=== 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: ac0226eb549c7ef1ea50d0597d545f29d4c8fc87 ################# 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 1.16.2, LIBGRPP 20231215 # # 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: 25.01.2024 # # \ 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; \ ./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; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 #USE_DEEPMD := 2.2.7 #USE_SIRIUS := 7.5.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS #LIBVDWXC_VER := 0.4.0 #SPFFT_VER := 1.0.6 #SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_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 SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/01 job id: 51730258 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/02 job id: 51730259 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/03 job id: 51730260 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/04 job id: 51730261 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/05 job id: 51730262 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/06 job id: 51730263 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/07 job id: 51730264 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/08 job id: 51730265 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/09 job id: 51730266 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/10 job id: 51730267 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/11 job id: 51730268 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/12 job id: 51730269 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/13 job id: 51730270 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/14 job id: 51730271 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/15 job id: 51730272 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/16 job id: 51730273 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/17 job id: 51730276 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/18 job id: 51730277 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/19 job id: 51730278 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/20 job id: 51730279 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/21 job id: 51730280 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/22 job id: 51730281 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/23 job id: 51730282 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/24 job id: 51730284 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/25 job id: 51730287 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/26 job id: 51730288 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/27 job id: 51730289 --- 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/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 0.018 0.033 134.117 134.117 farming_run 1 2.0 133.539 133.540 134.088 134.091 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.481052E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.005 0.026 114.623 114.624 qs_energies 1 2.0 0.001 0.001 114.436 114.438 mp2_main 1 3.0 0.000 0.000 112.337 112.339 mp2_gpw_main 1 4.0 0.030 0.037 111.332 111.334 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 92.582 92.669 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.550 54.636 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.160 40.951 45.705 get_2c_integrals 1 6.0 0.008 0.009 37.314 37.860 integrate_v_rspace 273 8.0 0.443 0.458 24.556 28.999 pw_transfer 6555 10.6 0.378 0.393 26.771 27.081 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 25.393 25.681 grid_integrate_task_list 273 9.0 20.463 25.327 20.463 25.327 fft_wrap_pw1pw2_100 2178 12.4 0.028 0.030 22.919 23.263 compute_2c_integrals 1 7.0 0.002 0.002 19.025 19.028 cp_fm_cholesky_decompose 12 8.2 18.274 18.876 18.274 18.876 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.571 18.751 mp2_eri_2c_integrate_gpw 1 9.0 2.394 2.431 18.569 18.750 rpa_ri_compute_en 1 5.0 0.021 0.050 18.644 18.738 cholesky_decomp 1 7.0 0.000 0.000 17.128 17.718 fft3d_s 5443 13.4 16.121 16.395 16.143 16.415 ao_to_mo_and_store_B_mult_1 272 7.0 10.771 15.320 10.771 15.320 calculate_wavefunction 272 8.0 5.373 5.450 12.300 12.989 rpa_num_int 1 6.0 0.000 0.004 10.736 10.746 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.669 10.704 calc_mat_Q 8 8.0 0.000 0.000 9.475 9.575 contract_S_to_Q 8 9.0 0.000 0.000 8.900 8.999 calc_potential_gpw 544 9.5 0.004 0.005 8.227 8.673 parallel_gemm_fm 14 9.1 0.000 0.000 8.505 8.596 parallel_gemm_fm_cosma 14 10.1 8.505 8.596 8.505 8.596 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.048 8.271 potential_pw2rs 545 10.0 0.106 0.108 7.463 8.136 create_integ_mat 1 6.0 0.015 0.028 7.701 7.710 collocate_single_gaussian 272 10.0 0.039 0.042 7.323 7.601 array2fm 1 7.0 0.000 0.000 6.597 7.130 pw_scatter_s 2720 13.7 4.344 4.475 4.344 4.475 pw_gather_s 2722 13.2 3.406 3.837 3.406 3.837 array2fm_buffer_send 1 8.0 2.925 3.093 2.925 3.093 pw_poisson_solve 545 10.5 1.124 1.173 2.208 2.333 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.331152, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2796.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 0.030 0.046 521.352 521.354 farming_run 1 2.0 520.504 520.508 521.309 521.314 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.234489E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 2188433. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.031 318.148 318.149 qs_energies 1 2.0 0.000 0.000 317.930 317.931 mp2_main 1 3.0 0.000 0.000 203.381 203.382 mp2_gpw_main 1 4.0 0.047 0.070 202.540 202.545 mp2_ri_gpw_compute_en 1 5.0 0.038 0.042 128.552 130.529 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.767 1.963 126.866 126.870 scf_env_do_scf 1 3.0 0.000 0.000 114.260 114.260 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.345 113.352 rebuild_ks_matrix 4 6.0 0.000 0.000 113.344 113.351 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.062 113.344 113.351 hfx_ks_matrix 4 8.0 0.001 0.001 112.993 112.997 integrate_four_center 4 9.0 0.153 0.474 112.992 112.996 mp2_ri_gpw_compute_en_expansio 172 7.0 0.466 0.495 111.288 111.979 local_gemm 172 8.0 110.822 111.502 110.822 111.502 integrate_four_center_main 4 10.0 0.108 0.494 101.790 104.305 integrate_four_center_bin 263 11.0 101.682 104.249 101.682 104.249 init_scf_loop 1 4.0 0.000 0.000 96.410 96.410 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 73.840 74.915 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.708 54.783 mp2_eri_3c_integrate_gpw 91 7.0 0.142 0.161 41.025 46.329 integrate_v_rspace 95 8.0 0.398 0.568 27.517 32.612 pw_transfer 2240 10.6 0.144 0.168 29.516 29.940 fft_wrap_pw1pw2 1868 11.4 0.017 0.021 28.528 28.992 grid_integrate_task_list 95 9.0 22.928 28.221 22.928 28.221 fft_wrap_pw1pw2_100 730 12.4 0.013 0.014 26.284 26.674 ao_to_mo_and_store_B_mult_1 91 7.0 11.020 26.194 11.020 26.194 get_2c_integrals 1 6.0 0.000 0.000 20.044 20.067 compute_2c_integrals 1 7.0 0.002 0.003 19.026 19.028 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.636 18.906 mp2_eri_2c_integrate_gpw 1 9.0 1.744 1.848 18.635 18.905 fft3d_s 1823 13.4 18.491 18.809 18.505 18.823 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.848 17.848 calculate_wavefunction 91 8.0 2.003 2.030 9.609 9.815 mp2_ri_gpw_compute_en_comm 22 7.0 0.497 0.513 7.710 9.269 potential_pw2rs 186 10.0 0.033 0.034 8.427 9.047 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.000 0.001 8.149 8.558 calc_potential_gpw 182 9.5 0.002 0.002 7.897 8.117 collocate_single_gaussian 91 10.0 0.016 0.018 7.763 7.967 mp_sendrecv_dm3 2068 8.0 5.748 7.322 5.748 7.322 integrate_four_center_load 4 10.0 0.000 0.000 6.788 6.792 hfx_load_balance 1 11.0 0.000 0.000 6.788 6.792 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=202.534442, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1491.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 445.759488E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. 