10.9 0.911 0.924 0.911 0.924 transfer_rs2pw_50 119 11.7 0.156 0.159 0.801 0.913 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=40.404000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=490.272727, yerr=1.420045 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 554.672128E+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 3679 72932. MP_Allreduce 10304 342. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 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 0.020 0.032 36.223 36.224 qs_mol_dyn_low 1 2.0 0.023 0.056 35.991 36.000 qs_forces 11 3.9 0.002 0.003 35.909 35.929 qs_energies 11 4.9 0.002 0.002 34.182 34.207 scf_env_do_scf 11 5.9 0.000 0.001 28.980 28.980 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.124 26.125 dbcsr_multiply_generic 2286 12.5 0.114 0.122 20.143 20.280 qs_scf_new_mos 108 7.5 0.001 0.001 18.064 18.115 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.063 18.115 velocity_verlet 10 3.0 0.002 0.002 18.080 18.082 ot_scf_mini 108 9.5 0.002 0.003 17.027 17.072 multiply_cannon 2286 13.5 0.216 0.224 16.332 16.819 multiply_cannon_loop 2286 14.5 1.546 1.623 15.366 15.760 ot_mini 108 10.5 0.001 0.001 10.527 10.584 multiply_cannon_multrec 27432 15.5 2.485 3.176 9.102 9.462 qs_ot_get_derivative 108 11.5 0.001 0.001 8.659 8.704 dbcsr_mm_accdrv_process 47916 15.9 5.994 7.810 6.516 7.874 rebuild_ks_matrix 119 8.3 0.000 0.000 6.398 6.458 qs_ks_build_kohn_sham_matrix 119 9.3 0.015 0.018 6.398 6.458 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.659 5.713 init_scf_run 11 5.9 0.000 0.001 3.825 3.825 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.825 3.825 qs_ot_get_p 119 10.4 0.001 0.001 3.626 3.695 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.232 3.626 sum_up_and_integrate 119 10.3 0.001 0.001 3.530 3.538 integrate_v_rspace 119 11.3 0.003 0.003 3.519 3.526 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.300 3.337 calculate_rho_elec 119 8.7 0.040 0.046 3.300 3.336 init_scf_loop 11 6.9 0.000 0.000 2.834 2.835 acc_transpose_blocks 27432 15.5 0.118 0.122 2.449 2.564 mp_waitall_1 137007 16.6 1.969 2.561 1.969 2.561 make_m2s 4572 13.5 0.054 0.056 2.369 2.487 calculate_first_density_matrix 1 7.0 0.000 0.000 2.416 2.418 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.355 2.377 make_images 4572 14.5 0.273 0.335 2.261 2.376 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.114 2.250 apply_single 119 13.6 0.000 0.000 2.113 2.250 calculate_dm_sparse 119 9.5 0.000 0.000 2.162 2.218 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.179 2.189 prepare_preconditioner 11 7.9 0.000 0.000 1.982 1.992 make_preconditioner 11 8.9 0.000 0.000 1.982 1.992 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.927 1.928 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.725 1.921 grid_integrate_task_list 119 12.3 1.829 1.913 1.829 1.913 density_rs2pw 119 9.7 0.003 0.004 1.757 1.871 ot_diis_step 108 11.5 0.012 0.012 1.826 1.826 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.773 1.773 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.733 1.763 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.707 1.718 acc_transpose_blocks_sync 82296 16.5 1.457 1.554 1.457 1.554 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.517 1.518 cp_fm_redistribute_end 50 14.0 0.761 1.495 0.765 1.497 cp_fm_diag_elpa_base 50 14.0 0.705 1.428 0.729 1.470 fft_wrap_pw1pw2_140 487 12.2 0.071 0.072 1.421 1.453 fft3d_ps 1201 13.6 0.573 0.635 1.395 1.420 wfi_extrapolate 11 7.9 0.001 0.001 1.354 1.354 multiply_cannon_metrocomm3 27432 15.5 0.039 0.040 0.801 1.350 potential_pw2rs 119 12.3 0.008 0.009 1.330 1.333 grid_collocate_task_list 119 9.7 1.249 1.327 1.249 1.327 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.301 1.319 mp_sum_l 7287 12.8 1.010 1.311 1.010 1.311 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.256 1.268 jit_kernel_multiply 6 16.0 0.456 1.153 0.456 1.153 qs_energies_init_hamiltonians 11 5.9 0.019 0.022 1.118 1.138 transfer_rs2pw 487 10.6 0.005 0.005 0.983 1.119 make_images_data 4572 15.5 0.047 0.051 0.903 1.034 hybrid_alltoall_any 4725 16.4 0.066 0.157 0.757 0.959 mp_alltoall_d11v 2130 13.8 0.793 0.919 0.793 0.919 dbcsr_complete_redistribute 329 12.2 0.138 0.159 0.720 0.883 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.799 0.877 acc_transpose_blocks_kernels 27432 16.5 0.272 0.279 0.846 0.863 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.826 0.831 transfer_pw2rs 487 13.2 0.004 0.004 0.822 0.826 mp_alltoall_z22v 1201 15.6 0.706 0.735 0.706 0.735 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.224000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.818182, yerr=3.009626 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 626.786304E+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 62672. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3629 73907. 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 0.024 0.060 29.108 29.109 qs_mol_dyn_low 1 2.0 0.003 0.003 28.835 28.847 qs_forces 11 3.9 0.002 0.003 28.761 28.762 qs_energies 11 4.9 0.002 0.002 27.066 27.069 scf_env_do_scf 11 5.9 0.000 0.001 21.909 21.909 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.582 19.582 velocity_verlet 10 3.0 0.006 0.019 14.426 14.429 dbcsr_multiply_generic 2286 12.5 0.124 0.128 13.732 13.806 qs_scf_new_mos 108 7.5 0.001 0.001 12.015 12.043 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.015 12.042 multiply_cannon 2286 13.5 0.222 0.227 11.033 11.495 ot_scf_mini 108 9.5 0.002 0.002 11.302 11.329 multiply_cannon_loop 2286 14.5 0.646 0.660 10.079 10.230 multiply_cannon_multrec 9144 15.5 1.755 1.964 6.401 6.617 ot_mini 108 10.5 0.001 0.001 6.446 6.478 rebuild_ks_matrix 119 8.3 0.000 0.000 5.824 5.845 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.824 5.844 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.193 5.212 qs_ot_get_derivative 108 11.5 0.001 0.001 5.141 5.168 dbcsr_mm_accdrv_process 12550 15.8 3.699 4.500 4.538 4.627 init_scf_run 11 5.9 0.000 0.001 3.703 3.703 scf_env_initial_rho_setup 11 6.9 0.021 0.024 3.702 3.702 sum_up_and_integrate 119 10.3 0.001 0.001 3.389 3.402 integrate_v_rspace 119 11.3 0.003 0.004 3.379 3.392 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.223 3.228 calculate_rho_elec 119 8.7 0.060 0.061 3.222 3.227 qs_ot_get_p 119 10.4 0.001 0.001 2.948 2.983 calculate_first_density_matrix 1 7.0 0.000 0.000 2.563 2.567 init_scf_loop 11 6.9 0.000 0.000 2.307 2.307 make_m2s 4572 13.5 0.034 0.036 1.875 2.034 calculate_dm_sparse 119 9.5 0.000 0.000 1.926 1.946 make_images 4572 14.5 0.270 0.301 1.785 1.942 grid_integrate_task_list 119 12.3 1.855 1.941 1.855 1.941 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.932 1.933 mp_waitall_1 115863 16.7 1.381 1.897 1.381 1.897 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.856 1.858 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.803 1.825 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.743 1.752 density_rs2pw 119 9.7 0.003 0.003 1.608 1.713 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.686 1.692 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.685 1.686 acc_transpose_blocks 9144 15.5 0.042 0.043 1.618 1.660 jit_kernel_multiply 8 15.4 0.799 1.659 0.799 1.659 prepare_preconditioner 11 7.9 0.000 0.000 1.546 1.550 make_preconditioner 11 8.9 0.000 0.000 1.546 1.549 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.442 1.466 fft_wrap_pw1pw2_140 487 12.2 0.094 0.096 1.439 1.449 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.381 1.391 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.366 1.374 cp_fm_diag_elpa_base 50 14.0 1.340 1.356 1.365 1.373 grid_collocate_task_list 119 9.7 1.299 1.372 1.299 1.372 fft3d_ps 1201 13.6 0.633 0.649 1.354 1.366 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.312 1.318 ot_diis_step 108 11.5 0.013 0.013 1.292 1.292 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.236 1.260 apply_single 119 13.6 0.000 0.000 1.235 1.260 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.224 1.225 potential_pw2rs 119 12.3 0.010 0.010 1.195 1.198 wfi_extrapolate 11 7.9 0.001 0.001 1.062 1.062 hybrid_alltoall_any 4725 16.4 0.066 0.177 0.785 1.016 make_images_data 4572 15.5 0.042 0.045 0.804 0.980 mp_alltoall_d11v 2130 13.8 0.833 0.957 0.833 0.957 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.874 0.923 transfer_rs2pw 487 10.6 0.004 0.005 0.789 0.887 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.836 0.861 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.833 0.839 cp_fm_cholesky_invert 11 10.9 0.752 0.758 0.752 0.758 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.751 0.753 acc_transpose_blocks_sync 27432 16.5 0.724 0.744 0.724 0.744 jit_kernel_transpose 5 15.6 0.718 0.742 0.718 0.742 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.671 0.721 mp_allgather_i34 2286 14.5 0.249 0.698 0.249 0.698 transfer_pw2rs 487 13.2 0.003 0.004 0.691 0.694 mp_alltoall_z22v 1201 15.6 0.597 0.634 0.597 0.634 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.313 0.625 dbcsr_complete_redistribute 329 12.2 0.157 0.165 0.556 0.583 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.109000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=593.727273, yerr=6.523955 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 776.634368E+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 3629 74142. 