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 0.015 0.047 54.229 54.229 qs_mol_dyn_low 1 2.0 0.005 0.024 53.908 53.915 qs_forces 11 3.9 0.002 0.003 53.801 53.803 qs_energies 11 4.9 0.001 0.001 52.323 52.339 scf_env_do_scf 11 5.9 0.000 0.001 46.351 46.351 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 44.264 44.265 qs_scf_new_mos 108 7.5 0.000 0.001 34.275 34.595 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.275 34.594 dbcsr_multiply_generic 2286 12.5 0.095 0.099 33.716 34.139 ot_scf_mini 108 9.5 0.002 0.002 32.589 32.833 multiply_cannon 2286 13.5 0.186 0.192 26.156 27.977 multiply_cannon_loop 2286 14.5 1.811 1.907 25.496 27.328 velocity_verlet 10 3.0 0.003 0.019 26.471 26.472 ot_mini 108 10.5 0.001 0.001 19.479 19.743 qs_ot_get_derivative 108 11.5 0.001 0.001 16.398 16.598 mp_waitall_1 245248 16.5 8.557 14.417 8.557 14.417 multiply_cannon_metrocomm3 54864 15.5 0.073 0.079 6.126 12.707 multiply_cannon_multrec 54864 15.5 3.625 5.598 7.539 10.890 qs_ot_get_p 119 10.4 0.001 0.001 8.327 8.592 rebuild_ks_matrix 119 8.3 0.000 0.000 7.823 8.007 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.822 8.007 mp_sum_l 7287 12.8 5.636 7.327 5.636 7.327 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.894 7.048 multiply_cannon_sync_h2d 54864 15.5 5.104 6.263 5.104 6.263 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.773 6.224 dbcsr_mm_accdrv_process 76910 16.1 1.832 2.888 3.829 5.486 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.331 5.433 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.337 5.408 init_scf_run 11 5.9 0.000 0.001 4.672 4.672 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.672 4.672 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.519 4.519 sum_up_and_integrate 119 10.3 0.001 0.002 4.425 4.433 integrate_v_rspace 119 11.3 0.002 0.003 4.414 4.423 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.324 4.325 cp_fm_redistribute_end 50 14.0 2.206 4.301 2.212 4.304 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.121 4.252 calculate_rho_elec 119 8.7 0.011 0.016 4.121 4.252 cp_fm_diag_elpa_base 50 14.0 2.086 4.200 2.090 4.210 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.940 3.145 apply_single 119 13.6 0.000 0.000 2.940 3.145 calculate_dm_sparse 119 9.5 0.000 0.000 2.899 3.026 multiply_cannon_metrocomm1 54864 15.5 0.056 0.062 1.799 2.856 ot_diis_step 108 11.5 0.006 0.006 2.799 2.800 acc_transpose_blocks 54864 15.5 0.221 0.243 2.201 2.713 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.572 2.649 jit_kernel_multiply 13 15.8 1.934 2.509 1.934 2.509 calculate_first_density_matrix 1 7.0 0.000 0.000 2.416 2.420 density_rs2pw 119 9.7 0.004 0.004 2.125 2.243 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.235 2.237 wfi_extrapolate 11 7.9 0.001 0.001 2.187 2.187 grid_integrate_task_list 119 12.3 2.011 2.090 2.011 2.090 init_scf_loop 11 6.9 0.000 0.000 2.069 2.069 mp_sum_d 4137 12.0 1.422 2.023 1.422 2.023 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.923 1.988 potential_pw2rs 119 12.3 0.004 0.004 1.802 1.812 pw_transfer 1439 11.6 0.051 0.055 1.669 1.743 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.594 1.670 make_m2s 4572 13.5 0.053 0.055 1.615 1.662 make_images 4572 14.5 0.132 0.138 1.534 1.580 mp_alltoall_d11v 2130 13.8 1.323 1.497 1.323 1.497 grid_collocate_task_list 119 9.7 1.354 1.435 1.354 1.435 transfer_rs2pw 487 10.6 0.005 0.006 1.347 1.432 acc_transpose_blocks_sync 164592 16.5 1.193 1.430 1.193 1.430 mp_waitany 12084 13.8 1.233 1.412 1.233 1.412 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.380 1.403 fft3d_ps 1201 14.6 0.371 0.476 1.306 1.379 fft_wrap_pw1pw2_140 487 13.2 0.109 0.115 1.242 1.318 transfer_pw2rs 487 13.2 0.006 0.006 1.309 1.316 dbcsr_dot_sd 1205 11.9 0.049 0.058 0.805 1.202 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.229000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=425.363636, yerr=0.979121 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 480.071680E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1032117. 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 0.073 0.105 39.978 39.978 qs_mol_dyn_low 1 2.0 0.003 0.003 39.562 39.570 qs_forces 11 3.9 0.022 0.024 39.270 39.271 qs_energies 11 4.9 0.001 0.002 37.567 37.572 scf_env_do_scf 11 5.9 0.001 0.001 32.490 32.490 scf_env_do_scf_inner_loop 108 6.5 0.013 0.022 29.908 29.908 dbcsr_multiply_generic 2286 12.5 0.104 0.107 22.164 22.537 qs_scf_new_mos 108 7.5 0.001 0.001 21.016 21.235 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.016 21.235 ot_scf_mini 108 9.5 0.002 0.003 20.088 20.243 velocity_verlet 10 3.0 0.001 0.002 18.964 18.970 multiply_cannon 2286 13.5 0.208 0.216 16.920 18.407 multiply_cannon_loop 2286 14.5 1.188 1.247 15.727 17.260 ot_mini 108 10.5 0.001 0.001 12.349 12.573 mp_waitall_1 200699 16.5 5.661 10.754 5.661 10.754 qs_ot_get_derivative 108 11.5 0.001 0.001 9.980 10.137 multiply_cannon_metrocomm3 27432 15.5 0.072 0.074 4.067 9.404 multiply_cannon_multrec 27432 15.5 1.830 4.202 6.157 9.055 rebuild_ks_matrix 119 8.3 0.000 0.000 6.908 7.050 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.020 6.907 7.049 dbcsr_mm_accdrv_process 47894 16.0 3.515 5.612 4.248 6.331 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.086 6.215 qs_ot_get_p 119 10.4 0.001 0.001 4.827 5.042 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.742 4.584 mp_sum_l 7287 12.8 2.115 4.197 2.115 4.197 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.999 4.046 apply_single 119 13.6 0.000 0.000 2.999 4.046 init_scf_run 11 5.9 0.000 0.001 3.863 3.863 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.862 3.863 sum_up_and_integrate 119 10.3 0.001 0.001 3.820 3.826 integrate_v_rspace 119 11.3 0.002 0.003 3.806 3.812 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.777 3.805 calculate_rho_elec 119 8.7 0.021 0.024 3.776 3.805 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.102 3.123 make_m2s 4572 13.5 0.051 0.053 2.674 3.012 make_images 4572 14.5 0.206 0.245 2.585 2.926 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.695 2.696 init_scf_loop 11 6.9 0.000 0.000 2.555 2.555 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.406 2.488 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.352 2.353 cp_fm_redistribute_end 50 14.0 1.192 2.323 1.196 2.325 ot_diis_step 108 11.5 0.011 0.011 2.317 2.318 calculate_first_density_matrix 1 7.0 0.000 0.001 2.300 2.302 cp_fm_diag_elpa_base 50 14.0 1.096 2.224 1.124 2.264 calculate_dm_sparse 119 9.5 0.000 0.001 2.072 2.151 multiply_cannon_sync_h2d 27432 15.5 1.672 2.148 1.672 2.148 density_rs2pw 119 9.7 0.004 0.004 2.071 2.134 pw_transfer 1439 11.6 0.064 0.069 1.923 1.969 acc_transpose_blocks 27432 15.5 0.114 0.120 1.617 1.946 grid_integrate_task_list 119 12.3 1.846 1.944 1.846 1.944 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.879 1.882 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.832 1.880 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.658 1.697 prepare_preconditioner 11 7.9 0.000 0.000 1.612 1.639 make_preconditioner 11 8.9 0.000 0.000 1.612 1.638 jit_kernel_multiply 8 16.5 0.673 1.620 0.673 1.620 make_images_data 4572 15.5 0.047 0.054 1.246 1.583 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.512 1.567 potential_pw2rs 119 12.3 0.006 0.006 1.555 1.565 fft3d_ps 1201 14.6 0.523 0.580 1.464 1.510 wfi_extrapolate 11 7.9 0.001 0.001 1.503 1.503 hybrid_alltoall_any 4725 16.4 0.053 0.116 1.120 1.491 fft_wrap_pw1pw2_140 487 13.2 0.127 0.131 1.420 1.466 mp_alltoall_d11v 2130 13.8 1.262 1.374 1.262 1.374 grid_collocate_task_list 119 9.7 1.290 1.368 1.290 1.368 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.227 1.275 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.255 1.263 transfer_rs2pw 487 10.6 0.005 0.005 1.197 1.256 mp_sum_d 4137 12.0 0.570 1.047 0.570 1.047 mp_allgather_i34 2286 14.5 0.615 1.011 0.615 1.011 transfer_pw2rs 487 13.2 0.004 0.005 0.989 0.997 acc_transpose_blocks_sync 82296 16.5 0.818 0.939 0.818 0.939 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.934 0.937 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.879 0.892 acc_transpose_blocks_kernels 27432 16.5 0.188 0.275 0.658 0.864 make_images_sizes 4572 15.5 0.005 0.005 0.564 0.801 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.978000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=457.181818, yerr=1.748671 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 512.270336E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.052 35.391 35.392 qs_mol_dyn_low 1 2.0 0.005 0.016 35.019 35.027 qs_forces 11 3.9 0.007 0.010 34.874 34.878 qs_energies 11 4.9 0.003 0.014 33.257 33.267 scf_env_do_scf 11 5.9 0.001 0.002 28.219 28.219 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 25.370 25.371 dbcsr_multiply_generic 2286 12.5 0.097 0.100 17.660 17.727 velocity_verlet 10 3.0 0.002 0.005 16.996 17.000 qs_scf_new_mos 108 7.5 0.001 0.001 16.520 16.535 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.519 16.534 ot_scf_mini 108 9.5 0.002 0.003 15.740 15.753 multiply_cannon 2286 13.5 0.194 0.199 13.826 14.451 multiply_cannon_loop 2286 14.5 0.857 0.888 12.956 13.621 ot_mini 108 10.5 0.001 0.001 9.623 9.636 qs_ot_get_derivative 108 11.5 0.001 0.001 8.102 8.115 multiply_cannon_multrec 18288 15.5 1.868 2.884 6.941 7.303 rebuild_ks_matrix 119 8.3 0.000 0.000 6.656 6.687 qs_ks_build_kohn_sham_matrix 119 9.3 0.016 0.041 6.655 6.687 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.873 5.900 dbcsr_mm_accdrv_process 38222 16.0 4.915 5.755 4.980 5.821 sum_up_and_integrate 119 10.3 0.001 0.003 4.150 4.156 integrate_v_rspace 119 11.3 0.003 0.003 4.138 4.145 mp_waitall_1 158411 16.6 3.346 4.062 3.346 4.062 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.991 3.995 calculate_rho_elec 119 8.7 0.033 0.053 3.991 3.995 qs_ot_get_p 119 10.4 0.001 0.002 3.805 3.825 init_scf_run 11 5.9 0.000 0.001 3.698 3.698 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.698 3.698 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.072 3.695 init_scf_loop 11 6.9 0.001 0.006 2.827 2.828 density_rs2pw 119 9.7 0.004 0.004 2.492 2.583 qs_ot_p2m_diag 50 11.0 0.012 0.014 2.512 2.517 make_m2s 4572 13.5 0.043 0.044 2.284 2.469 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.051 2.426 apply_single 119 13.6 0.000 0.000 2.051 2.426 make_images 4572 14.5 0.192 0.204 2.198 2.382 calculate_first_density_matrix 1 7.0 0.001 0.003 2.280 2.282 multiply_cannon_metrocomm3 18288 15.5 0.048 0.049 1.468 2.248 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.201 2.202 pw_transfer 1439 11.6 0.064 0.068 2.106 2.138 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.014 2.047 prepare_preconditioner 11 7.9 0.000 0.000 1.963 1.968 make_preconditioner 11 8.9 0.000 0.001 1.963 1.968 potential_pw2rs 119 12.3 0.007 0.008 1.905 1.927 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.904 1.914 cp_fm_diag_elpa_base 50 14.0 1.878 1.895 1.903 1.912 grid_integrate_task_list 119 12.3 1.805 1.903 1.805 1.903 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.797 1.896 calculate_dm_sparse 119 9.5 0.000 0.001 1.871 1.881 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.858 1.866 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.711 1.713 acc_transpose_blocks 18288 15.5 0.079 0.080 1.630 1.672 transfer_rs2pw 487 10.6 0.005 0.005 1.541 1.651 fft3d_ps 1201 14.6 0.549 0.569 1.570 1.605 fft_wrap_pw1pw2_140 487 13.2 0.178 0.183 1.532 1.564 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.533 1.540 mp_sum_l 7287 12.8 1.180 1.538 1.180 1.538 ot_diis_step 108 11.5 0.011 0.011 1.502 1.503 wfi_extrapolate 11 7.9 0.001 0.002 1.316 1.316 grid_collocate_task_list 119 9.7 1.235 1.304 1.235 1.304 transfer_pw2rs 487 13.2 0.004 0.004 1.279 1.301 make_images_data 4572 15.5 0.047 0.051 0.958 1.214 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.198 1.202 multiply_cannon_sync_h2d 18288 15.5 1.051 1.192 1.051 1.192 hybrid_alltoall_any 4725 16.4 0.058 0.118 0.800 1.115 mp_alltoall_d11v 2130 13.8 0.866 1.017 0.866 1.017 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.002 1.011 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.988 1.005 mp_alltoall_z22v 1201 16.6 0.848 0.925 0.848 0.925 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.857 0.859 cp_fm_cholesky_invert 11 10.9 0.851 0.856 0.851 0.856 acc_transpose_blocks_kernels 18288 16.5 0.217 0.223 0.800 0.828 make_images_sizes 4572 15.5 0.005 0.005 0.572 0.793 mp_alltoall_i44 4572 16.5 0.567 0.789 0.567 0.789 acc_transpose_blocks_sync 54864 16.5 0.731 0.784 0.731 0.784 mp_sendrecv_dv 11067 12.7 0.763 0.781 0.763 0.781 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.667 0.716 mp_waitany 9880 13.7 0.585 0.714 0.585 0.714 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=35.392000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=488.090909, yerr=0.792527 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 553.537536E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.062 36.563 36.564 qs_mol_dyn_low 1 2.0 0.003 0.003 36.115 36.189 qs_forces 11 3.9 0.010 0.013 36.053 36.054 qs_energies 11 4.9 0.001 0.001 34.338 34.344 scf_env_do_scf 11 5.9 0.001 0.001 29.334 29.336 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.937 25.938 dbcsr_multiply_generic 2286 12.5 0.109 0.119 19.825 19.957 velocity_verlet 10 3.0 0.001 0.001 18.450 18.451 qs_scf_new_mos 108 7.5 0.001 0.001 17.851 17.915 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.850 17.914 ot_scf_mini 108 9.5 0.002 0.003 16.815 16.867 multiply_cannon 2286 13.5 0.218 0.224 16.072 16.543 multiply_cannon_loop 2286 14.5 1.540 1.607 15.128 15.500 ot_mini 108 10.5 0.001 0.001 10.341 10.403 multiply_cannon_multrec 27432 15.5 2.446 3.112 9.017 9.290 qs_ot_get_derivative 108 11.5 0.001 0.001 8.467 8.521 dbcsr_mm_accdrv_process 47916 15.9 6.152 7.683 6.468 7.810 rebuild_ks_matrix 119 8.3 0.000 0.000 6.349 6.394 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.348 6.394 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.629 5.668 qs_ot_get_p 119 10.4 0.001 0.002 3.618 3.695 init_scf_run 11 5.9 0.000 0.001 3.604 3.604 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.603 3.604 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.076 3.535 sum_up_and_integrate 119 10.3 0.001 0.001 3.494 3.503 integrate_v_rspace 119 11.3 0.003 0.003 3.482 3.492 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.355 3.391 calculate_rho_elec 119 8.7 0.040 0.046 3.354 3.390 init_scf_loop 11 6.9 0.000 0.000 3.378 3.379 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.177 2.625 apply_single 119 13.6 0.000 0.000 2.177 2.625 prepare_preconditioner 11 7.9 0.000 0.000 2.533 2.541 make_preconditioner 11 8.9 0.000 0.000 2.533 2.541 acc_transpose_blocks 27432 15.5 0.119 0.121 2.425 2.524 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.141 2.468 make_m2s 4572 13.5 0.053 0.056 2.294 2.395 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.341 2.368 make_images 4572 14.5 0.273 0.331 2.188 2.286 mp_waitall_1 137007 16.6 1.700 2.253 1.700 2.253 calculate_first_density_matrix 1 7.0 0.000 0.000 2.230 2.231 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.201 2.212 calculate_dm_sparse 119 9.5 0.000 0.000 2.139 2.203 pw_transfer 1439 11.6 0.064 0.068 1.989 2.024 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.897 1.936 grid_integrate_task_list 119 12.3 1.835 1.924 1.835 1.924 density_rs2pw 119 9.7 0.004 0.004 1.818 1.921 ot_diis_step 108 11.5 0.012 0.012 1.833 1.834 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.809 1.810 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.798 1.799 mp_sum_l 7287 12.8 1.016 1.684 1.016 1.684 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.637 1.648 fft_wrap_pw1pw2_140 487 13.2 0.227 0.234 1.573 1.612 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.555 1.567 cp_fm_diag_elpa_base 50 14.0 1.522 1.540 1.553 1.565 acc_transpose_blocks_sync 82296 16.5 1.447 1.535 1.447 1.535 fft3d_ps 1201 14.6 0.592 0.653 1.386 1.411 grid_collocate_task_list 119 9.7 1.247 1.337 1.247 1.337 wfi_extrapolate 11 7.9 0.001 0.001 1.321 1.321 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.299 1.317 potential_pw2rs 119 12.3 0.009 0.009 1.292 1.296 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.231 1.239 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.709 1.221 cp_fm_upper_to_full 72 14.2 0.817 1.162 0.817 1.162 qs_energies_init_hamiltonians 11 5.9 0.000 0.000 1.144 1.146 dbcsr_complete_redistribute 329 12.2 0.119 0.143 0.759 1.032 transfer_rs2pw 487 10.6 0.004 0.005 0.908 1.014 jit_kernel_multiply 5 15.9 0.250 0.989 0.250 0.989 make_images_data 4572 15.5 0.048 0.053 0.849 0.963 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.729 0.902 mp_alltoall_d11v 2130 13.8 0.727 0.885 0.727 0.885 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.801 0.879 acc_transpose_blocks_kernels 27432 16.5 0.270 0.279 0.828 0.850 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.570 0.839 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.809 0.813 cp_fm_cholesky_invert 11 10.9 0.750 0.753 0.750 0.753 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.564000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=524.545455, yerr=4.097974 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/07/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 117.977176E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1384136 0.0% 0.0% 100.0% average stack size 0.0 0.0 106.2 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 621.850624E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.027 29.417 29.417 qs_mol_dyn_low 1 2.0 0.003 0.003 29.212 29.219 qs_forces 11 3.9 0.002 0.002 29.155 29.156 qs_energies 11 4.9 0.003 0.003 27.432 27.437 scf_env_do_scf 11 5.9 0.000 0.001 22.342 22.343 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.871 19.872 velocity_verlet 10 3.0 0.006 0.008 14.806 14.809 dbcsr_multiply_generic 2286 12.5 0.096 0.099 13.462 13.572 qs_scf_new_mos 108 7.5 0.001 0.001 12.132 12.155 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.132 12.155 ot_scf_mini 108 9.5 0.002 0.002 11.419 11.443 multiply_cannon 2286 13.5 0.224 0.229 10.820 11.368 multiply_cannon_loop 2286 14.5 0.642 0.658 9.891 10.085 ot_mini 108 10.5 0.001 0.001 6.599 6.630 multiply_cannon_multrec 9144 15.5 1.763 1.948 6.182 6.367 rebuild_ks_matrix 119 8.3 0.000 0.000 5.832 5.857 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.832 5.857 qs_ot_get_derivative 108 11.5 0.001 0.001 5.303 5.328 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.184 