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 0.015 0.034 41.521 41.521 qs_mol_dyn_low 1 2.0 0.003 0.004 41.322 41.329 qs_forces 11 3.9 0.002 0.002 41.211 41.211 qs_energies 11 4.9 0.002 0.002 39.174 39.179 scf_env_do_scf 11 5.9 0.001 0.001 33.026 33.027 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.961 26.963 velocity_verlet 10 3.0 0.016 0.057 22.700 22.706 dbcsr_multiply_generic 2286 12.5 0.106 0.107 19.367 19.559 qs_scf_new_mos 108 7.5 0.001 0.001 17.542 17.635 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.541 17.634 ot_scf_mini 108 9.5 0.002 0.002 16.363 16.455 multiply_cannon 2286 13.5 0.297 0.305 15.166 16.108 multiply_cannon_loop 2286 14.5 0.863 0.893 13.883 14.794 ot_mini 108 10.5 0.001 0.001 9.909 10.027 multiply_cannon_multrec 9144 15.5 3.394 4.849 8.828 8.928 qs_ot_get_derivative 108 11.5 0.001 0.001 7.829 7.925 rebuild_ks_matrix 119 8.3 0.000 0.000 7.153 7.301 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.018 7.152 7.300 dbcsr_mm_accdrv_process 12550 15.8 4.585 6.559 5.302 6.618 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.441 6.573 init_scf_loop 11 6.9 0.000 0.000 6.035 6.038 prepare_preconditioner 11 7.9 0.000 0.000 5.035 5.049 make_preconditioner 11 8.9 0.000 0.000 5.035 5.048 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.053 4.930 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.102 4.171 calculate_rho_elec 119 8.7 0.118 0.122 4.102 4.170 init_scf_run 11 5.9 0.000 0.001 4.004 4.004 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.004 4.004 sum_up_and_integrate 119 10.3 0.001 0.001 3.806 3.813 integrate_v_rspace 119 11.3 0.004 0.004 3.796 3.803 qs_ot_get_p 119 10.4 0.001 0.001 3.469 3.612 mp_waitall_1 94719 16.7 2.519 3.410 2.519 3.410 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.893 3.317 cp_fm_upper_to_full 72 14.2 2.053 2.945 2.053 2.945 make_m2s 4572 13.5 0.038 0.038 2.531 2.694 make_images 4572 14.5 0.351 0.383 2.411 2.574 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.240 2.546 apply_single 119 13.6 0.000 0.000 2.240 2.546 calculate_first_density_matrix 1 7.0 0.000 0.000 2.410 2.483 calculate_dm_sparse 119 9.5 0.000 0.000 2.393 2.464 fft_wrap_pw1pw2 1201 11.6 0.014 0.014 2.456 2.462 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.382 2.223 grid_integrate_task_list 119 12.3 2.087 2.149 2.087 2.149 dbcsr_complete_redistribute 329 12.2 0.255 0.258 1.615 2.135 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.083 2.133 density_rs2pw 119 9.7 0.003 0.003 2.056 2.074 fft_wrap_pw1pw2_140 487 12.2 0.179 0.180 2.061 2.068 ot_diis_step 108 11.5 0.014 0.014 2.050 2.050 mp_sum_l 7287 12.8 1.123 1.994 1.123 1.994 qs_ot_p2m_diag 50 11.0 0.042 0.043 1.989 1.990 fft3d_ps 1201 13.6 0.855 0.873 1.909 1.915 acc_transpose_blocks 9144 15.5 0.044 0.045 1.833 1.907 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.892 1.893 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.857 1.858 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.263 1.781 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.607 1.653 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.629 1.631 grid_collocate_task_list 119 9.7 1.526 1.564 1.526 1.564 mp_alltoall_i22 627 13.8 0.960 1.503 0.960 1.503 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 0.977 1.493 wfi_extrapolate 11 7.9 0.001 0.001 1.451 1.451 hybrid_alltoall_any 4725 16.4 0.090 0.150 1.186 1.441 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.375 1.440 mp_alltoall_d11v 2130 13.8 1.331 1.433 1.331 1.433 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.353 1.379 make_images_data 4572 15.5 0.045 0.048 1.157 1.378 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.324 1.324 cp_fm_diag_elpa_base 50 14.0 1.238 1.251 1.322 1.323 potential_pw2rs 119 12.3 0.014 0.014 1.264 1.267 cp_fm_cholesky_invert 11 10.9 1.198 1.205 1.198 1.205 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.175 1.197 acc_transpose_blocks_sync 27432 16.5 1.100 1.167 1.100 1.167 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.000 1.046 mp_alltoall_z22v 1201 15.6 0.919 0.951 0.919 0.951 qs_create_task_list 11 7.9 0.000 0.000 0.933 0.942 generate_qs_task_list 11 8.9 0.368 0.388 0.933 0.942 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.924 0.939 transfer_rs2pw 487 10.6 0.005 0.005 0.868 0.896 jit_kernel_multiply 5 15.4 0.687 0.853 0.687 0.853 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.521000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=729.545455, yerr=14.487384 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 496.541696E+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 3528 76079. MP_Allreduce 9920 483. MP_Sync 52 MP_Alltoall 1938 401165. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 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 0.026 83.535 83.536 qs_mol_dyn_low 1 2.0 0.003 0.004 83.292 83.303 qs_forces 11 3.9 0.003 0.003 83.218 83.219 qs_energies 11 4.9 0.002 0.002 80.343 80.366 scf_env_do_scf 11 5.9 0.000 0.001 71.322 71.322 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 66.035 66.035 dbcsr_multiply_generic 2055 12.4 0.108 0.111 52.396 52.867 qs_scf_new_mos 99 7.5 0.000 0.001 49.034 49.174 qs_scf_loop_do_ot 99 8.5 0.000 0.001 49.034 49.173 ot_scf_mini 99 9.5 0.002 0.002 46.570 46.689 multiply_cannon 2055 13.4 0.175 0.180 43.010 44.200 multiply_cannon_loop 2055 14.4 1.842 1.910 42.010 43.282 velocity_verlet 10 3.0 0.001 0.002 42.335 42.335 ot_mini 99 10.5 0.001 0.001 27.871 27.998 qs_ot_get_derivative 99 11.5 0.001 0.001 20.935 21.065 multiply_cannon_multrec 49320 15.4 11.141 11.875 17.296 18.050 rebuild_ks_matrix 110 8.3 0.000 0.000 14.483 14.646 qs_ks_build_kohn_sham_matrix 110 9.3 0.010 0.011 14.482 14.645 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.705 12.854 mp_waitall_1 220248 16.4 11.412 12.561 11.412 12.561 qs_ot_get_p 110 10.4 0.001 0.001 10.090 10.217 multiply_cannon_sync_h2d 49320 15.4 9.360 10.120 9.360 10.120 multiply_cannon_metrocomm3 49320 15.4 0.079 0.083 6.916 8.534 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.369 7.905 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.229 7.782 apply_single 110 13.6 0.000 0.000 7.228 7.782 init_scf_run 11 5.9 0.000 0.001 6.982 6.983 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.982 6.983 sum_up_and_integrate 110 10.3 0.002 0.003 6.889 6.903 integrate_v_rspace 110 11.3 0.003 0.003 6.864 6.884 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.848 6.879 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.602 6.676 ot_diis_step 99 11.5 0.006 0.006 6.651 6.652 dbcsr_mm_accdrv_process 87628 16.1 3.192 3.398 6.027 6.401 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.234 6.388 calculate_rho_elec 110 8.6 0.020 0.024 6.234 6.388 mp_sum_l 6594 12.7 4.992 6.108 4.992 6.108 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.927 5.927 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.413 5.415 cp_fm_redistribute_end 48 14.0 2.747 5.387 2.760 5.392 cp_fm_diag_elpa_base 48 14.0 2.621 5.267 2.629 5.284 init_scf_loop 11 6.9 0.000 0.000 5.254 5.254 make_m2s 4110 13.4 0.061 0.066 3.929 4.055 wfi_extrapolate 11 7.9 0.001 0.001 3.977 3.978 calculate_dm_sparse 110 9.5 0.001 0.001 3.879 3.969 make_images 4110 14.4 0.177 0.190 3.833 3.960 multiply_cannon_metrocomm1 49320 15.4 0.064 0.067 2.622 3.620 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.582 3.584 density_rs2pw 110 9.6 0.004 0.004 3.198 3.438 grid_integrate_task_list 110 12.3 3.252 3.401 3.252 3.401 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.318 3.363 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.199 3.242 prepare_preconditioner 11 7.9 0.000 0.000 3.081 3.105 make_preconditioner 11 8.9 0.000 0.000 3.081 3.105 calculate_first_density_matrix 1 7.0 0.000 0.000 2.903 2.908 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.832 2.873 acc_transpose_blocks 49320 15.4 0.214 0.227 2.682 2.862 fft_wrap_pw1pw2 1111 11.6 0.013 0.018 2.671 2.734 mp_alltoall_d11v 2046 13.8 2.259 2.683 2.259 2.683 jit_kernel_multiply 13 15.9 2.552 2.585 2.552 2.585 potential_pw2rs 110 12.3 0.006 0.006 2.491 2.509 fft_wrap_pw1pw2_140 451 12.1 0.092 0.098 2.251 2.312 fft3d_ps 1111 13.6 0.725 0.812 2.210 2.271 grid_collocate_task_list 110 9.6 2.159 2.265 2.159 2.265 transfer_rs2pw 451 10.6 0.006 0.007 1.912 2.119 mp_waitany 14300 13.8 1.807 2.108 1.807 2.108 mp_sum_d 3891 11.9 1.484 2.085 1.484 2.085 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.916 1.935 make_images_data 4110 15.4 0.043 0.047 1.780 1.931 hybrid_alltoall_any 4261 16.3 0.084 0.482 1.542 1.807 transfer_pw2rs 451 13.1 0.006 0.007 1.751 1.760 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.724 1.751 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.536000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=472.363636, yerr=2.422348 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 590.397440E+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 3528 76294. MP_Allreduce 9919 558. MP_Sync 52 MP_Alltoall 1717 1935329. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 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 0.021 0.078 69.981 69.983 qs_mol_dyn_low 1 2.0 0.005 0.015 69.588 69.599 qs_forces 11 3.9 0.005 0.025 69.512 69.515 qs_energies 11 4.9 0.002 0.008 66.162 66.167 scf_env_do_scf 11 5.9 0.001 0.002 57.249 57.250 scf_env_do_scf_inner_loop 99 6.5 0.005 0.024 50.137 50.137 dbcsr_multiply_generic 2055 12.4 0.118 0.122 38.574 38.786 velocity_verlet 10 3.0 0.003 0.012 36.071 36.073 qs_scf_new_mos 99 7.5 0.001 0.001 34.160 34.321 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.160 34.321 multiply_cannon 2055 13.4 0.224 0.243 31.656 32.731 ot_scf_mini 99 9.5 0.003 0.004 32.468 32.607 multiply_cannon_loop 2055 14.4 1.159 1.182 30.298 31.280 ot_mini 99 10.5 0.001 0.001 18.834 18.985 multiply_cannon_multrec 24660 15.4 6.953 8.591 14.061 15.744 rebuild_ks_matrix 110 8.3 0.000 0.000 13.729 13.874 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.019 13.729 13.874 qs_ot_get_derivative 99 11.5 0.001 0.001 12.979 13.119 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.101 12.223 mp_waitall_1 176588 16.5 7.874 10.665 7.874 10.665 multiply_cannon_metrocomm3 24660 15.4 0.071 0.076 5.316 7.851 multiply_cannon_sync_h2d 24660 15.4 6.374 7.300 6.374 7.300 dbcsr_mm_accdrv_process 52282 16.1 5.515 6.086 6.940 7.251 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.557 7.135 apply_single 110 13.6 0.000 0.001 6.556 7.135 init_scf_loop 11 6.9 0.001 0.004 7.069 7.070 qs_ot_get_p 110 10.4 0.001 0.001 6.890 7.063 sum_up_and_integrate 110 10.3 0.002 0.009 6.404 6.417 integrate_v_rspace 110 11.3 0.003 0.003 6.377 6.389 init_scf_run 11 5.9 0.000 0.001 6.323 6.324 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.323 6.324 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.872 5.882 calculate_rho_elec 110 8.6 0.039 0.048 5.871 5.881 ot_diis_step 99 11.5 0.010 0.010 5.797 5.797 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.857 5.570 prepare_preconditioner 11 7.9 0.000 0.000 5.002 5.021 make_preconditioner 11 8.9 0.000 0.001 5.002 5.021 qs_ot_p2m_diag 48 11.0 0.031 0.043 4.965 4.987 make_m2s 4110 13.4 0.058 0.060 4.341 4.811 make_full_inverse_cholesky 11 9.9 0.000 0.001 4.648 4.742 make_images 4110 14.4 0.406 0.464 4.229 4.694 cp_dbcsr_syevd 48 12.0 0.005 0.019 4.484 4.487 cp_fm_diag_elpa 48 13.0 0.000 0.002 3.904 3.905 cp_fm_redistribute_end 48 14.0 1.962 3.865 1.968 3.868 cp_fm_diag_elpa_base 48 14.0 1.833 3.720 1.890 3.800 wfi_extrapolate 11 7.9 0.001 0.003 3.551 3.551 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.402 3.489 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.424 3.425 fft_wrap_pw1pw2 1111 11.6 0.014 0.017 3.238 3.386 grid_integrate_task_list 110 12.3 3.155 3.335 3.155 3.335 density_rs2pw 110 9.6 0.004 0.005 3.080 3.285 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 3.043 3.083 calculate_dm_sparse 110 9.5 0.001 0.001 3.004 3.029 hybrid_alltoall_any 4261 16.3 0.105 0.455 2.048 2.844 fft_wrap_pw1pw2_140 451 12.1 0.111 0.128 2.695 2.838 make_images_data 4110 15.4 0.048 0.052 2.347 2.836 fft3d_ps 1111 13.6 1.069 1.274 2.616 2.751 mp_sum_l 6594 12.7 1.971 2.731 1.971 2.731 calculate_first_density_matrix 1 7.0 0.000 0.003 2.659 2.662 cp_fm_cholesky_invert 11 10.9 2.462 2.480 2.462 2.480 grid_collocate_task_list 110 9.6 2.170 2.333 2.170 2.333 potential_pw2rs 110 12.3 0.008 0.009 2.275 2.287 mp_alltoall_d11v 2046 13.8 1.906 2.103 1.906 2.103 qs_energies_init_hamiltonians 11 5.9 0.003 0.014 2.075 2.078 acc_transpose_blocks 24660 15.4 0.117 0.120 1.987 2.063 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.021 2.033 jit_kernel_multiply 10 16.4 1.056 1.927 1.056 1.927 transfer_rs2pw 451 10.6 0.007 0.008 1.592 1.829 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.791 1.815 multiply_cannon_metrocomm4 22605 15.4 0.078 0.082 0.771 1.783 mp_allgather_i34 2055 14.4 0.662 1.711 0.662 1.711 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 1.548 1.564 mp_irecv_dv 57340 16.2 0.641 1.563 0.641 1.563 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.386 1.486 mp_waitany 10164 13.8 1.249 1.448 1.249 1.448 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.983000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=557.272727, yerr=6.876718 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 666.087424E+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 3480 77281. 