5.205 dbcsr_mm_accdrv_process 12550 15.8 3.623 4.365 4.311 4.408 init_scf_run 11 5.9 0.000 0.001 3.627 3.627 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.627 3.627 sum_up_and_integrate 119 10.3 0.001 0.001 3.468 3.473 integrate_v_rspace 119 11.3 0.003 0.003 3.458 3.463 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.398 3.404 calculate_rho_elec 119 8.7 0.060 0.061 3.398 3.403 qs_ot_get_p 119 10.4 0.001 0.002 2.928 2.964 calculate_first_density_matrix 1 7.0 0.000 0.000 2.472 2.472 init_scf_loop 11 6.9 0.000 0.000 2.450 2.451 pw_transfer 1439 11.6 0.065 0.067 2.143 2.154 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.050 2.062 make_m2s 4572 13.5 0.033 0.034 1.837 2.000 mp_waitall_1 115863 16.7 1.418 1.991 1.418 1.991 grid_integrate_task_list 119 12.3 1.858 1.941 1.858 1.941 make_images 4572 14.5 0.269 0.300 1.748 1.908 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.901 1.903 density_rs2pw 119 9.7 0.003 0.003 1.792 1.870 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.796 1.815 calculate_dm_sparse 119 9.5 0.000 0.000 1.797 1.813 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.719 1.720 prepare_preconditioner 11 7.9 0.000 0.000 1.710 1.715 make_preconditioner 11 8.9 0.000 0.000 1.710 1.715 fft_wrap_pw1pw2_140 487 13.2 0.324 0.331 1.689 1.700 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.656 1.656 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.600 1.630 acc_transpose_blocks 9144 15.5 0.043 0.044 1.564 1.587 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.527 1.539 jit_kernel_multiply 8 15.6 0.648 1.471 0.648 1.471 fft3d_ps 1201 14.6 0.650 0.664 1.404 1.414 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.378 1.390 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.380 1.389 cp_fm_diag_elpa_base 50 14.0 1.352 1.370 1.378 1.387 grid_collocate_task_list 119 9.7 1.295 1.360 1.295 1.360 ot_diis_step 108 11.5 0.013 0.013 1.280 1.281 potential_pw2rs 119 12.3 0.010 0.010 1.274 1.275 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.213 1.248 apply_single 119 13.6 0.000 0.000 1.213 1.248 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.230 1.231 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.189 1.196 wfi_extrapolate 11 7.9 0.001 0.001 1.104 1.104 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.758 0.997 make_images_data 4572 15.5 0.042 0.046 0.788 0.979 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.877 0.926 mp_alltoall_d11v 2130 13.8 0.789 0.902 0.789 0.902 cp_fm_cholesky_invert 11 10.9 0.851 0.855 0.851 0.855 transfer_rs2pw 487 10.6 0.004 0.005 0.783 0.851 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.826 0.832 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 0.788 0.799 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.770 0.774 acc_transpose_blocks_sync 27432 16.5 0.718 0.743 0.718 0.743 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.677 0.726 jit_kernel_transpose 5 15.6 0.670 0.683 0.670 0.683 mp_allgather_i34 2286 14.5 0.244 0.672 0.244 0.672 transfer_pw2rs 487 13.2 0.003 0.004 0.662 0.664 mp_alltoall_z22v 1201 16.6 0.628 0.661 0.628 0.661 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.368 0.625 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.417000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=587.727273, yerr=8.001033 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 769.835008E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.030 42.906 42.907 qs_mol_dyn_low 1 2.0 0.003 0.003 42.710 42.718 qs_forces 11 3.9 0.002 0.002 42.646 42.648 qs_energies 11 4.9 0.001 0.002 40.635 40.640 scf_env_do_scf 11 5.9 0.001 0.001 34.720 34.721 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.859 26.861 velocity_verlet 10 3.0 0.002 0.004 24.214 24.219 dbcsr_multiply_generic 2286 12.5 0.104 0.105 19.009 19.212 qs_scf_new_mos 108 7.5 0.001 0.001 17.171 17.267 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.170 17.266 ot_scf_mini 108 9.5 0.002 0.002 16.001 16.102 multiply_cannon 2286 13.5 0.308 0.314 15.054 15.990 multiply_cannon_loop 2286 14.5 0.862 0.879 13.794 14.734 ot_mini 108 10.5 0.001 0.001 9.738 9.856 multiply_cannon_multrec 9144 15.5 3.377 4.663 8.848 8.949 init_scf_loop 11 6.9 0.000 0.000 7.835 7.838 qs_ot_get_derivative 108 11.5 0.001 0.001 7.668 7.769 rebuild_ks_matrix 119 8.3 0.000 0.000 7.267 7.412 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.267 7.411 prepare_preconditioner 11 7.9 0.000 0.000 6.850 6.864 make_preconditioner 11 8.9 0.000 0.000 6.850 6.864 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.397 6.742 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.550 6.681 dbcsr_mm_accdrv_process 12550 15.8 4.720 6.535 5.339 6.634 cp_fm_upper_to_full 72 14.2 3.221 4.618 3.221 4.618 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.247 4.300 calculate_rho_elec 119 8.7 0.117 0.120 4.246 4.299 sum_up_and_integrate 119 10.3 0.001 0.001 3.851 3.857 integrate_v_rspace 119 11.3 0.004 0.004 3.841 3.847 init_scf_run 11 5.9 0.000 0.001 3.816 3.817 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.816 3.816 qs_ot_get_p 119 10.4 0.002 0.003 3.346 3.485 mp_waitall_1 94719 16.7 2.272 3.210 2.272 3.210 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.806 3.201 pw_transfer 1439 11.6 0.068 0.068 3.018 3.024 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.920 2.926 dbcsr_complete_redistribute 329 12.2 0.282 0.287 1.979 2.860 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.679 2.544 make_m2s 4572 13.5 0.037 0.037 2.400 2.544 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.213 2.509 apply_single 119 13.6 0.000 0.000 2.212 2.509 fft_wrap_pw1pw2_140 487 13.2 0.652 0.655 2.496 2.503 make_images 4572 14.5 0.352 0.386 2.280 2.423 calculate_dm_sparse 119 9.5 0.000 0.000 2.350 2.396 calculate_first_density_matrix 1 7.0 0.000 0.000 2.297 2.351 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.448 2.307 mp_alltoall_i22 627 13.8 1.429 2.256 1.429 2.256 density_rs2pw 119 9.7 0.003 0.004 2.226 2.248 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.283 2.163 grid_integrate_task_list 119 12.3 2.087 2.116 2.087 2.116 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.041 2.092 ot_diis_step 108 11.5 0.014 0.014 2.042 2.043 qs_ot_p2m_diag 50 11.0 0.043 0.043 1.952 1.954 acc_transpose_blocks 9144 15.5 0.045 0.046 1.825 1.863 fft3d_ps 1201 14.6 0.871 0.886 1.839 1.843 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.840 1.841 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.817 1.819 mp_sum_l 7287 12.8 1.002 1.779 1.002 1.779 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.591 1.628 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.601 1.602 grid_collocate_task_list 119 9.7 1.528 1.566 1.528 1.566 cp_fm_cholesky_invert 11 10.9 1.416 1.419 1.416 1.419 wfi_extrapolate 11 7.9 0.001 0.001 1.401 1.402 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.343 1.393 potential_pw2rs 119 12.3 0.013 0.014 1.344 1.345 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.315 1.337 hybrid_alltoall_any 4725 16.4 0.091 0.152 1.115 1.337 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.327 1.327 cp_fm_diag_elpa_base 50 14.0 1.183 1.235 1.325 1.325 mp_alltoall_d11v 2130 13.8 1.187 1.254 1.187 1.254 make_images_data 4572 15.5 0.046 0.049 1.063 1.241 qs_env_update_s_mstruct 11 6.9 0.031 0.044 1.133 1.154 acc_transpose_blocks_sync 27432 16.5 1.104 1.140 1.104 1.140 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.998 1.037 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.925 0.939 qs_create_task_list 11 7.9 0.001 0.001 0.926 0.938 generate_qs_task_list 11 8.9 0.367 0.386 0.926 0.938 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.907000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=725.636364, yerr=11.695164 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 493.293568E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.041 81.500 81.503 qs_mol_dyn_low 1 2.0 0.004 0.009 81.080 81.090 qs_forces 11 3.9 0.003 0.003 81.007 81.009 qs_energies 11 4.9 0.001 0.003 78.147 78.160 scf_env_do_scf 11 5.9 0.000 0.001 69.322 69.324 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 63.897 63.898 dbcsr_multiply_generic 2055 12.4 0.106 0.109 50.695 51.013 qs_scf_new_mos 99 7.5 0.000 0.001 46.910 47.016 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.910 47.016 ot_scf_mini 99 9.5 0.002 0.002 44.557 44.636 multiply_cannon 2055 13.4 0.182 0.186 42.377 43.329 velocity_verlet 10 3.0 0.002 0.004 42.791 42.792 multiply_cannon_loop 2055 14.4 1.783 1.826 41.421 42.389 ot_mini 99 10.5 0.001 0.001 26.327 26.434 qs_ot_get_derivative 99 11.5 0.001 0.001 19.509 19.641 multiply_cannon_multrec 49320 15.4 11.420 12.066 17.419 18.078 rebuild_ks_matrix 110 8.3 0.000 0.000 14.446 14.662 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.015 14.445 14.662 