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 0.030 0.056 60.249 60.251 qs_mol_dyn_low 1 2.0 0.022 0.053 59.653 59.666 qs_forces 11 3.9 0.003 0.003 59.402 59.424 qs_energies 11 4.9 0.005 0.016 56.186 56.222 scf_env_do_scf 11 5.9 0.001 0.002 48.095 48.097 scf_env_do_scf_inner_loop 99 6.5 0.006 0.018 40.226 40.228 velocity_verlet 10 3.0 0.008 0.018 31.987 31.994 dbcsr_multiply_generic 2055 12.4 0.115 0.123 29.076 29.381 qs_scf_new_mos 99 7.5 0.001 0.001 25.609 25.717 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.608 25.716 ot_scf_mini 99 9.5 0.003 0.003 24.348 24.491 multiply_cannon 2055 13.4 0.212 0.219 22.580 23.574 multiply_cannon_loop 2055 14.4 0.818 0.841 21.412 22.496 ot_mini 99 10.5 0.001 0.001 13.907 14.047 rebuild_ks_matrix 110 8.3 0.000 0.000 12.329 12.486 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.021 12.329 12.486 multiply_cannon_multrec 16440 15.4 3.646 4.955 9.895 11.038 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.853 10.994 mp_waitall_1 139946 16.5 7.005 10.121 7.005 10.121 qs_ot_get_derivative 99 11.5 0.001 0.001 9.467 9.608 init_scf_loop 11 6.9 0.001 0.003 7.832 7.834 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.364 7.357 sum_up_and_integrate 110 10.3 0.010 0.059 6.296 6.359 integrate_v_rspace 110 11.3 0.003 0.003 6.262 6.277 dbcsr_mm_accdrv_process 34862 16.1 5.408 5.830 6.096 6.248 prepare_preconditioner 11 7.9 0.000 0.000 5.890 5.905 make_preconditioner 11 8.9 0.000 0.002 5.890 5.905 qs_ot_get_p 110 10.4 0.014 0.050 5.704 5.863 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.648 5.664 calculate_rho_elec 110 8.6 0.058 0.059 5.647 5.663 init_scf_run 11 5.9 0.000 0.001 5.592 5.592 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.592 5.592 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.305 5.538 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.985 5.434 apply_single 110 13.6 0.000 0.000 4.985 5.434 make_m2s 4110 13.4 0.050 0.051 4.222 4.565 make_images 4110 14.4 0.399 0.523 4.105 4.445 ot_diis_step 99 11.5 0.011 0.011 4.407 4.407 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.929 3.932 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.245 3.921 multiply_cannon_sync_h2d 16440 15.4 3.207 3.902 3.207 3.902 cp_dbcsr_syevd 48 12.0 0.003 0.005 3.532 3.533 grid_integrate_task_list 110 12.3 3.190 3.392 3.190 3.392 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.178 3.191 density_rs2pw 110 9.6 0.008 0.028 2.850 3.033 wfi_extrapolate 11 7.9 0.004 0.022 2.953 2.953 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.929 2.938 cp_fm_diag_elpa_base 48 14.0 2.882 2.904 2.926 2.935 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.861 2.863 make_images_data 4110 15.4 0.045 0.050 2.343 2.812 hybrid_alltoall_any 4261 16.3 0.108 0.380 2.070 2.768 fft_wrap_pw1pw2_140 451 12.1 0.127 0.132 2.698 2.708 calculate_dm_sparse 110 9.5 0.001 0.001 2.587 2.623 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.503 2.586 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.470 2.527 cp_fm_cholesky_invert 11 10.9 2.511 2.523 2.511 2.523 fft3d_ps 1111 13.6 1.085 1.098 2.488 2.514 calculate_first_density_matrix 1 7.0 0.004 0.025 2.509 2.512 multiply_cannon_metrocomm4 14385 15.4 0.047 0.053 0.899 2.474 grid_collocate_task_list 110 9.6 2.219 2.363 2.219 2.363 mp_irecv_dv 48980 15.7 0.823 2.341 0.823 2.341 mp_sum_l 6594 12.7 1.661 2.285 1.661 2.285 mp_alltoall_d11v 2046 13.8 1.951 2.199 1.951 2.199 potential_pw2rs 110 12.3 0.010 0.010 2.116 2.121 qs_energies_init_hamiltonians 11 5.9 0.019 0.034 2.059 2.081 dbcsr_complete_redistribute 325 12.2 0.311 0.362 1.458 1.768 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.676 1.692 acc_transpose_blocks 16440 15.4 0.078 0.080 1.592 1.630 transfer_rs2pw 451 10.6 0.005 0.006 1.387 1.616 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.002 0.012 1.516 1.534 mp_waitany 17072 13.8 1.268 1.485 1.268 1.485 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.359 1.482 mp_allgather_i34 2055 14.4 0.476 1.463 0.476 1.463 cp_fm_cholesky_decompose 22 10.9 1.380 1.403 1.380 1.403 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.292 1.306 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.978 1.279 transfer_pw2rs 451 13.1 0.005 0.005 1.253 1.257 mp_alltoall_z22v 1111 15.6 1.192 1.233 1.192 1.233 cp_fm_upper_to_full 70 14.2 0.952 1.212 0.952 1.212 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.122 1.210 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.251000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.545455, yerr=9.451923 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 746.156032E+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 3480 77279. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.048 63.529 63.530 qs_mol_dyn_low 1 2.0 0.003 0.004 63.186 63.198 qs_forces 11 3.9 0.003 0.003 63.114 63.114 qs_energies 11 4.9 0.001 0.002 59.730 59.733 scf_env_do_scf 11 5.9 0.000 0.001 51.422 51.427 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.069 42.069 velocity_verlet 10 3.0 0.003 0.008 35.090 35.093 dbcsr_multiply_generic 2055 12.4 0.122 0.125 30.856 31.082 qs_scf_new_mos 99 7.5 0.001 0.001 27.611 27.722 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.611 27.722 ot_scf_mini 99 9.5 0.002 0.003 25.919 25.998 multiply_cannon 2055 13.4 0.237 0.252 23.701 24.955 multiply_cannon_loop 2055 14.4 1.419 1.476 22.310 23.043 ot_mini 99 10.5 0.001 0.001 14.929 15.034 multiply_cannon_multrec 24660 15.4 4.070 6.847 13.047 14.166 rebuild_ks_matrix 110 8.3 0.000 0.000 11.944 12.052 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.944 12.051 qs_ot_get_derivative 99 11.5 0.001 0.001 10.726 10.811 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.552 10.646 dbcsr_mm_accdrv_process 52304 16.0 7.814 9.216 8.821 9.734 init_scf_loop 11 6.9 0.000 0.000 9.311 9.311 prepare_preconditioner 11 7.9 0.000 0.000 7.570 7.584 make_preconditioner 11 8.9 0.000 0.000 7.570 7.584 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.429 7.239 mp_waitall_1 121746 16.5 4.321 6.189 4.321 6.189 sum_up_and_integrate 110 10.3 0.001 0.002 6.146 6.159 integrate_v_rspace 110 11.3 0.003 0.003 6.120 6.132 qs_ot_get_p 110 10.4 0.001 0.001 5.863 5.987 make_m2s 4110 13.4 0.060 0.062 5.492 5.752 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.706 5.717 calculate_rho_elec 110 8.6 0.077 0.081 5.705 5.717 make_images 4110 14.4 0.575 0.694 5.350 5.605 init_scf_run 11 5.9 0.000 0.001 5.601 5.602 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.601 5.602 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.079 4.164 apply_single 110 13.6 0.000 0.000 4.078 4.164 ot_diis_step 99 11.5 0.011 0.011 4.162 4.162 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.920 3.936 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.575 3.620 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.422 3.473 grid_integrate_task_list 110 12.3 3.293 3.434 3.293 3.434 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.340 3.341 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.234 3.260 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.444 3.150 calculate_dm_sparse 110 9.5 0.001 0.001 3.017 3.059 make_images_data 4110 15.4 0.048 0.052 2.705 2.987 hybrid_alltoall_any 4261 16.3 0.122 0.461 2.358 2.969 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.966 2.968 density_rs2pw 110 9.6 0.003 0.004 2.828 2.968 wfi_extrapolate 11 7.9 0.001 0.001 2.964 2.964 cp_fm_upper_to_full 70 14.2 2.063 2.916 2.063 2.916 dbcsr_complete_redistribute 325 12.2 0.400 0.476 2.147 2.834 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.778 2.788 cp_fm_diag_elpa_base 48 14.0 2.713 2.736 2.774 2.784 fft_wrap_pw1pw2_140 451 12.1 0.145 0.149 2.744 2.772 acc_transpose_blocks 24660 15.4 0.111 0.113 2.470 2.569 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.521 2.561 calculate_first_density_matrix 1 7.0 0.000 0.000 2.536 2.539 fft3d_ps 1111 13.6 1.125 1.166 2.503 2.527 cp_fm_cholesky_invert 11 10.9 2.477 2.487 2.477 2.487 multiply_cannon_sync_h2d 24660 15.4 2.365 2.486 2.365 2.486 grid_collocate_task_list 110 9.6 2.264 2.372 2.264 2.372 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.595 2.263 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.224 2.225 mp_alltoall_d11v 2046 13.8 1.909 2.129 1.909 2.129 potential_pw2rs 110 12.3 0.012 0.013 1.954 1.960 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.819 1.853 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.132 1.783 mp_alltoall_i22 605 13.7 1.035 1.769 1.035 1.769 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.610 1.707 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.639 1.650 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.577 1.596 mp_sum_l 6594 12.7 0.977 1.579 0.977 1.579 mp_allgather_i34 2055 14.4 0.474 1.572 0.474 1.572 acc_transpose_blocks_sync 73980 16.4 1.423 1.526 1.423 1.526 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.851 1.438 transfer_rs2pw 451 10.6 0.005 0.006 1.259 1.416 cp_fm_cholesky_decompose 22 10.9 1.357 1.397 1.357 1.397 mp_irecv_dv 62702 16.1 0.748 1.353 0.748 1.353 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=63.530000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=699.636364, yerr=8.647361 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 862.720000E+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 258767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3480 77277. 