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.664 12.867 mp_waitall_1 220248 16.4 10.499 11.140 10.499 11.140 multiply_cannon_sync_h2d 49320 15.4 9.657 10.303 9.657 10.303 qs_ot_get_p 110 10.4 0.005 0.019 9.716 9.825 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.554 8.068 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.231 7.701 apply_single 110 13.6 0.000 0.000 7.231 7.701 multiply_cannon_metrocomm3 49320 15.4 0.082 0.086 6.408 7.252 sum_up_and_integrate 110 10.3 0.002 0.003 6.950 6.963 integrate_v_rspace 110 11.3 0.003 0.003 6.925 6.944 init_scf_run 11 5.9 0.000 0.001 6.768 6.768 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.768 6.768 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.572 6.604 ot_diis_step 99 11.5 0.005 0.006 6.599 6.599 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.279 6.421 calculate_rho_elec 110 8.6 0.020 0.025 6.278 6.420 dbcsr_mm_accdrv_process 87628 16.1 3.061 3.153 5.869 6.158 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.698 5.699 init_scf_loop 11 6.9 0.000 0.001 5.398 5.398 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.234 5.283 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.132 5.155 cp_fm_diag_elpa_base 48 14.0 5.119 5.143 5.130 5.153 mp_sum_l 6594 12.7 4.030 4.829 4.030 4.829 wfi_extrapolate 11 7.9 0.001 0.001 4.016 4.016 make_m2s 4110 13.4 0.059 0.064 3.842 3.963 make_images 4110 14.4 0.177 0.189 3.747 3.873 calculate_dm_sparse 110 9.5 0.001 0.001 3.713 3.798 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 3.548 3.552 density_rs2pw 110 9.6 0.004 0.004 3.307 3.464 grid_integrate_task_list 110 12.3 3.259 3.387 3.259 3.387 multiply_cannon_metrocomm1 49320 15.4 0.065 0.068 2.283 3.293 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.236 3.289 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.192 3.268 prepare_preconditioner 11 7.9 0.000 0.000 3.234 3.263 make_preconditioner 11 8.9 0.000 0.000 3.234 3.263 pw_transfer 1331 11.6 0.054 0.063 3.049 3.125 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.021 3.070 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.961 3.039 calculate_first_density_matrix 1 7.0 0.000 0.001 2.655 2.659 fft_wrap_pw1pw2_140 451 13.1 0.309 0.331 2.538 2.630 acc_transpose_blocks 49320 15.4 0.208 0.215 2.560 2.618 potential_pw2rs 110 12.3 0.005 0.006 2.548 2.566 jit_kernel_multiply 13 15.9 2.528 2.550 2.528 2.550 mp_alltoall_d11v 2046 13.8 2.040 2.499 2.040 2.499 grid_collocate_task_list 110 9.6 2.157 2.317 2.157 2.317 fft3d_ps 1111 14.6 0.798 0.888 2.255 2.310 mp_waitany 14300 13.8 1.840 2.136 1.840 2.136 transfer_rs2pw 451 10.6 0.005 0.006 1.871 2.030 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.912 1.940 make_images_data 4110 15.4 0.043 0.047 1.746 1.872 mp_sum_d 3891 11.9 1.354 1.848 1.354 1.848 cp_fm_cholesky_invert 11 10.9 1.758 1.761 1.758 1.761 hybrid_alltoall_any 4261 16.3 0.084 0.487 1.503 1.746 transfer_pw2rs 451 13.1 0.006 0.007 1.727 1.735 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.640 1.671 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.503000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=468.363636, yerr=2.804954 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 586.551296E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 3418652. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.027 68.623 68.624 qs_mol_dyn_low 1 2.0 0.003 0.003 68.381 68.392 qs_forces 11 3.9 0.003 0.003 68.310 68.312 qs_energies 11 4.9 0.001 0.002 64.989 64.993 scf_env_do_scf 11 5.9 0.000 0.001 56.487 56.491 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.942 48.943 dbcsr_multiply_generic 2055 12.4 0.115 0.119 37.767 37.938 velocity_verlet 10 3.0 0.005 0.012 35.848 35.850 qs_scf_new_mos 99 7.5 0.001 0.001 33.156 33.280 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.156 33.279 multiply_cannon 2055 13.4 0.223 0.241 31.211 32.095 ot_scf_mini 99 9.5 0.003 0.003 31.489 31.618 multiply_cannon_loop 2055 14.4 1.160 1.187 29.992 30.813 ot_mini 99 10.5 0.001 0.001 18.308 18.435 multiply_cannon_multrec 24660 15.4 6.983 8.631 14.037 15.606 rebuild_ks_matrix 110 8.3 0.000 0.000 13.522 13.666 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.521 13.666 qs_ot_get_derivative 99 11.5 0.001 0.001 12.558 12.689 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.894 12.025 mp_waitall_1 176588 16.5 7.421 9.778 7.421 9.778 init_scf_loop 11 6.9 0.000 0.000 7.511 7.512 multiply_cannon_metrocomm3 24660 15.4 0.073 0.076 5.078 7.476 multiply_cannon_sync_h2d 24660 15.4 6.348 7.458 6.348 7.458 dbcsr_mm_accdrv_process 52282 16.1 5.502 6.279 6.884 7.247 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.422 7.017 apply_single 110 13.6 0.000 0.001 6.422 7.016 qs_ot_get_p 110 10.4 0.001 0.001 6.552 6.709 sum_up_and_integrate 110 10.3 0.001 0.002 6.354 6.365 integrate_v_rspace 110 11.3 0.003 0.003 6.327 6.338 init_scf_run 11 5.9 0.000 0.001 6.114 6.115 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.114 6.114 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.866 5.874 calculate_rho_elec 110 8.6 0.039 0.047 5.866 5.874 ot_diis_step 99 11.5 0.010 0.010 5.700 5.700 prepare_preconditioner 11 7.9 0.000 0.000 5.472 5.489 make_preconditioner 11 8.9 0.000 0.000 5.472 5.489 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.697 5.420 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.068 5.221 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.636 4.656 make_m2s 4110 13.4 0.056 0.059 4.162 4.572 make_images 4110 14.4 0.407 0.466 4.051 4.456 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.165 4.166 pw_transfer 1331 11.6 0.065 0.073 3.539 3.679 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.570 3.584 cp_fm_diag_elpa_base 48 14.0 3.523 3.544 3.567 3.580 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.433 3.576 wfi_extrapolate 11 7.9 0.001 0.001 3.481 3.481 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.283 3.352 density_rs2pw 110 9.6 0.004 0.004 3.112 3.306 grid_integrate_task_list 110 12.3 3.173 3.305 3.173 3.305 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.242 3.244 fft_wrap_pw1pw2_140 451 13.1 0.358 0.381 2.943 3.093 calculate_dm_sparse 110 9.5 0.001 0.001 3.001 3.032 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.865 2.914 hybrid_alltoall_any 4261 16.3 0.106 0.462 1.951 2.726 fft3d_ps 1111 14.6 1.116 1.353 2.524 2.681 make_images_data 4110 15.4 0.049 0.053 2.248 2.676 cp_fm_cholesky_invert 11 10.9 2.537 2.544 2.537 2.544 calculate_first_density_matrix 1 7.0 0.000 0.000 2.536 2.538 mp_sum_l 6594 12.7 1.759 2.495 1.759 2.495 grid_collocate_task_list 110 9.6 2.172 2.298 2.172 2.298 potential_pw2rs 110 12.3 0.008 0.009 2.229 2.240 acc_transpose_blocks 24660 15.4 0.116 0.119 1.982 2.022 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.964 1.981 mp_alltoall_d11v 2046 13.8 1.724 1.900 1.724 1.900 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.866 1.867 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.783 1.793 jit_kernel_multiply 9 16.4 1.011 1.757 1.011 1.757 transfer_rs2pw 451 10.6 0.006 0.007 1.420 1.656 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.567 1.577 multiply_cannon_metrocomm4 22605 15.4 0.074 0.079 0.788 1.555 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.386 1.495 mp_allgather_i34 2055 14.4 0.521 1.492 0.521 1.492 mp_irecv_dv 57340 16.2 0.661 1.436 0.661 1.436 dbcsr_complete_redistribute 325 12.2 0.268 0.330 1.157 1.423 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.624000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=552.818182, yerr=7.233988 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 660.299776E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.137 60.054 60.055 qs_mol_dyn_low 1 2.0 0.007 0.017 59.607 59.617 qs_forces 11 3.9 0.003 0.003 59.515 59.517 qs_energies 11 4.9 0.013 0.036 56.335 56.339 scf_env_do_scf 11 5.9 0.001 0.002 47.986 47.986 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.576 39.576 velocity_verlet 10 3.0 0.004 0.017 32.303 32.308 dbcsr_multiply_generic 2055 12.4 0.110 0.114 28.682 28.956 qs_scf_new_mos 99 7.5 0.001 0.001 25.007 25.100 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.006 25.099 ot_scf_mini 99 9.5 0.002 0.003 23.755 23.863 multiply_cannon 2055 13.4 0.212 0.222 22.473 23.782 multiply_cannon_loop 2055 14.4 0.814 0.843 21.300 22.285 ot_mini 99 10.5 0.001 0.001 13.628 13.736 rebuild_ks_matrix 110 8.3 0.000 0.000 12.070 12.199 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.017 12.069 12.199 multiply_cannon_multrec 16440 15.4 3.656 5.183 9.889 11.387 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.624 10.741 mp_waitall_1 139946 16.5 6.749 9.905 6.749 9.905 qs_ot_get_derivative 99 11.5 0.001 0.001 9.196 9.307 init_scf_loop 11 6.9 0.001 0.003 8.377 8.378 multiply_cannon_metrocomm3 16440 15.4 0.046 0.047 4.260 7.222 prepare_preconditioner 