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 0.012 0.040 54.849 54.849 qs_mol_dyn_low 1 2.0 0.003 0.003 54.538 54.557 qs_forces 11 3.9 0.003 0.003 54.467 54.468 qs_energies 11 4.9 0.002 0.002 50.838 50.843 scf_env_do_scf 11 5.9 0.000 0.001 42.643 42.643 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.448 35.449 velocity_verlet 10 3.0 0.002 0.002 30.353 30.357 dbcsr_multiply_generic 2055 12.4 0.111 0.112 23.588 23.779 qs_scf_new_mos 99 7.5 0.001 0.001 21.103 21.174 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.103 21.174 ot_scf_mini 99 9.5 0.002 0.002 19.840 19.870 multiply_cannon 2055 13.4 0.236 0.248 17.891 19.399 multiply_cannon_loop 2055 14.4 0.601 0.623 16.598 17.015 rebuild_ks_matrix 110 8.3 0.000 0.000 11.610 11.623 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.609 11.623 ot_mini 99 10.5 0.001 0.001 10.928 10.949 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.343 10.356 multiply_cannon_multrec 8220 15.4 3.225 4.345 7.753 8.684 mp_waitall_1 103326 16.6 5.916 7.400 5.916 7.400 qs_ot_get_derivative 99 11.5 0.001 0.001 7.233 7.262 init_scf_loop 11 6.9 0.000 0.000 7.147 7.148 sum_up_and_integrate 110 10.3 0.001 0.002 6.111 6.123 integrate_v_rspace 110 11.3 0.003 0.003 6.084 6.096 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.752 5.768 calculate_rho_elec 110 8.6 0.116 0.116 5.752 5.767 prepare_preconditioner 11 7.9 0.000 0.000 5.442 5.445 make_preconditioner 11 8.9 0.000 0.000 5.442 5.445 dbcsr_mm_accdrv_process 17442 15.9 3.238 4.333 4.389 5.380 init_scf_run 11 5.9 0.000 0.001 5.111 5.111 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.111 5.111 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.026 5.099 qs_ot_get_p 110 10.4 0.005 0.015 4.968 4.995 make_m2s 4110 13.4 0.038 0.040 4.284 4.576 make_images 4110 14.4 0.646 0.703 4.153 4.447 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 3.003 4.440 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.734 3.768 apply_single 110 13.6 0.000 0.000 3.734 3.768 ot_diis_step 99 11.5 0.012 0.012 3.669 3.669 grid_integrate_task_list 110 12.3 3.387 3.520 3.387 3.520 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.506 3.510 fft_wrap_pw1pw2 1111 11.6 0.014 0.016 3.406 3.414 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.161 3.162 density_rs2pw 110 9.6 0.004 0.004 2.752 2.945 fft_wrap_pw1pw2_140 451 12.1 0.192 0.194 2.926 2.944 hybrid_alltoall_any 4261 16.3 0.201 0.849 2.289 2.741 make_images_data 4110 15.4 0.040 0.047 2.299 2.672 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.655 2.655 calculate_dm_sparse 110 9.5 0.001 0.001 2.551 2.596 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.578 2.586 cp_fm_diag_elpa_base 48 14.0 2.521 2.548 2.576 2.584 fft3d_ps 1111 13.6 1.268 1.292 2.564 2.579 cp_fm_cholesky_invert 11 10.9 2.569 2.576 2.569 2.576 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.573 2.574 wfi_extrapolate 11 7.9 0.001 0.001 2.568 2.568 grid_collocate_task_list 110 9.6 2.363 2.563 2.363 2.563 multiply_cannon_sync_h2d 8220 15.4 2.388 2.487 2.388 2.487 calculate_first_density_matrix 1 7.0 0.000 0.000 2.451 2.453 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.185 2.199 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.085 2.117 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.789 2.000 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.969 1.991 mp_alltoall_d11v 2046 13.8 1.743 1.906 1.743 1.906 potential_pw2rs 110 12.3 0.015 0.015 1.895 1.900 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.481 1.600 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.597 1.599 dbcsr_complete_redistribute 325 12.2 0.536 0.563 1.488 1.586 mp_allgather_i34 2055 14.4 0.458 1.563 0.458 1.563 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.500 1.510 cp_fm_cholesky_decompose 22 10.9 1.446 1.466 1.446 1.466 acc_transpose_blocks 8220 15.4 0.040 0.041 1.398 1.441 transfer_rs2pw 451 10.6 0.005 0.005 1.158 1.364 qs_create_task_list 11 7.9 0.000 0.000 1.204 1.305 generate_qs_task_list 11 8.9 0.375 0.442 1.204 1.305 mp_waitany 9240 13.8 1.074 1.290 1.074 1.290 jit_kernel_multiply 7 15.4 0.830 1.286 0.830 1.286 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.205 1.224 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.775 1.161 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.849000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=815.000000, yerr=9.409859 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 1.327235E+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 3469 77990. 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 0.096 0.135 91.821 91.822 qs_mol_dyn_low 1 2.0 0.003 0.004 91.219 91.230 qs_forces 11 3.9 0.003 0.003 91.145 91.146 qs_energies 11 4.9 0.003 0.003 86.752 86.764 scf_env_do_scf 11 5.9 0.001 0.001 75.626 75.627 velocity_verlet 10 3.0 0.002 0.002 55.273 55.280 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 52.741 52.743 dbcsr_multiply_generic 2055 12.4 0.134 0.137 34.188 34.766 qs_scf_new_mos 99 7.5 0.001 0.001 32.813 32.965 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.812 32.964 ot_scf_mini 99 9.5 0.002 0.002 30.795 30.935 multiply_cannon 2055 13.4 0.343 0.382 24.547 25.288 multiply_cannon_loop 2055 14.4 0.825 0.850 22.101 23.935 init_scf_loop 11 6.9 0.000 0.000 22.802 22.806 prepare_preconditioner 11 7.9 0.000 0.000 20.573 20.596 make_preconditioner 11 8.9 0.000 0.000 20.573 20.595 make_full_inverse_cholesky 11 9.9 0.000 0.000 16.664 19.993 ot_mini 99 10.5 0.001 0.001 17.170 17.351 rebuild_ks_matrix 110 8.3 0.000 0.001 15.204 15.451 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 15.204 15.450 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.751 13.975 mp_waitall_1 84994 16.7 10.097 13.752 10.097 13.752 qs_ot_get_derivative 99 11.5 0.001 0.001 11.895 12.038 cp_fm_upper_to_full 70 14.2 8.361 11.797 8.361 11.797 multiply_cannon_multrec 8220 15.4 4.111 4.285 9.894 10.002 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.725 8.780 calculate_rho_elec 110 8.6 0.227 0.227 8.724 8.779 dbcsr_complete_redistribute 325 12.2 0.929 0.943 5.942 7.943 qs_ot_get_p 110 10.4 0.001 0.001 7.673 7.888 make_m2s 4110 13.4 0.043 0.044 6.764 7.838 sum_up_and_integrate 110 10.3 0.002 0.002 7.817 7.834 integrate_v_rspace 110 11.3 0.003 0.003 7.788 7.805 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 6.071 7.802 make_images 4110 14.4 0.883 0.943 6.573 7.647 copy_fm_to_dbcsr 174 11.2 0.001 0.001 4.698 6.694 init_scf_run 11 5.9 0.000 0.001 6.637 6.637 scf_env_initial_rho_setup 11 6.9 0.002 0.003 6.637 6.637 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 3.893 5.898 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.237 5.877 apply_single 110 13.6 0.000 0.000 5.237 5.877 dbcsr_mm_accdrv_process 11614 15.7 3.782 4.128 5.632 5.826 fft_wrap_pw1pw2 1111 11.6 0.016 0.016 5.727 5.774 mp_alltoall_i22 605 13.7 3.586 5.687 3.586 5.687 qs_ot_p2m_diag 48 11.0 0.151 0.155 5.418 5.434 make_images_data 4110 15.4 0.044 0.047 3.665 5.415 cp_fm_cholesky_invert 11 10.9 5.325 5.337 5.325 5.337 ot_diis_step 99 11.5 0.015 0.016 5.198 5.199 fft_wrap_pw1pw2_140 451 12.1 0.364 0.365 4.904 4.959 hybrid_alltoall_any 4261 16.3 0.268 0.587 3.384 4.910 density_rs2pw 110 9.6 0.004 0.004 4.820 4.859 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.793 4.794 fft3d_ps 1111 13.6 1.835 1.883 4.585 4.628 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.775 4.287 wfi_extrapolate 11 7.9 0.001 0.001 3.974 3.974 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.964 3.964 cp_fm_diag_elpa_base 48 14.0 3.661 3.737 3.961 3.962 grid_integrate_task_list 110 12.3 3.742 3.807 3.742 3.807 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.798 3.799 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.643 3.706 calculate_dm_sparse 110 9.5 0.001 0.001 3.540 3.557 mp_alltoall_d11v 2046 13.8 3.204 3.300 3.204 3.300 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.283 3.289 multiply_cannon_sync_h2d 8220 15.4 3.136 3.164 3.136 3.164 potential_pw2rs 110 12.3 0.021 0.021 2.954 2.978 mp_sum_l 6594 12.7 1.987 2.938 1.987 2.938 grid_collocate_task_list 110 9.6 2.717 2.755 2.717 2.755 mp_allgather_i34 2055 14.4 1.168 2.739 1.168 2.739 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.568 2.646 mp_alltoall_z22v 1111 15.6 2.499 2.540 2.499 2.540 calculate_first_density_matrix 1 7.0 0.000 0.000 2.503 2.504 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.324 2.380 cp_fm_cholesky_decompose 22 10.9 2.187 2.215 2.187 2.215 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.106 2.215 copy_dbcsr_to_fm 151 11.3 0.003 0.003 2.057 2.131 qs_ot_get_orbitals 99 10.5 0.000 0.000 2.055 2.094 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.057 2.078 transfer_rs2pw 451 10.6 0.006 0.006 2.028 2.068 qs_create_task_list 11 7.9 0.000 0.000 1.895 1.941 generate_qs_task_list 11 8.9 0.734 0.788 1.895 1.941 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.880 1.903 acc_transpose_blocks 8220 15.4 0.041 0.041 1.778 1.850 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=91.822000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1210.272727, yerr=41.446770 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 634.073088E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4009 67203. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 3705764. 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 0.051 0.165 205.631 205.656 qs_mol_dyn_low 1 2.0 0.016 0.048 204.720 204.786 qs_forces 11 3.9 0.006 0.014 204.539 204.543 qs_energies 11 4.9 0.017 0.070 198.878 198.913 scf_env_do_scf 11 5.9 0.003 0.018 181.923 181.940 scf_env_do_scf_inner_loop 117 6.6 0.007 0.030 162.830 162.833 dbcsr_multiply_generic 2507 12.6 0.178 0.182 125.591 126.296 qs_scf_new_mos 117 7.6 0.001 0.001 123.471 123.749 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.470 123.748 velocity_verlet 10 3.0 0.003 0.006 121.973 121.975 ot_scf_mini 117 9.6 0.003 0.004 116.773 117.053 multiply_cannon 2507 13.6 0.237 0.245 101.830 103.979 multiply_cannon_loop 2507 14.6 2.417 2.475 99.640 101.588 ot_mini 117 10.6 0.001 0.001 65.688 65.923 multiply_cannon_multrec 60168 15.6 31.795 33.514 41.678 43.565 qs_ot_get_derivative 117 11.6 0.001 0.001 40.875 41.082 rebuild_ks_matrix 128 8.3 0.001 0.001 33.840 34.095 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.038 33.839 34.094 mp_waitall_1 267128 16.5 28.990 32.395 28.990 32.395 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.405 30.642 qs_ot_get_p 128 10.4 0.001 0.002 28.829 29.119 multiply_cannon_sync_h2d 60168 15.6 26.503 28.468 26.503 28.468 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.373 24.873 apply_single 128 13.6 0.001 0.001 24.372 24.873 ot_diis_step 117 11.6 0.007 0.008 24.586 24.587 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.238 22.358 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.393 19.614 cp_dbcsr_syevd 83 12.4 0.005 0.007 19.224 19.225 init_scf_loop 11 6.9 0.002 0.010 18.989 18.992 multiply_cannon_metrocomm3 60168 15.6 0.116 0.121 15.971 17.797 cp_fm_diag_elpa 83 13.4 0.001 0.002 16.127 16.163 cp_fm_diag_elpa_base 83 14.4 16.074 16.114 16.119 16.158 make_m2s 5014 13.6 0.106 0.112 14.000 14.368 prepare_preconditioner 11 7.9 0.000 0.001 14.297 14.349 make_preconditioner 11 8.9 0.000 0.001 14.297 14.349 make_images 5014 14.6 0.399 0.417 13.814 14.192 sum_up_and_integrate 128 10.3 0.002 0.006 13.920 13.944 integrate_v_rspace 128 11.3 0.004 0.004 13.861 13.888 make_full_inverse_cholesky 11 9.9 0.000 0.001 13.546 13.704 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.894 13.027 calculate_rho_elec 128 8.7 0.045 0.064 12.894 13.026 init_scf_run 11 5.9 0.000 0.002 12.623 12.624 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.622 12.623 dbcsr_mm_accdrv_process 124484 16.2 4.723 4.922 9.448 10.006 mp_sum_l 7950 12.9 8.675 10.004 8.675 10.004 wfi_extrapolate 11 7.9 0.001 0.004 8.942 8.942 calculate_dm_sparse 128 9.5 0.001 0.001 8.603 8.690 multiply_cannon_metrocomm1 60168 15.6 0.094 0.097 6.381 8.345 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.142 8.246 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.121 8.233 cp_fm_cholesky_invert 11 10.9 8.087 8.104 8.087 8.104 make_images_data 5014 15.6 0.064 0.070 6.831 7.765 grid_integrate_task_list 128 12.3 7.053 7.492 7.053 7.492 density_rs2pw 128 9.7 0.006 0.009 6.462 7.365 hybrid_alltoall_any 5200 16.5 0.294 2.256 6.012 7.073 fft_wrap_pw1pw2 1291 11.7 0.017 0.025 6.681 6.967 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 6.945 6.959 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.993 6.107 fft_wrap_pw1pw2_140 523 12.2 0.234 0.245 5.664 5.848 fft3d_ps 1291 13.7 2.090 2.581 5.500 5.727 mp_alltoall_d11v 2415 14.1 4.612 5.615 4.612 5.615 grid_collocate_task_list 128 9.7 4.840 5.238 4.840 5.238 potential_pw2rs 128 12.3 0.009 0.010 4.413 4.449 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.656000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.090909, yerr=7.292020 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 848.793600E+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 4009 67655. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 4729031. 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 0.074 0.232 183.792 183.795 qs_mol_dyn_low 1 2.0 0.022 0.042 183.134 183.150 qs_forces 11 3.9 0.004 0.005 182.995 183.002 qs_energies 11 4.9 0.003 0.009 176.189 176.206 scf_env_do_scf 11 5.9 0.001 0.002 159.866 159.882 scf_env_do_scf_inner_loop 117 6.6 0.005 0.025 133.356 133.359 velocity_verlet 10 3.0 0.003 0.005 113.220 113.222 dbcsr_multiply_generic 2507 12.6 0.188 0.194 97.661 98.823 qs_scf_new_mos 117 7.6 0.001 0.001 94.967 95.526 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.967 95.525 ot_scf_mini 117 9.6 0.004 0.005 90.133 90.754 multiply_cannon 2507 13.6 0.505 0.559 77.331 81.692 multiply_cannon_loop 2507 14.6 1.564 1.625 73.931 76.708 ot_mini 117 10.6 0.001 0.001 50.319 50.878 mp_waitall_1 214728 16.6 24.373 38.502 24.373 38.502 multiply_cannon_multrec 30084 15.6 20.890 25.700 31.577 36.841 rebuild_ks_matrix 128 8.3 0.001 0.001 32.596 33.402 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 32.596 33.402 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.328 30.073 qs_ot_get_derivative 117 11.6 0.001 0.002 28.250 28.853 multiply_cannon_metrocomm3 30084 15.6 0.096 0.102 15.761 28.360 init_scf_loop 11 6.9 0.003 0.020 26.411 26.413 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.212 23.378 apply_single 128 13.6 0.001 0.001 22.212 23.378 qs_ot_get_p 128 10.4 0.005 0.033 21.861 22.517 prepare_preconditioner 11 7.9 0.000 0.000 21.935 22.009 make_preconditioner 11 8.9 0.000 0.001 21.935 22.009 ot_diis_step 117 11.6 0.014 0.015 21.883 21.885 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.766 21.171 multiply_cannon_sync_h2d 30084 15.6 17.901 20.638 17.901 20.638 qs_ot_p2m_diag 83 11.4 0.188 0.217 16.867 16.903 make_m2s 5014 13.6 0.088 0.094 14.272 15.936 make_images 5014 14.6 1.162 1.359 14.055 15.713 cp_dbcsr_syevd 83 12.4 0.006 0.006 15.653 15.653 sum_up_and_integrate 128 10.3 0.002 0.004 13.821 13.851 integrate_v_rspace 128 11.3 0.003 0.004 13.761 13.793 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.970 13.003 calculate_rho_elec 128 8.7 0.088 0.106 12.969 13.003 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.232 12.265 cp_fm_diag_elpa_base 83 14.4 12.054 12.099 12.221 12.251 cp_fm_cholesky_invert 11 10.9 12.000 12.035 12.000 12.035 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.214 11.667 init_scf_run 11 5.9 0.000 0.001 11.493 11.494 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.492 11.494 multiply_cannon_metrocomm4 27577 15.6 0.104 0.118 3.864 10.991 dbcsr_mm_accdrv_process 62242 16.2 5.536 6.218 10.135 10.667 mp_irecv_dv 69486 16.3 3.656 10.578 3.656 10.578 make_images_data 5014 15.6 0.065 0.075 8.473 10.462 hybrid_alltoall_any 5200 16.5 0.349 1.513 7.302 9.847 wfi_extrapolate 11 7.9 0.001 0.001 8.254 8.254 fft_wrap_pw1pw2 1291 11.7 0.019 0.024 7.476 7.568 grid_integrate_task_list 128 12.3 7.164 7.558 7.164 7.558 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.280 7.080 density_rs2pw 128 9.7 0.006 0.006 6.677 7.068 calculate_dm_sparse 128 9.5 0.001 0.001 6.533 6.695 fft_wrap_pw1pw2_140 523 12.2 0.251 0.261 6.524 6.629 mp_sum_l 7950 12.9 4.271 6.343 4.271 6.343 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.249 6.258 fft3d_ps 1291 13.7 2.779 2.980 5.922 5.989 cp_fm_cholesky_decompose 22 10.9 5.636 5.722 5.636 5.722 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.408 5.501 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.268 5.449 grid_collocate_task_list 128 9.7 5.001 5.390 5.001 5.390 mp_alltoall_d11v 2415 14.1 4.520 5.143 4.520 5.143 mp_allgather_i34 2507 14.6 1.906 5.135 1.906 5.135 potential_pw2rs 128 12.3 0.014 0.015 4.352 4.378 dbcsr_complete_redistribute 395 12.7 0.718 0.862 3.397 4.065 mp_sum_d 4472 12.1 2.675 3.987 2.675 3.987 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.684 3.759 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=183.795000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=807.727273, yerr=2.004128 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.030316E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3999 67815. MP_Allreduce 11057 1000. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.033 171.157 171.158 qs_mol_dyn_low 1 2.0 0.004 0.004 170.734 170.752 qs_forces 11 3.9 0.004 0.005 170.638 170.639 qs_energies 11 4.9 0.002 0.003 164.055 164.063 scf_env_do_scf 11 5.9 0.001 0.001 148.383 148.384 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 119.183 119.184 velocity_verlet 10 3.0 0.001 0.002 107.704 107.706 dbcsr_multiply_generic 2507 12.6 0.190 0.195 82.420 83.451 qs_scf_new_mos 117 7.6 0.001 0.001 82.231 82.561 qs_scf_loop_do_ot 117 8.6 0.001 0.001 82.230 82.560 ot_scf_mini 117 9.6 0.003 0.004 78.016 78.379 multiply_cannon 2507 13.6 0.506 0.527 62.239 66.903 multiply_cannon_loop 2507 14.6 1.131 1.192 59.520 62.252 ot_mini 117 10.6 0.001 0.001 42.864 43.248 mp_waitall_1 170520 16.6 25.572 35.188 25.572 35.188 rebuild_ks_matrix 128 8.3 0.001 0.001 30.554 31.101 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 30.554 31.101 init_scf_loop 11 6.9 0.000 0.000 29.091 29.092 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.571 28.079 multiply_cannon_multrec 20056 15.6 12.940 16.281 22.377 25.669 prepare_preconditioner 11 7.9 0.000 0.000 25.083 25.133 make_preconditioner 11 8.9 0.000 0.000 25.083 25.133 multiply_cannon_metrocomm3 20056 15.6 0.059 0.063 15.747 25.039 make_full_inverse_cholesky 11 9.9 0.000 0.000 23.071 24.037 qs_ot_get_derivative 117 11.6 0.001 0.002 23.122 23.481 qs_ot_get_p 128 10.4 0.001 0.002 20.399 20.877 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.741 20.822 apply_single 128 13.6 0.001 0.001 19.740 20.821 ot_diis_step 117 11.6 0.018 0.018 19.628 19.629 qs_ot_p2m_diag 83 11.4 0.266 0.275 16.104 16.116 make_m2s 5014 13.6 0.079 0.082 15.083 15.711 make_images 5014 14.6 1.139 1.221 14.847 15.470 multiply_cannon_sync_h2d 20056 15.6 13.495 15.305 13.495 15.305 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.990 14.991 sum_up_and_integrate 128 10.3 0.002 0.003 13.896 13.922 integrate_v_rspace 128 11.3 0.003 0.004 13.837 13.862 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.070 13.105 calculate_rho_elec 128 8.7 0.132 0.146 13.070 13.104 cp_fm_cholesky_invert 11 10.9 11.988 12.014 11.988 12.014 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.463 11.482 cp_fm_diag_elpa_base 83 14.4 11.210 11.287 11.460 11.479 init_scf_run 11 5.9 0.000 0.001 10.659 10.660 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.659 10.659 make_images_data 5014 15.6 0.062 0.071 9.160 10.199 hybrid_alltoall_any 5200 16.5 0.452 2.070 7.862 9.666 multiply_cannon_metrocomm4 17549 15.6 0.066 0.077 3.556 9.491 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.155 9.425 mp_irecv_dv 50230 16.2 3.424 9.228 3.424 9.228 dbcsr_mm_accdrv_process 41502 16.2 5.742 6.037 8.893 9.080 fft_wrap_pw1pw2 1291 11.7 0.019 0.024 7.679 7.780 grid_integrate_task_list 128 12.3 7.307 7.759 7.307 7.759 wfi_extrapolate 11 7.9 0.001 0.001 7.413 7.413 density_rs2pw 128 9.7 0.005 0.006 6.535 7.007 fft_wrap_pw1pw2_140 523 12.2 0.281 0.290 6.688 6.802 cp_fm_cholesky_decompose 22 10.9 6.095 6.155 6.095 6.155 fft3d_ps 1291 13.7 2.756 2.976 5.990 6.065 calculate_dm_sparse 128 9.5 0.001 0.001 5.916 5.998 dbcsr_complete_redistribute 395 12.7 1.039 1.100 4.696 5.893 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.592 5.600 mp_alltoall_d11v 2415 14.1 4.822 5.578 4.822 5.578 grid_collocate_task_list 128 9.7 5.162 5.544 5.162 5.544 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.770 5.376 cp_fm_upper_to_full 105 14.8 3.927 4.922 3.927 4.922 mp_sum_l 7950 12.9 3.271 4.872 3.271 4.872 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.581 4.717 mp_allgather_i34 2507 14.6 1.193 4.605 1.193 4.605 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.430 4.601 potential_pw2rs 128 12.3 0.020 0.022 4.323 4.341 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.029 4.072 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.748 3.749 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.493 3.535 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=171.158000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=980.000000, yerr=4.431294 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 1.106817E+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 4014 67561. 