11 7.9 0.000 0.000 6.593 6.612 make_preconditioner 11 8.9 0.000 0.001 6.593 6.612 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.934 6.297 sum_up_and_integrate 110 10.3 0.002 0.002 6.219 6.232 integrate_v_rspace 110 11.3 0.003 0.003 6.193 6.206 dbcsr_mm_accdrv_process 34862 16.1 5.434 5.745 6.076 6.154 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.734 5.746 calculate_rho_elec 110 8.6 0.059 0.060 5.734 5.745 init_scf_run 11 5.9 0.000 0.001 5.610 5.610 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.609 5.610 qs_ot_get_p 110 10.4 0.002 0.005 5.325 5.449 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.953 5.409 apply_single 110 13.6 0.000 0.000 4.952 5.409 make_m2s 4110 13.4 0.049 0.053 4.142 4.514 ot_diis_step 99 11.5 0.011 0.011 4.402 4.403 make_images 4110 14.4 0.402 0.524 4.026 4.400 multiply_cannon_sync_h2d 16440 15.4 3.265 3.913 3.265 3.913 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.139 3.788 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.684 3.687 pw_transfer 1331 11.6 0.065 0.072 3.581 3.588 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.475 3.485 grid_integrate_task_list 110 12.3 3.202 3.355 3.202 3.355 cp_dbcsr_syevd 48 12.0 0.003 0.005 3.312 3.312 density_rs2pw 110 9.6 0.004 0.004 2.959 3.091 fft_wrap_pw1pw2_140 451 13.1 0.460 0.471 3.022 3.033 wfi_extrapolate 11 7.9 0.001 0.003 2.964 2.964 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.834 2.835 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.781 2.790 cp_fm_diag_elpa_base 48 14.0 2.714 2.744 2.778 2.787 make_images_data 4110 15.4 0.047 0.051 2.297 2.779 hybrid_alltoall_any 4261 16.3 0.108 0.379 2.043 2.718 calculate_dm_sparse 110 9.5 0.001 0.001 2.560 2.599 cp_fm_cholesky_invert 11 10.9 2.557 2.563 2.557 2.563 calculate_first_density_matrix 1 7.0 0.000 0.003 2.555 2.557 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.487 2.539 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.405 2.463 multiply_cannon_metrocomm4 14385 15.4 0.047 0.051 0.890 2.449 grid_collocate_task_list 110 9.6 2.220 2.406 2.220 2.406 fft3d_ps 1111 14.6 1.119 1.131 2.392 2.405 mp_irecv_dv 48980 15.7 0.814 2.306 0.814 2.306 qs_energies_init_hamiltonians 11 5.9 0.028 0.041 2.236 2.239 potential_pw2rs 110 12.3 0.011 0.012 2.070 2.074 mp_sum_l 6594 12.7 1.491 2.058 1.491 2.058 mp_alltoall_d11v 2046 13.8 1.748 2.001 1.748 2.001 dbcsr_complete_redistribute 325 12.2 0.329 0.358 1.449 1.915 cp_fm_upper_to_full 70 14.2 1.393 1.753 1.393 1.753 acc_transpose_blocks 16440 15.4 0.078 0.081 1.575 1.693 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.638 1.651 cp_fm_cholesky_decompose 22 10.9 1.534 1.558 1.534 1.558 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.493 1.507 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.370 1.500 mp_allgather_i34 2055 14.4 0.485 1.484 0.485 1.484 transfer_rs2pw 451 10.6 0.005 0.006 1.312 1.440 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.969 1.416 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.284 1.354 mp_waitany 17072 13.8 1.175 1.339 1.175 1.339 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.266 1.273 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.055000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.000000, yerr=9.638936 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 723.939328E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.032 65.790 65.792 qs_mol_dyn_low 1 2.0 0.004 0.009 65.462 65.479 qs_forces 11 3.9 0.003 0.003 65.389 65.396 qs_energies 11 4.9 0.002 0.006 62.000 62.008 scf_env_do_scf 11 5.9 0.001 0.001 53.662 53.664 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.107 42.107 velocity_verlet 10 3.0 0.002 0.005 37.122 37.124 dbcsr_multiply_generic 2055 12.4 0.118 0.124 30.765 30.996 qs_scf_new_mos 99 7.5 0.001 0.001 27.480 27.587 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.480 27.587 ot_scf_mini 99 9.5 0.003 0.003 25.793 25.906 multiply_cannon 2055 13.4 0.239 0.256 23.752 24.873 multiply_cannon_loop 2055 14.4 1.386 1.445 22.217 22.817 ot_mini 99 10.5 0.001 0.001 14.917 15.030 multiply_cannon_multrec 24660 15.4 4.118 6.788 13.075 14.306 rebuild_ks_matrix 110 8.3 0.000 0.000 11.992 12.117 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 11.992 12.116 init_scf_loop 11 6.9 0.000 0.002 11.513 11.514 qs_ot_get_derivative 99 11.5 0.001 0.001 10.762 10.872 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.593 10.703 prepare_preconditioner 11 7.9 0.000 0.000 9.776 9.789 make_preconditioner 11 8.9 0.000 0.001 9.776 9.789 dbcsr_mm_accdrv_process 52304 16.0 7.779 9.162 8.799 9.705 make_full_inverse_cholesky 11 9.9 0.000 0.001 8.039 9.481 mp_waitall_1 121746 16.5 4.272 6.367 4.272 6.367 sum_up_and_integrate 110 10.3 0.001 0.002 6.205 6.220 integrate_v_rspace 110 11.3 0.003 0.004 6.180 6.194 qs_ot_get_p 110 10.4 0.001 0.001 5.749 5.901 make_m2s 4110 13.4 0.059 0.060 5.393 5.883 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.843 5.855 calculate_rho_elec 110 8.6 0.078 0.081 5.843 5.854 make_images 4110 14.4 0.576 0.703 5.254 5.740 init_scf_run 11 5.9 0.000 0.001 5.566 5.566 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.566 5.566 cp_fm_upper_to_full 70 14.2 3.366 4.850 3.366 4.850 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.032 4.120 apply_single 110 13.6 0.000 0.000 4.031 4.120 ot_diis_step 99 11.5 0.011 0.011 4.115 4.115 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.891 3.906 pw_transfer 1331 11.6 0.065 0.073 3.748 3.782 dbcsr_complete_redistribute 325 12.2 0.412 0.455 2.600 3.749 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.641 3.679 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.486 3.538 grid_integrate_task_list 110 12.3 3.272 3.439 3.272 3.439 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.336 3.394 cp_dbcsr_syevd 48 12.0 0.006 0.029 3.333 3.334 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.088 3.221 hybrid_alltoall_any 4261 16.3 0.123 0.453 2.314 3.190 fft_wrap_pw1pw2_140 451 13.1 0.530 0.542 3.141 3.180 make_images_data 4110 15.4 0.049 0.053 2.666 3.143 density_rs2pw 110 9.6 0.004 0.004 2.962 3.132 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.462 3.074 calculate_dm_sparse 110 9.5 0.001 0.001 2.999 3.039 wfi_extrapolate 11 7.9 0.001 0.001 3.015 3.015 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.859 2.860 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.728 2.847 mp_alltoall_i22 605 13.7 1.625 2.793 1.625 2.793 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.780 2.789 cp_fm_diag_elpa_base 48 14.0 2.631 2.692 2.777 2.787 cp_fm_cholesky_invert 11 10.9 2.620 2.628 2.620 2.628 multiply_cannon_sync_h2d 24660 15.4 2.361 2.522 2.361 2.522 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.483 2.522 fft3d_ps 1111 14.6 1.152 1.194 2.469 2.493 acc_transpose_blocks 24660 15.4 0.111 0.114 2.371 2.467 calculate_first_density_matrix 1 7.0 0.000 0.003 2.454 2.457 grid_collocate_task_list 110 9.6 2.262 2.405 2.262 2.405 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.273 2.279 potential_pw2rs 110 12.3 0.012 0.013 2.015 2.018 mp_alltoall_d11v 2046 13.8 1.775 1.984 1.775 1.984 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.781 1.811 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.613 1.708 cp_fm_cholesky_decompose 22 10.9 1.640 1.691 1.640 1.691 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.639 1.652 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.556 1.573 multiply_cannon_metrocomm4 20550 15.4 0.060 0.063 0.853 1.530 mp_sum_l 6594 12.7 0.938 1.494 0.938 1.494 mp_irecv_dv 62702 16.1 0.746 1.448 0.746 1.448 acc_transpose_blocks_sync 73980 16.4 1.340 1.419 1.340 1.419 mp_allgather_i34 2055 14.4 0.622 1.374 0.622 1.374 transfer_rs2pw 451 10.6 0.005 0.005 1.179 1.368 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.792000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=686.636364, yerr=8.059695 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 841.564160E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.030 55.679 55.680 qs_mol_dyn_low 1 2.0 0.003 0.003 55.404 55.421 qs_forces 11 3.9 0.004 0.011 55.339 55.339 qs_energies 11 4.9 0.001 0.002 51.676 51.683 scf_env_do_scf 11 5.9 0.000 0.001 43.254 43.255 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.419 35.421 velocity_verlet 10 3.0 0.002 0.002 31.352 31.356 dbcsr_multiply_generic 2055 12.4 0.107 0.109 23.293 23.515 qs_scf_new_mos 99 7.5 0.001 0.001 20.796 20.867 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.796 20.867 ot_scf_mini 99 9.5 0.002 0.002 19.540 19.588 multiply_cannon 2055 13.4 0.270 0.282 17.804 18.984 multiply_cannon_loop 2055 14.4 0.604 0.626 16.530 16.830 rebuild_ks_matrix 110 8.3 0.000 0.000 11.662 11.703 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.662 11.702 ot_mini 99 10.5 0.001 0.001 10.873 10.918 