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 0.012 0.028 177.334 177.334 qs_mol_dyn_low 1 2.0 0.003 0.004 176.940 176.955 qs_forces 11 3.9 0.004 0.004 176.825 176.834 qs_energies 11 4.9 0.002 0.002 169.698 169.709 scf_env_do_scf 11 5.9 0.001 0.001 153.130 153.145 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.819 115.820 velocity_verlet 10 3.0 0.001 0.002 115.171 115.174 dbcsr_multiply_generic 2485 12.5 0.194 0.200 81.046 81.808 qs_scf_new_mos 116 7.6 0.001 0.001 80.176 80.521 qs_scf_loop_do_ot 116 8.6 0.001 0.001 80.175 80.520 ot_scf_mini 116 9.6 0.003 0.004 75.461 75.803 multiply_cannon 2485 13.5 0.546 0.583 55.581 59.389 multiply_cannon_loop 2485 14.5 1.835 1.914 51.872 53.650 ot_mini 116 10.6 0.001 0.001 43.005 43.337 init_scf_loop 11 6.9 0.000 0.000 37.174 37.175 prepare_preconditioner 11 7.9 0.000 0.000 33.024 33.062 make_preconditioner 11 8.9 0.000 0.000 33.024 33.062 make_full_inverse_cholesky 11 9.9 0.016 0.022 28.284 31.832 multiply_cannon_multrec 29820 15.5 13.435 18.816 26.395 31.409 rebuild_ks_matrix 127 8.3 0.001 0.001 29.265 29.613 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 29.265 29.612 mp_waitall_1 146592 16.7 17.487 27.192 17.487 27.192 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.393 26.696 qs_ot_get_derivative 116 11.6 0.001 0.002 23.344 23.693 make_m2s 4970 13.5 0.093 0.099 21.056 21.974 make_images 4970 14.5 1.992 2.307 20.748 21.673 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 19.145 19.691 apply_single 127 13.6 0.001 0.001 19.145 19.690 ot_diis_step 116 11.6 0.017 0.018 19.535 19.537 qs_ot_get_p 127 10.4 0.010 0.015 18.571 18.914 multiply_cannon_metrocomm3 29820 15.5 0.048 0.050 6.271 14.886 qs_ot_p2m_diag 82 11.4 0.340 0.385 14.342 14.391 sum_up_and_integrate 127 10.3 0.002 0.003 13.592 13.621 integrate_v_rspace 127 11.3 0.003 0.004 13.532 13.564 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.125 13.171 calculate_rho_elec 127 8.7 0.173 0.188 13.124 13.171 dbcsr_mm_accdrv_process 61748 16.2 8.388 9.174 12.531 13.035 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.945 12.947 make_images_data 4970 15.5 0.063 0.069 11.402 12.849 cp_fm_upper_to_full 104 14.8 8.779 12.461 8.779 12.461 hybrid_alltoall_any 5155 16.4 0.527 2.206 10.001 12.224 cp_fm_cholesky_invert 11 10.9 12.156 12.178 12.156 12.178 multiply_cannon_sync_h2d 29820 15.5 10.401 11.238 10.401 11.238 init_scf_run 11 5.9 0.000 0.001 10.697 10.700 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.696 10.699 dbcsr_complete_redistribute 393 12.7 1.408 1.501 7.540 10.316 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.859 10.117 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.682 9.704 cp_fm_diag_elpa_base 82 14.4 9.195 9.338 9.675 9.690 copy_fm_to_dbcsr 208 11.6 0.001 0.002 6.131 8.904 grid_integrate_task_list 127 12.3 7.499 7.877 7.499 7.877 fft_wrap_pw1pw2 1281 11.7 0.018 0.023 7.715 7.765 wfi_extrapolate 11 7.9 0.001 0.001 7.526 7.526 transfer_fm_to_dbcsr 11 9.9 0.003 0.014 4.717 7.419 fft_wrap_pw1pw2_140 519 12.2 0.321 0.330 6.838 6.909 multiply_cannon_metrocomm4 24850 15.5 0.079 0.088 2.800 6.884 mp_alltoall_i22 712 14.1 4.102 6.859 4.102 6.859 mp_irecv_dv 75445 16.2 2.646 6.608 2.646 6.608 density_rs2pw 127 9.7 0.005 0.005 6.207 6.587 calculate_dm_sparse 127 9.5 0.001 0.001 6.429 6.530 cp_fm_cholesky_decompose 22 10.9 6.101 6.166 6.101 6.166 fft3d_ps 1281 13.7 2.911 2.972 5.922 5.963 grid_collocate_task_list 127 9.7 5.239 5.658 5.239 5.658 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.625 5.655 mp_alltoall_d11v 2401 14.1 4.961 5.301 4.961 5.301 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.497 4.626 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.531 4.623 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.401 4.401 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.211 4.304 potential_pw2rs 127 12.3 0.021 0.022 3.942 3.950 mp_allgather_i34 2485 14.5 1.794 3.943 1.794 3.943 dbcsr_special_finalize 12425 15.5 0.066 0.068 3.477 3.643 mp_sum_l 7884 12.9 2.489 3.575 2.489 3.575 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.519 3.553 dbcsr_merge_single_wm 12425 16.5 0.319 0.336 3.387 3.550 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=177.334000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1048.545455, yerr=13.753437 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 5.910120E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.443697E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526746E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592214928 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 750341. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4050 66971. MP_Allreduce 11184 1163. MP_Sync 88 MP_Alltoall 1724 18848066. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.090 167.467 167.468 qs_mol_dyn_low 1 2.0 0.003 0.003 166.795 166.834 qs_forces 11 3.9 0.004 0.004 166.652 166.654 qs_energies 11 4.9 0.008 0.009 159.211 159.221 scf_env_do_scf 11 5.9 0.001 0.001 141.988 141.999 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 110.677 110.679 velocity_verlet 10 3.0 0.002 0.003 108.771 108.778 dbcsr_multiply_generic 2529 12.6 0.191 0.195 73.446 73.888 qs_scf_new_mos 118 7.6 0.001 0.001 73.696 73.826 qs_scf_loop_do_ot 118 8.6 0.001 0.001 73.695 73.826 ot_scf_mini 118 9.6 0.003 0.004 69.225 69.400 multiply_cannon 2529 13.6 0.566 0.590 54.252 57.436 multiply_cannon_loop 2529 14.6 0.812 0.848 51.191 52.128 ot_mini 118 10.6 0.001 0.001 38.305 38.482 mp_waitall_1 126876 16.7 25.790 32.695 25.790 32.695 init_scf_loop 11 6.9 0.000 0.000 31.151 31.152 rebuild_ks_matrix 129 8.3 0.001 0.001 29.599 29.761 qs_ks_build_kohn_sham_matrix 129 9.3 0.035 0.071 29.598 29.760 prepare_preconditioner 11 7.9 0.000 0.000 27.261 27.293 make_preconditioner 11 8.9 0.000 0.000 27.261 27.293 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.028 27.176 make_full_inverse_cholesky 11 9.9 0.022 0.027 25.344 25.604 multiply_cannon_multrec 10116 15.6 10.339 14.993 18.070 22.929 qs_ot_get_derivative 118 11.6 0.002 0.002 20.949 21.116 multiply_cannon_metrocomm3 10116 15.6 0.025 0.026 13.160 20.858 qs_ot_get_p 129 10.4 0.002 0.002 17.951 18.140 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.378 17.569 apply_single 129 13.6 0.001 0.001 17.377 17.569 ot_diis_step 118 11.6 0.020 0.021 17.283 17.283 make_m2s 5058 13.6 0.065 0.069 15.253 16.158 make_images 5058 14.6 2.177 2.593 14.943 15.851 cp_fm_cholesky_invert 11 10.9 15.054 15.067 15.054 15.067 sum_up_and_integrate 129 10.3 0.002 0.002 14.068 14.125 qs_ot_p2m_diag 84 11.4 0.501 0.507 14.074 14.088 integrate_v_rspace 129 11.3 0.004 0.004 14.007 14.066 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.695 13.737 calculate_rho_elec 129 8.7 0.259 0.270 13.695 13.736 cp_dbcsr_syevd 84 12.4 0.005 0.006 12.849 12.850 multiply_cannon_sync_h2d 10116 15.6 10.765 11.280 10.765 11.280 init_scf_run 11 5.9 0.000 0.001 10.396 10.397 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.396 10.397 make_images_data 5058 15.6 0.054 0.066 8.837 10.098 hybrid_alltoall_any 5245 16.5 0.852 3.809 8.579 10.009 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.829 9.841 cp_fm_diag_elpa_base 84 14.4 9.588 9.666 9.825 9.838 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.290 8.420 fft_wrap_pw1pw2 1301 11.7 0.017 0.019 8.232 8.269 grid_integrate_task_list 129 12.3 7.846 8.158 7.846 8.158 dbcsr_mm_accdrv_process 20934 16.1 3.304 4.518 7.379 8.134 multiply_cannon_metrocomm1 10116 15.6 0.030 0.031 4.506 7.900 fft_wrap_pw1pw2_140 527 12.2 0.440 0.443 7.226 7.273 wfi_extrapolate 11 7.9 0.001 0.001 7.239 7.239 cp_fm_cholesky_decompose 22 10.9 6.617 6.674 6.617 6.674 density_rs2pw 129 9.7 0.005 0.005 6.292 6.537 calculate_dm_sparse 129 9.5 0.001 0.001 6.214 6.324 fft3d_ps 1301 13.7 3.140 3.265 6.170 6.212 grid_collocate_task_list 129 9.7 5.622 5.868 5.622 5.868 multiply_cannon_metrocomm4 7587 15.6 0.025 0.028 1.886 5.716 mp_irecv_dv 29102 15.9 1.846 5.620 1.846 5.620 dbcsr_complete_redistribute 397 12.7 2.000 2.026 5.189 5.550 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.289 5.306 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.238 5.239 mp_alltoall_d11v 2429 14.1 4.779 5.185 4.779 5.185 mp_allgather_i34 2529 14.6 1.226 4.804 1.226 4.804 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.141 4.210 potential_pw2rs 129 12.3 0.026 0.027 4.111 4.122 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.592 3.876 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.684 3.727 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.631 3.715 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.295 3.625 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.541 3.607 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.485 3.499 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=167.468000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1347.000000, yerr=30.475027 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410024443904 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 11444712984576 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 92.796586E+12 0.0% 0.0% 100.0% flops max/rank 11.606414E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705502176 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.065311E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+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 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108700816 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 68558. MP_Allreduce 11085 1504. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.058 249.825 249.826 qs_mol_dyn_low 1 2.0 0.003 0.003 249.145 249.168 qs_forces 11 3.9 0.004 0.004 249.042 249.046 qs_energies 11 4.9 0.028 0.029 240.059 240.065 scf_env_do_scf 11 5.9 0.001 0.001 216.970 216.983 velocity_verlet 10 3.0 0.002 0.002 172.492 172.501 scf_env_do_scf_inner_loop 116 6.6 0.004 0.008 133.097 133.099 qs_scf_new_mos 116 7.6 0.001 0.001 89.874 89.955 qs_scf_loop_do_ot 116 8.6 0.001 0.001 89.873 89.954 ot_scf_mini 116 9.6 0.003 0.004 84.880 84.911 dbcsr_multiply_generic 2485 12.5 0.249 0.302 83.909 84.490 init_scf_loop 11 6.9 0.000 0.000 83.535 83.539 prepare_preconditioner 11 7.9 0.000 0.000 78.701 78.718 make_preconditioner 11 8.9 0.000 0.000 78.701 78.718 make_full_inverse_cholesky 11 9.9 0.038 0.040 62.884 75.809 multiply_cannon 2485 13.5 0.666 0.692 59.773 61.171 multiply_cannon_loop 2485 14.5 1.043 1.051 55.639 57.270 cp_fm_upper_to_full 104 14.8 34.313 48.359 34.313 48.359 ot_mini 116 10.6 0.001 0.001 45.143 45.194 rebuild_ks_matrix 127 8.3 0.001 0.001 34.179 34.254 qs_ks_build_kohn_sham_matrix 127 9.3 0.031 0.042 34.178 34.254 mp_waitall_1 102768 16.8 29.440 33.568 29.440 33.568 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.707 31.763 dbcsr_complete_redistribute 393 12.7 3.842 3.918 22.116 30.245 copy_fm_to_dbcsr 208 11.6 0.001 0.001 18.431 26.491 qs_ot_get_derivative 116 11.6 0.002 0.002 25.027 25.061 qs_ot_get_p 127 10.4 0.002 0.003 24.726 24.828 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 15.775 23.746 mp_alltoall_i22 712 14.1 13.469 21.357 13.469 21.357 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 19.516 21.232 make_m2s 4970 13.5 0.074 0.077 19.339 20.935 qs_ot_p2m_diag 82 11.4 0.869 0.874 20.538 20.570 make_images 4970 14.5 3.071 3.261 18.861 20.456 ot_diis_step 116 11.6 0.022 0.022 20.083 20.083 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.463 19.614 apply_single 127 13.6 0.001 0.001 19.463 19.614 cp_fm_cholesky_invert 11 10.9 19.176 19.189 19.176 19.189 cp_dbcsr_syevd 82 12.4 0.006 0.006 18.686 18.688 multiply_cannon_multrec 9940 15.5 10.219 12.024 18.089 18.164 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.068 16.096 calculate_rho_elec 127 8.7 0.478 0.479 16.067 16.096 sum_up_and_integrate 127 10.3 0.002 0.002 15.980 16.067 integrate_v_rspace 127 11.3 0.004 0.004 15.917 16.003 cp_fm_diag_elpa 82 13.4 0.000 0.000 15.442 15.443 cp_fm_diag_elpa_base 82 14.4 12.631 13.567 15.437 15.437 multiply_cannon_sync_h2d 9940 15.5 14.195 14.213 14.195 14.213 hybrid_alltoall_any 5155 16.4 1.301 3.041 11.183 13.589 make_images_data 4970 15.5 0.061 0.066 11.054 13.418 init_scf_run 11 5.9 0.000 0.001 12.581 12.581 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.580 12.581 fft_wrap_pw1pw2 1281 11.7 0.020 0.020 10.348 10.365 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.035 10.060 dbcsr_mm_accdrv_process 20590 16.0 4.248 6.114 7.621 9.471 fft_wrap_pw1pw2_140 519 12.2 0.818 0.820 9.117 9.128 wfi_extrapolate 11 7.9 0.001 0.001 8.978 8.978 grid_integrate_task_list 127 12.3 8.494 8.701 8.494 8.701 qs_energies_init_hamiltonians 11 5.9 0.012 0.012 8.197 8.198 fft3d_ps 1281 13.7 3.941 3.965 7.659 7.708 mp_alltoall_d11v 2401 14.1 7.548 7.693 7.548 7.693 cp_fm_cholesky_decompose 22 10.9 7.612 7.674 7.612 7.674 density_rs2pw 127 9.7 0.005 0.005 7.126 7.183 calculate_dm_sparse 127 9.5 0.001 0.001 6.924 6.969 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.885 6.939 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.388 6.465 grid_collocate_task_list 127 9.7 6.360 6.442 6.360 6.442 qs_env_update_s_mstruct 11 6.9 0.000 0.000 5.007 5.216 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.689 5.105 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=249.826000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2739.909091, yerr=156.707866 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 1.270440E+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 1033 290947. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.031 84.951 84.953 qs_energies 1 2.0 0.000 0.000 84.478 84.492 ls_scf 1 3.0 0.000 0.000 83.595 83.608 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.644 72.836 multiply_cannon 111 7.7 0.017 0.021 55.899 57.004 multiply_cannon_loop 111 8.7 0.228 0.243 52.468 53.735 ls_scf_main 1 4.0 0.000 0.000 52.381 52.382 density_matrix_trs4 2 5.0 0.002 0.003 46.867 46.937 ls_scf_init_scf 1 4.0 0.000 0.001 28.137 28.139 ls_scf_init_matrix_S 1 5.0 0.000 0.000 26.992 27.053 mp_waitall_1 11031 10.9 22.630 26.262 22.630 26.262 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.890 24.906 multiply_cannon_multrec 2664 9.7 8.131 8.914 15.617 17.425 multiply_cannon_sync_h2d 2664 9.7 13.349 14.805 13.349 14.805 make_m2s 222 7.7 0.009 0.011 13.107 13.668 make_images 222 8.7 0.099 0.109 13.085 13.649 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.791 12.361 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.482 8.547 make_images_data 222 9.7 0.004 0.005 7.705 8.256 dbcsr_mm_accdrv_process 4760 10.4 0.590 0.712 7.104 8.089 hybrid_alltoall_any 227 10.6 0.216 1.832 6.611 7.751 dbcsr_mm_accdrv_process_sort 4760 11.4 6.315 7.243 6.315 7.243 calculate_norms 4752 9.8 5.491 6.071 5.491 6.071 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.048 5.186 mp_sum_l 887 5.1 3.094 4.422 3.094 4.422 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.342 3.375 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.081 3.353 mp_irecv_dv 6231 10.9 2.064 3.330 2.064 3.330 make_images_sizes 222 9.7 0.000 0.000 0.722 3.294 mp_alltoall_i44 222 10.7 0.722 3.294 0.722 3.294 arnoldi_extremal 4 6.8 0.000 0.000 3.192 3.219 arnoldi_normal_ev 4 7.8 0.001 0.002 3.192 3.219 build_subspace 16 8.4 0.009 0.012 3.101 3.103 ls_scf_post 1 4.0 0.000 0.000 3.077 3.091 ls_scf_store_result 1 5.0 0.000 0.000 2.857 2.903 dbcsr_special_finalize 555 9.7 0.005 0.006 2.321 2.775 dbcsr_merge_single_wm 555 10.7 0.454 0.612 2.313 2.766 make_images_pack 222 9.7 2.211 2.646 2.213 2.647 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.307 2.565 dbcsr_sort_data 658 11.4 2.118 2.504 2.118 2.504 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.471 2.066 2.473 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.267 2.335 buffer_matrices_ensure_size 222 8.7 1.744 2.084 1.744 2.084 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.797 1.799 rebuild_ks_matrix 3 7.3 0.000 0.000 1.788 1.790 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.788 1.790 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.953000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 2.159264E+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 1033 301899. 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 0.031 0.043 90.192 90.193 qs_energies 1 2.0 0.000 0.000 89.468 89.484 ls_scf 1 3.0 0.000 0.000 88.028 88.044 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.042 74.354 multiply_cannon 111 7.7 0.028 0.042 52.637 56.564 ls_scf_main 1 4.0 0.000 0.000 54.295 54.299 multiply_cannon_loop 111 8.7 0.136 0.147 49.908 52.925 density_matrix_trs4 2 5.0 0.002 0.003 48.525 48.718 ls_scf_init_scf 1 4.0 0.000 0.001 30.009 30.011 mp_waitall_1 9105 10.9 20.522 29.662 20.522 29.662 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.788 28.891 multiply_cannon_multrec 1332 9.7 13.434 17.707 22.843 28.284 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.351 26.364 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.208 20.708 make_m2s 222 7.7 0.006 0.007 15.107 15.659 make_images 222 8.7 1.374 1.686 15.077 15.629 dbcsr_mm_accdrv_process 4041 10.4 0.342 0.554 9.005 10.582 dbcsr_mm_accdrv_process_sort 4041 11.4 8.515 10.060 8.515 10.060 make_images_data 222 9.7 0.004 0.005 8.749 9.600 hybrid_alltoall_any 227 10.6 0.541 2.557 8.117 9.320 mp_sum_l 887 5.1 4.829 7.780 4.829 7.780 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.160 7.778 mp_irecv_dv 3311 11.0 3.139 7.724 3.139 7.724 calculate_norms 2376 9.8 6.065 6.850 6.065 6.850 multiply_cannon_sync_h2d 1332 9.7 4.904 6.428 4.904 6.428 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.580 6.128 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.002 5.223 arnoldi_extremal 4 6.8 0.000 0.000 4.662 4.682 arnoldi_normal_ev 4 7.8 0.001 0.005 4.662 4.682 build_subspace 16 8.4 0.014 0.021 4.413 4.416 ls_scf_post 1 4.0 0.000 0.000 3.723 3.739 ls_scf_store_result 1 5.0 0.000 0.000 3.428 3.538 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.161 3.408 dbcsr_matrix_vector_mult_local 304 10.0 2.744 3.221 2.746 3.223 mp_allgather_i34 111 8.7 0.715 2.941 0.715 2.941 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.654 2.774 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.210 2.577 dbcsr_data_new 4174 10.1 2.115 2.404 2.115 2.404 make_images_pack 222 9.7 1.839 2.133 1.841 2.135 dbcsr_sort_data 436 11.2 1.839 2.110 1.839 2.110 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.903 1.905 rebuild_ks_matrix 3 7.3 0.000 0.000 1.891 1.893 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.026 1.891 1.893 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.193000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1802.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 2.970542E+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 1033 300682. 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 0.034 0.077 96.789 96.800 qs_energies 1 2.0 0.000 0.002 96.002 96.009 ls_scf 1 3.0 0.000 0.001 94.542 94.548 dbcsr_multiply_generic 111 6.7 0.016 0.016 79.092 79.427 multiply_cannon 111 7.7 0.034 0.084 55.416 59.399 multiply_cannon_loop 111 8.7 0.117 0.129 52.636 57.323 ls_scf_main 1 4.0 0.000 0.001 57.001 57.005 density_matrix_trs4 2 5.0 0.002 0.006 51.090 51.300 mp_waitall_1 7281 11.0 24.016 34.394 24.016 34.394 ls_scf_init_scf 1 4.0 0.000 0.002 33.808 33.810 ls_scf_init_matrix_S 1 5.0 0.001 0.009 32.592 32.669 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.123 30.135 multiply_cannon_multrec 888 9.7 12.588 15.305 21.295 24.614 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.244 24.028 make_m2s 222 7.7 0.006 0.007 16.451 17.219 make_images 222 8.7 1.586 1.864 16.413 17.179 make_images_data 222 9.7 0.004 0.004 9.530 10.694 hybrid_alltoall_any 227 10.6 0.646 2.947 9.101 10.285 dbcsr_mm_accdrv_process 3754 10.4 0.325 0.526 8.220 9.489 mp_sum_l 887 5.1 5.601 9.036 5.601 9.036 dbcsr_mm_accdrv_process_sort 3754 11.4 7.775 8.963 7.775 8.963 multiply_cannon_sync_h2d 888 9.7 5.959 7.440 5.959 7.440 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.887 7.267 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.322 7.258 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.518 7.248 mp_irecv_dv 2335 11.1 2.501 7.199 2.501 7.199 arnoldi_extremal 4 6.8 0.000 0.000 5.077 5.091 arnoldi_normal_ev 4 7.8 0.001 0.004 5.076 5.090 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.811 5.022 build_subspace 16 8.4 0.014 0.020 4.775 4.781 calculate_norms 1584 9.8 4.374 4.737 4.374 4.737 compute_matrix_preconditioner 1 6.0 0.000 0.001 4.042 4.050 mp_allgather_i34 111 8.7 0.883 3.809 0.883 3.809 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.425 3.767 ls_scf_post 1 4.0 0.000 0.000 3.734 3.739 dbcsr_matrix_vector_mult_local 304 10.0 3.018 3.602 3.021 3.604 ls_scf_store_result 1 5.0 0.000 0.000 3.422 3.515 acc_transpose_blocks 888 9.7 0.005 0.006 3.101 3.413 acc_transpose_blocks_kernels 888 10.7 0.013 0.014 3.038 3.348 jit_kernel_transpose 1 13.0 3.025 3.334 3.025 3.334 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.818 2.915 dbcsr_data_new 4116 9.9 2.120 2.472 2.120 2.472 make_images_sizes 222 9.7 0.000 0.000 1.087 2.428 mp_alltoall_i44 222 10.7 1.086 2.427 1.086 2.427 dbcsr_sort_data 325 11.1 1.860 2.174 1.860 2.174 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=96.800000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2336.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 3.398320E+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 1033 301899. 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 0.033 0.050 97.033 97.034 qs_energies 1 2.0 0.000 0.002 96.175 96.183 ls_scf 1 3.0 0.000 0.003 94.499 94.507 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.198 78.509 ls_scf_main 1 4.0 0.000 0.006 58.622 58.623 multiply_cannon 111 7.7 0.050 0.138 51.282 56.222 density_matrix_trs4 2 5.0 0.004 0.017 52.475 52.571 multiply_cannon_loop 111 8.7 0.152 0.166 46.219 49.588 ls_scf_init_scf 1 4.0 0.000 0.002 32.576 32.579 ls_scf_init_matrix_S 1 5.0 0.000 0.002 31.369 31.431 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.895 28.906 mp_waitall_1 6369 11.0 22.189 28.585 22.189 28.585 multiply_cannon_multrec 1332 9.7 14.148 17.466 22.181 25.207 make_m2s 222 7.7 0.007 0.008 21.159 22.605 make_images 222 8.7 3.155 3.617 21.109 22.556 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 8.706 17.364 make_images_data 222 9.7 0.004 0.004 11.786 13.528 hybrid_alltoall_any 227 10.6 0.804 3.860 11.065 12.952 dbcsr_mm_accdrv_process 3641 10.4 0.288 0.479 7.665 9.205 dbcsr_mm_accdrv_process_sort 3641 11.4 7.224 8.701 7.224 8.701 mp_sum_l 887 5.1 4.175 7.500 4.175 7.500 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.104 6.069 mp_irecv_dv 3229 10.9 2.080 5.993 2.080 5.993 multiply_cannon_sync_h2d 1332 9.7 5.436 5.967 5.436 5.967 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.171 5.908 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.534 5.811 arnoldi_extremal 4 6.8 0.000 0.000 5.187 5.198 arnoldi_normal_ev 4 7.8 0.001 0.005 5.187 5.198 build_subspace 16 8.4 0.014 0.020 4.856 4.863 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.592 4.782 calculate_norms 2376 9.8 4.162 4.500 4.162 4.500 mp_allgather_i34 111 8.7 2.160 4.485 2.160 4.485 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.574 3.869 dbcsr_matrix_vector_mult_local 304 10.0 3.184 3.682 3.186 3.684 dbcsr_sort_data 658 11.4 3.017 3.551 3.017 3.551 ls_scf_post 1 4.0 0.000 0.002 3.302 3.310 dbcsr_special_finalize 555 9.7 0.006 0.007 2.747 3.233 dbcsr_merge_single_wm 555 10.7 0.532 0.670 2.739 3.225 ls_scf_store_result 1 5.0 0.000 0.000 2.975 3.054 ls_scf_dm_to_ks 2 5.0 0.000 0.003 2.988 3.043 dbcsr_data_release 10477 10.7 1.586 2.458 1.586 2.458 dbcsr_finalize 304 7.8 0.049 0.061 1.823 2.086 dbcsr_merge_all 275 8.9 0.320 0.369 1.704 1.961 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.949 1.952 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.034000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2755.