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.375 10.408 multiply_cannon_multrec 8220 15.4 3.153 4.686 7.588 8.766 init_scf_loop 11 6.9 0.000 0.000 7.788 7.791 mp_waitall_1 103326 16.6 5.934 7.586 5.934 7.586 qs_ot_get_derivative 99 11.5 0.001 0.001 7.095 7.142 sum_up_and_integrate 110 10.3 0.001 0.002 6.245 6.258 integrate_v_rspace 110 11.3 0.003 0.003 6.218 6.230 prepare_preconditioner 11 7.9 0.000 0.000 6.149 6.155 make_preconditioner 11 8.9 0.000 0.000 6.149 6.155 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.025 6.043 calculate_rho_elec 110 8.6 0.113 0.114 6.024 6.043 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.723 5.797 init_scf_run 11 5.9 0.000 0.001 5.274 5.274 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.274 5.274 dbcsr_mm_accdrv_process 17442 15.9 3.008 4.163 4.293 5.223 qs_ot_get_p 110 10.4 0.001 0.002 4.733 4.792 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 3.155 4.661 make_m2s 4110 13.4 0.037 0.039 4.147 4.333 make_images 4110 14.4 0.648 0.715 4.017 4.198 pw_transfer 1331 11.6 0.065 0.070 4.115 4.125 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.008 4.021 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.804 3.864 apply_single 110 13.6 0.000 0.000 3.804 3.863 ot_diis_step 99 11.5 0.012 0.012 3.757 3.757 grid_integrate_task_list 110 12.3 3.387 3.506 3.387 3.506 fft_wrap_pw1pw2_140 451 13.1 0.721 0.733 3.481 3.493 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.374 3.378 density_rs2pw 110 9.6 0.004 0.004 3.039 3.149 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.035 3.036 cp_fm_cholesky_invert 11 10.9 2.997 3.001 2.997 3.001 wfi_extrapolate 11 7.9 0.001 0.001 2.694 2.694 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.676 2.678 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 2.677 2.678 fft3d_ps 1111 14.6 1.299 1.306 2.547 2.561 calculate_dm_sparse 110 9.5 0.001 0.001 2.514 2.550 hybrid_alltoall_any 4261 16.3 0.199 0.846 2.156 2.536 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.524 2.531 cp_fm_diag_elpa_base 48 14.0 2.469 2.489 2.522 2.530 grid_collocate_task_list 110 9.6 2.367 2.505 2.367 2.505 calculate_first_density_matrix 1 7.0 0.000 0.000 2.484 2.485 multiply_cannon_sync_h2d 8220 15.4 2.348 2.464 2.348 2.464 make_images_data 4110 15.4 0.041 0.047 2.190 2.453 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.190 2.211 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.012 2.051 potential_pw2rs 110 12.3 0.015 0.015 2.039 2.042 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.789 2.007 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.917 1.942 mp_alltoall_d11v 2046 13.8 1.640 1.867 1.640 1.867 cp_fm_cholesky_decompose 22 10.9 1.706 1.723 1.706 1.723 dbcsr_complete_redistribute 325 12.2 0.653 0.699 1.541 1.634 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.616 1.624 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.502 1.616 jit_kernel_multiply 9 15.8 0.963 1.539 0.963 1.539 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.461 1.471 mp_allgather_i34 2055 14.4 0.399 1.428 0.399 1.428 acc_transpose_blocks 8220 15.4 0.041 0.042 1.374 1.417 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.482 1.315 qs_create_task_list 11 7.9 0.001 0.001 1.212 1.310 generate_qs_task_list 11 8.9 0.372 0.437 1.211 1.309 mp_irecv_dv 24056 15.7 0.454 1.269 0.454 1.269 transfer_rs2pw 451 10.6 0.005 0.005 1.109 1.243 mp_waitany 9240 13.8 1.051 1.178 1.051 1.178 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.139 1.175 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.680000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=792.363636, yerr=11.299821 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 1.357435E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.033 87.998 87.998 qs_mol_dyn_low 1 2.0 0.003 0.003 87.692 87.714 qs_forces 11 3.9 0.003 0.003 87.521 87.522 qs_energies 11 4.9 0.001 0.002 83.354 83.356 scf_env_do_scf 11 5.9 0.000 0.001 73.158 73.159 velocity_verlet 10 3.0 0.017 0.042 56.262 56.276 scf_env_do_scf_inner_loop 99 6.5 0.004 0.007 44.615 44.616 dbcsr_multiply_generic 2055 12.4 0.135 0.153 29.490 29.573 init_scf_loop 11 6.9 0.000 0.000 28.473 28.474 qs_scf_new_mos 99 7.5 0.001 0.001 26.681 26.735 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.680 26.734 prepare_preconditioner 11 7.9 0.000 0.000 26.430 26.439 make_preconditioner 11 8.9 0.000 0.000 26.430 26.439 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.419 25.913 ot_scf_mini 99 9.5 0.002 0.002 24.882 24.930 multiply_cannon 2055 13.4 0.329 0.358 22.520 23.164 multiply_cannon_loop 2055 14.4 0.822 0.835 20.767 21.089 cp_fm_upper_to_full 70 14.2 12.819 18.560 12.819 18.560 ot_mini 99 10.5 0.001 0.001 14.093 14.139 rebuild_ks_matrix 110 8.3 0.000 0.001 13.965 14.007 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.965 14.006 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.667 12.705 dbcsr_complete_redistribute 325 12.2 1.008 1.027 7.574 10.898 multiply_cannon_multrec 8220 15.4 4.061 4.233 9.767 9.892 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.553 9.876 qs_ot_get_derivative 99 11.5 0.001 0.001 9.488 9.536 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.995 9.278 mp_alltoall_i22 605 13.7 5.624 8.970 5.624 8.970 mp_waitall_1 84994 16.7 7.458 8.293 7.458 8.293 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.481 7.519 calculate_rho_elec 110 8.6 0.223 0.224 7.480 7.519 sum_up_and_integrate 110 10.3 0.002 0.002 6.914 6.926 integrate_v_rspace 110 11.3 0.004 0.004 6.885 6.897 init_scf_run 11 5.9 0.000 0.001 5.930 5.930 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.929 5.930 pw_transfer 1331 11.6 0.074 0.074 5.816 5.826 dbcsr_mm_accdrv_process 11614 15.7 3.917 4.160 5.557 5.768 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.698 5.708 make_m2s 4110 13.4 0.042 0.043 5.240 5.693 qs_ot_get_p 110 10.4 0.010 0.012 5.507 5.563 make_images 4110 14.4 0.888 0.945 5.052 5.504 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.957 5.298 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.750 5.210 apply_single 110 13.6 0.000 0.000 4.750 5.210 cp_fm_cholesky_invert 11 10.9 5.126 5.131 5.126 5.131 fft_wrap_pw1pw2_140 451 13.1 1.367 1.373 5.015 5.028 ot_diis_step 99 11.5 0.015 0.017 4.584 4.584 density_rs2pw 110 9.6 0.004 0.004 3.936 3.969 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.934 3.942 grid_integrate_task_list 110 12.3 3.694 3.779 3.694 3.779 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.666 3.668 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.036 3.484 fft3d_ps 1111 14.6 1.895 1.907 3.402 3.419 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.418 3.419 hybrid_alltoall_any 4261 16.3 0.267 0.580 2.774 3.382 wfi_extrapolate 11 7.9 0.001 0.001 3.324 3.324 make_images_data 4110 15.4 0.045 0.048 2.693 3.294 calculate_dm_sparse 110 9.5 0.001 0.001 3.164 3.196 multiply_cannon_sync_h2d 8220 15.4 3.129 3.152 3.129 3.152 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.906 2.911 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.841 2.841 cp_fm_diag_elpa_base 48 14.0 2.308 2.501 2.838 2.839 grid_collocate_task_list 110 9.6 2.682 2.705 2.682 2.705 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.644 2.670 calculate_first_density_matrix 1 7.0 0.000 0.000 2.488 2.488 potential_pw2rs 110 12.3 0.021 0.021 2.434 2.438 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.415 2.428 qs_env_update_s_mstruct 11 6.9 0.001 0.003 2.185 2.247 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.097 2.235 mp_alltoall_d11v 2046 13.8 1.985 2.064 1.985 2.064 cp_fm_cholesky_decompose 22 10.9 1.953 1.969 1.953 1.969 qs_create_task_list 11 7.9 0.001 0.001 1.873 1.923 generate_qs_task_list 11 8.9 0.728 0.781 1.872 1.922 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.896 1.900 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.732 1.762 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.998000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1223.909091, yerr=54.071425 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/15/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 1.086553E+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 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 625.250304E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 57688. MP_Allreduce 11097 796. MP_Sync 86 MP_Alltoall 2210 3925666. MP_SendRecv 24130 18752. MP_ISendRecv 24130 18752. MP_Wait 42150 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.051 205.438 205.440 qs_mol_dyn_low 1 2.0 0.011 0.037 204.853 204.866 qs_forces 11 3.9 0.005 0.005 204.675 204.680 qs_energies 11 4.9 0.001 0.002 199.102 199.120 scf_env_do_scf 11 5.9 0.001 0.001 182.399 182.403 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 162.112 162.114 dbcsr_multiply_generic 2485 12.5 0.202 0.236 125.357 125.866 qs_scf_new_mos 