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 4.809695E+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 1033 300769. 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 0.040 0.056 92.913 92.914 qs_energies 1 2.0 0.000 0.000 92.041 92.046 ls_scf 1 3.0 0.000 0.000 90.099 90.104 dbcsr_multiply_generic 111 6.7 0.017 0.019 71.645 71.825 ls_scf_main 1 4.0 0.000 0.000 56.966 56.966 multiply_cannon 111 7.7 0.075 0.136 53.055 56.108 multiply_cannon_loop 111 8.7 0.088 0.093 50.426 52.069 density_matrix_trs4 2 5.0 0.002 0.003 50.010 50.073 ls_scf_init_scf 1 4.0 0.000 0.001 29.772 29.777 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.513 28.554 mp_waitall_1 5436 11.0 24.688 28.499 24.688 28.499 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.182 26.195 multiply_cannon_multrec 444 9.7 13.801 16.247 21.080 22.199 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.851 15.852 make_m2s 222 7.7 0.005 0.005 13.676 14.640 make_images 222 8.7 2.037 2.474 13.609 14.570 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.187 13.987 hybrid_alltoall_any 227 10.6 0.800 3.817 8.272 9.952 make_images_data 222 9.7 0.003 0.004 8.463 9.783 dbcsr_mm_accdrv_process 3003 10.4 0.435 0.604 6.967 8.115 multiply_cannon_sync_h2d 444 9.7 6.732 8.038 6.732 8.038 dbcsr_mm_accdrv_process_sort 3003 11.4 6.517 7.619 6.517 7.619 arnoldi_extremal 4 6.8 0.000 0.000 5.861 5.875 arnoldi_normal_ev 4 7.8 0.001 0.004 5.861 5.875 build_subspace 16 8.4 0.015 0.019 5.460 5.475 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.632 4.830 mp_sum_l 887 5.1 2.907 4.568 2.907 4.568 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.173 4.378 dbcsr_matrix_vector_mult_local 304 10.0 3.695 4.164 3.698 4.166 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.554 3.837 mp_irecv_dv 1241 11.2 1.542 3.795 1.542 3.795 mp_allgather_i34 111 8.7 1.171 3.780 1.171 3.780 calculate_norms 792 9.8 3.595 3.720 3.595 3.720 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.500 3.579 ls_scf_post 1 4.0 0.000 0.000 3.362 3.366 make_images_sizes 222 9.7 0.000 0.000 0.829 3.322 mp_alltoall_i44 222 10.7 0.828 3.322 0.828 3.322 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.979 3.315 ls_scf_store_result 1 5.0 0.000 0.000 3.150 3.186 dbcsr_finalize 304 7.8 0.062 0.077 2.204 2.314 dbcsr_data_new 4608 9.7 1.789 2.217 1.789 2.217 dbcsr_merge_all 275 8.9 0.479 0.524 2.059 2.149 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.071 2.072 rebuild_ks_matrix 3 7.3 0.000 0.000 2.038 2.039 qs_ks_build_kohn_sham_matrix 3 8.3 0.006 0.032 2.038 2.039 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 1.926 1.926 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.914000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3720.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 9.060614E+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 1033 319197. 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 0.076 0.097 107.408 107.408 qs_energies 1 2.0 0.000 0.000 106.074 106.080 ls_scf 1 3.0 0.000 0.000 103.109 103.115 dbcsr_multiply_generic 111 6.7 0.024 0.027 75.904 76.086 ls_scf_main 1 4.0 0.000 0.000 65.170 65.171 density_matrix_trs4 2 5.0 0.002 0.003 56.060 56.136 multiply_cannon 111 7.7 0.123 0.163 49.066 50.905 multiply_cannon_loop 111 8.7 0.098 0.100 46.050 47.041 ls_scf_init_scf 1 4.0 0.001 0.001 34.112 34.112 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.572 32.592 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.857 29.865 mp_waitall_1 4527 11.1 22.489 26.828 22.489 26.828 make_m2s 222 7.7 0.005 0.005 22.772 23.888 make_images 222 8.7 3.575 3.887 22.664 23.780 multiply_cannon_multrec 444 9.7 17.867 18.473 22.584 23.180 hybrid_alltoall_any 227 10.6 1.653 3.605 12.895 15.845 make_images_data 222 9.7 0.003 0.004 13.154 15.290 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.731 11.497 multiply_cannon_sync_h2d 444 9.7 8.791 8.849 8.791 8.849 arnoldi_extremal 4 6.8 0.000 0.000 8.142 8.152 arnoldi_normal_ev 4 7.8 0.002 0.008 8.142 8.151 build_subspace 16 8.4 0.025 0.035 7.583 7.594 dbcsr_matrix_vector_mult 304 9.0 0.016 0.033 5.774 6.090 dbcsr_matrix_vector_mult_local 304 10.0 5.033 5.353 5.035 5.355 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.976 5.244 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.121 5.204 dbcsr_mm_accdrv_process 1814 10.4 0.245 0.360 4.532 4.661 dbcsr_mm_accdrv_process_sort 1814 11.4 4.195 4.322 4.195 4.322 ls_scf_post 1 4.0 0.000 0.000 3.827 3.834 make_images_sizes 222 9.7 0.000 0.000 1.492 3.827 mp_alltoall_i44 222 10.7 1.491 3.826 1.491 3.826 mp_allgather_i34 111 8.7 1.132 3.663 1.132 3.663 ls_scf_store_result 1 5.0 0.000 0.000 3.523 3.561 calculate_norms 792 9.8 3.229 3.275 3.229 3.275 dbcsr_finalize 304 7.8 0.082 0.090 3.079 3.178 dbcsr_merge_all 275 8.9 0.886 0.918 2.864 2.956 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.934 2.934 dbcsr_complete_redistribute 5 7.6 1.430 1.468 2.786 2.885 mp_sum_l 887 5.1 1.665 2.639 1.665 2.639 matrix_ls_to_qs 2 6.0 0.000 0.000 2.432 2.525 dbcsr_sort_data 325 11.1 2.438 2.509 2.438 2.509 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.443 2.445 rebuild_ks_matrix 3 7.3 0.000 0.000 2.377 2.379 qs_ks_build_kohn_sham_matrix 3 8.3 0.010 0.020 2.377 2.379 dbcsr_new_transposed 4 7.5 0.242 0.253 2.337 2.359 dbcsr_data_new 6591 9.6 1.875 2.336 1.875 2.336 dbcsr_frobenius_norm 74 6.6 2.056 2.141 2.203 2.241 dbcsr_add_d 103 6.2 0.000 0.000 2.126 2.202 dbcsr_add_anytype 103 7.2 0.859 0.891 2.126 2.202 dbcsr_data_release 12724 10.6 1.977 2.178 1.977 2.178 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.408000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7118.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ee725ffc68fafd11fbf186e4af07b2db94632368_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 585.166848E+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 77937. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 1406441232. 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 0.037 0.113 242.544 242.546 qs_mol_dyn_low 1 2.0 0.008 0.033 241.046 241.093 qs_forces 5 3.8 0.004 0.012 240.877 240.886 qs_energies 5 4.8 0.005 0.054 237.662 237.694 scf_env_do_scf 5 5.8 0.000 0.003 222.920 222.921 scf_env_do_scf_inner_loop 105 6.6 0.002 0.010 195.714 195.715 qs_scf_new_mos 105 7.6 0.000 0.001 152.626 152.800 qs_scf_loop_do_ot 105 8.6 0.001 0.002 152.626 152.799 ot_scf_mini 105 9.6 0.003 0.005 142.712 142.868 dbcsr_multiply_generic 1445 12.2 0.142 0.169 135.806 136.265 multiply_cannon 1445 13.2 0.274 0.284 115.534 117.968 multiply_cannon_loop 1445 14.2 2.866 3.020 113.583 115.149 velocity_verlet 4 3.0 0.020 0.117 108.450 108.453 ot_mini 105 10.6 0.001 0.001 61.239 61.368 qs_ot_get_p 112 10.4 0.001 0.001 51.181 51.401 multiply_cannon_multrec 69360 15.2 29.560 34.826 40.038 44.887 mp_waitall_1 488190 16.1 36.930 43.771 36.930 43.771 qs_ot_p2m_diag 40 11.0 0.020 0.033 39.784 39.877 qs_ot_get_derivative 55 11.6 0.001 0.002 39.400 39.501 cp_dbcsr_syevd 40 12.0 0.002 0.004 36.498 36.499 multiply_cannon_metrocomm3 69360 15.2 0.199 0.216 26.502 33.419 multiply_cannon_sync_h2d 69360 15.2 28.901 33.230 28.901 33.230 cp_fm_syevd 40 13.0 0.000 0.005 30.988 31.137 rebuild_ks_matrix 110 8.4 0.000 0.003 30.734 30.903 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.038 30.733 30.903 qs_ks_update_qs_env 112 7.6 0.001 0.001 28.257 28.419 init_scf_loop 7 6.6 0.002 0.044 27.157 27.158 cp_fm_redistribute_end 40 14.0 13.081 26.073 13.100 26.080 cp_fm_syevd_base 40 14.0 12.965 25.952 12.965 25.952 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.108 23.415 apply_single 62 13.6 0.000 0.000 23.107 23.415 prepare_preconditioner 7 7.6 0.000 0.000 21.776 21.807 make_preconditioner 7 8.6 0.000 0.002 21.776 21.807 ot_new_cg_direction 55 11.6 0.001 0.001 21.134 21.135 qs_rho_update_rho_low 110 7.6 0.000 0.001 18.845 19.216 calculate_rho_elec 110 8.6 0.028 0.031 18.845 19.216 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.967 15.124 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.051 14.159 mp_sum_l 4764 12.2 12.819 13.814 12.819 13.814 make_full_inverse_cholesky 7 9.6 0.000 0.001 13.692 13.755 density_rs2pw 110 9.6 0.004 0.023 12.897 13.288 fft_wrap_pw1pw2 1425 12.5 0.019 0.027 12.522 12.796 init_scf_run 5 5.8 0.000 0.001 12.214 12.215 scf_env_initial_rho_setup 5 6.8 0.001 0.003 12.214 12.214 calculate_dm_sparse 110 9.5 0.000 0.001 11.598 11.819 fft_wrap_pw1pw2_240 915 14.0 0.196 0.242 11.011 11.235 dbcsr_mm_accdrv_process 154766 15.8 6.276 6.446 10.347 11.198 qs_vxc_create 110 10.4 0.002 0.003 10.847 10.879 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.690 10.771 fft3d_pb 915 15.0 2.225 2.430 10.004 10.220 check_diag 80 13.5 8.650 8.925 9.768 9.913 transfer_rs2pw 445 10.6 0.008 0.010 9.270 9.690 sum_up_and_integrate 60 10.3 0.001 0.005 9.475 9.494 integrate_v_rspace 60 11.3 0.002 0.003 9.458 9.477 cp_fm_cholesky_invert 7 10.6 9.040 9.058 9.040 9.058 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 7.927 7.947 make_m2s 2890 13.2 0.079 0.089 7.324 7.933 make_images 2890 14.2 0.240 0.260 7.215 7.825 xc_rho_set_and_dset_create 110 12.4 0.071 0.080 7.566 7.813 make_full_single_inverse 7 9.6 0.001 0.004 7.752 7.783 acc_transpose_blocks 69360 15.2 0.350 0.380 7.206 7.712 multiply_cannon_metrocomm1 69360 15.2 0.096 0.103 4.962 7.527 mp_alltoall_z22v 2340 16.7 7.266 7.502 7.266 7.502 calculate_first_density_matrix 1 7.0 0.000 0.006 7.451 7.470 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.010 7.346 7.393 xc_vxc_pw_create 60 11.3 0.036 0.040 7.291 7.327 xc_pw_derive 510 13.4 0.005 0.006 6.625 6.700 potential_pw2rs 60 12.3 0.002 0.003 5.550 5.584 mp_waitany 7680 13.5 4.757 5.471 4.757 5.471 multiply_cannon_metrocomm4 67915 15.2 0.183 0.209 2.035 4.920 mp_alltoall_d11v 1300 13.8 4.111 4.916 4.111 4.916 acc_transpose_blocks_kernels 69360 16.2 0.860 0.897 4.452 4.890 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=242.546000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=557.200000, yerr=1.939072 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: ee725ffc68fafd11fbf186e4af07b2db94632368 Summary: empty Status: OK