116 7.6 0.001 0.001 123.371 123.602 qs_scf_loop_do_ot 116 8.6 0.001 0.001 123.371 123.601 velocity_verlet 10 3.0 0.025 0.069 121.584 121.586 ot_scf_mini 116 9.6 0.003 0.003 116.751 116.964 multiply_cannon 2485 13.5 0.232 0.240 101.281 102.635 multiply_cannon_loop 2485 14.5 2.372 2.432 99.190 100.678 ot_mini 116 10.6 0.001 0.001 66.294 66.504 multiply_cannon_multrec 59640 15.5 31.664 33.183 41.536 43.233 qs_ot_get_derivative 116 11.6 0.001 0.001 41.580 41.812 rebuild_ks_matrix 127 8.3 0.001 0.001 33.142 33.503 qs_ks_build_kohn_sham_matrix 127 9.3 0.015 0.017 33.141 33.503 mp_waitall_1 264810 16.5 28.768 31.667 28.768 31.667 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.727 30.027 qs_ot_get_p 127 10.4 0.001 0.001 28.430 28.746 multiply_cannon_sync_h2d 59640 15.5 26.124 27.941 26.124 27.941 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.210 25.054 apply_single 127 13.6 0.001 0.001 24.209 25.053 ot_diis_step 116 11.6 0.007 0.008 24.432 24.433 qs_ot_p2m_diag 82 11.4 0.078 0.091 21.914 21.992 init_scf_loop 11 6.9 0.000 0.001 20.213 20.214 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 19.897 20.048 cp_dbcsr_syevd 82 12.4 0.005 0.005 19.105 19.106 multiply_cannon_metrocomm3 59640 15.5 0.115 0.119 16.172 17.605 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.226 16.264 cp_fm_diag_elpa_base 82 14.4 16.160 16.200 16.222 16.261 prepare_preconditioner 11 7.9 0.000 0.000 15.633 15.678 make_preconditioner 11 8.9 0.000 0.000 15.633 15.678 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.887 15.058 make_m2s 4970 13.5 0.103 0.113 13.677 14.113 make_images 4970 14.5 0.392 0.412 13.494 13.938 sum_up_and_integrate 127 10.3 0.002 0.004 13.746 13.760 integrate_v_rspace 127 11.3 0.003 0.004 13.688 13.706 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.895 13.078 calculate_rho_elec 127 8.7 0.045 0.063 12.895 13.077 init_scf_run 11 5.9 0.000 0.001 12.504 12.505 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.504 12.505 mp_sum_l 7884 12.9 9.135 10.402 9.135 10.402 dbcsr_mm_accdrv_process 123452 16.2 4.703 4.835 9.438 9.969 wfi_extrapolate 11 7.9 0.001 0.001 9.032 9.032 cp_fm_cholesky_invert 11 10.9 8.892 8.900 8.892 8.900 calculate_dm_sparse 127 9.5 0.001 0.001 8.501 8.623 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.249 8.386 qs_ot_get_orbitals 116 10.6 0.001 0.001 8.006 8.073 multiply_cannon_metrocomm1 59640 15.5 0.094 0.098 6.299 7.650 pw_transfer 1535 11.6 0.072 0.088 7.467 7.646 make_images_data 4970 15.5 0.066 0.072 6.666 7.596 fft_wrap_pw1pw2 1281 12.7 0.009 0.011 7.269 7.440 grid_integrate_task_list 127 12.3 7.001 7.387 7.001 7.387 density_rs2pw 127 9.7 0.006 0.007 6.585 7.117 hybrid_alltoall_any 5155 16.4 0.290 2.239 5.874 7.074 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.005 6.792 6.802 fft_wrap_pw1pw2_140 519 13.2 0.852 0.891 6.411 6.574 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.841 5.970 fft3d_ps 1281 14.7 2.183 2.828 5.286 5.604 mp_alltoall_d11v 2401 14.1 4.248 5.211 4.248 5.211 grid_collocate_task_list 127 9.7 4.804 5.203 4.804 5.203 cp_fm_cholesky_decompose 22 10.9 4.588 4.602 4.588 4.602 potential_pw2rs 127 12.3 0.009 0.010 4.328 4.347 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.440000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=590.636364, yerr=6.905501 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 835.371008E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57903. MP_Allreduce 11138 958. MP_Sync 87 MP_Alltoall 1983 6484349. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.032 192.527 192.528 qs_mol_dyn_low 1 2.0 0.010 0.061 192.081 192.116 qs_forces 11 3.9 0.018 0.021 191.895 191.905 qs_energies 11 4.9 0.001 0.002 185.114 185.136 scf_env_do_scf 11 5.9 0.001 0.001 168.586 168.596 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 135.429 135.432 velocity_verlet 10 3.0 0.015 0.108 121.459 121.462 dbcsr_multiply_generic 2527 12.6 0.198 0.205 98.708 100.122 qs_scf_new_mos 118 7.6 0.001 0.001 96.289 96.858 qs_scf_loop_do_ot 118 8.6 0.001 0.001 96.288 96.857 ot_scf_mini 118 9.6 0.004 0.004 91.388 92.009 multiply_cannon 2527 13.6 0.507 0.572 77.910 83.221 multiply_cannon_loop 2527 14.6 1.579 1.644 74.483 77.591 ot_mini 118 10.6 0.001 0.001 51.320 51.951 mp_waitall_1 216438 16.6 24.929 39.156 24.929 39.156 multiply_cannon_multrec 30324 15.6 20.797 24.703 31.520 35.675 rebuild_ks_matrix 129 8.3 0.001 0.001 32.579 33.312 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 32.578 33.311 init_scf_loop 11 6.9 0.000 0.000 33.064 33.065 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.315 29.992 qs_ot_get_derivative 118 11.6 0.001 0.002 28.883 29.480 multiply_cannon_metrocomm3 30324 15.6 0.099 0.106 16.260 29.188 prepare_preconditioner 11 7.9 0.000 0.000 28.747 28.832 make_preconditioner 11 8.9 0.000 0.000 28.746 28.832 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.443 27.973 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.505 23.688 apply_single 129 13.6 0.001 0.001 22.505 23.687 qs_ot_get_p 129 10.4 0.001 0.002 21.901 22.582 ot_diis_step 118 11.6 0.014 0.015 22.240 22.243 multiply_cannon_sync_h2d 30324 15.6 17.901 19.666 17.901 19.666 qs_ot_p2m_diag 83 11.4 0.189 0.217 16.911 16.952 cp_fm_cholesky_invert 11 10.9 16.772 16.785 16.772 16.785 make_m2s 5054 13.6 0.087 0.092 14.584 16.230 make_images 5054 14.6 1.187 1.392 14.367 16.017 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.695 15.695 sum_up_and_integrate 129 10.3 0.002 0.004 14.160 14.195 integrate_v_rspace 129 11.3 0.003 0.004 14.100 14.138 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.693 13.729 calculate_rho_elec 129 8.7 0.088 0.105 13.692 13.728 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.387 12.425 cp_fm_diag_elpa_base 83 14.4 12.129 12.239 12.380 12.414 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.429 11.865 init_scf_run 11 5.9 0.000 0.001 11.604 11.605 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.604 11.605 multiply_cannon_metrocomm4 27797 15.6 0.101 0.116 3.942 11.210 mp_irecv_dv 70031 16.3 3.736 10.798 3.736 10.798 dbcsr_mm_accdrv_process 62734 16.2 5.490 6.291 10.171 10.746 make_images_data 5054 15.6 0.068 0.076 8.593 10.464 hybrid_alltoall_any 5240 16.5 0.350 1.534 7.512 10.136 pw_transfer 1559 11.6 0.085 0.097 8.746 8.797 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.520 8.571 wfi_extrapolate 11 7.9 0.001 0.001 8.426 8.426 fft_wrap_pw1pw2_140 527 13.2 0.952 0.985 7.523 7.585 grid_integrate_task_list 129 12.3 7.236 7.577 7.236 7.577 density_rs2pw 129 9.7 0.006 0.006 7.188 7.562 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.515 7.316 cp_fm_cholesky_decompose 22 10.9 7.010 7.083 7.010 7.083 calculate_dm_sparse 129 9.5 0.001 0.001 6.582 6.748 mp_sum_l 8010 12.9 4.387 6.671 4.387 6.671 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.158 6.166 fft3d_ps 1301 14.7 2.866 3.073 6.039 6.088 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.506 5.579 grid_collocate_task_list 129 9.7 5.046 5.491 5.046 5.491 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.232 5.403 mp_alltoall_d11v 2423 14.1 4.646 5.366 4.646 5.366 mp_allgather_i34 2527 14.6 1.933 5.315 1.933 5.315 potential_pw2rs 129 12.3 0.016 0.018 4.592 4.616 mp_sum_d 4501 12.1 2.689 4.122 2.689 4.122 dbcsr_complete_redistribute 395 12.7 0.769 0.848 3.199 4.048 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=192.528000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=795.727273, yerr=3.193485 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.950886E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 4015680 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.9 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 924.020736E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1051232 MPI messages size (bytes): total size 2.737039E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.603648E+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 284160 37245419520 131072 < size <= 4194304 665408 1004401590272 4194304 < size <= 16777216 66080 937889764224 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57897. MP_Allreduce 11137 997. MP_Sync 87 MP_Alltoall 1724 9394172. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.084 177.561 177.563 qs_mol_dyn_low 1 2.0 0.015 0.039 176.893 176.906 qs_forces 11 3.9 0.004 0.006 176.635 176.642 qs_energies 11 4.9 0.009 0.027 169.991 170.002 scf_env_do_scf 11 5.9 0.001 0.002 153.927 153.928 scf_env_do_scf_inner_loop 118 6.6 0.003 0.010 118.714 118.714 velocity_verlet 10 3.0 0.032 0.088 113.799 113.815 dbcsr_multiply_generic 2527 12.6 0.192 0.201 82.740 83.760 qs_scf_new_mos 118 7.6 0.001 0.001 81.125 81.454 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.124 81.453 ot_scf_mini 118 9.6 0.004 0.004 76.810 77.193 multiply_cannon 2527 13.6 0.511 0.535 62.913 67.636 multiply_cannon_loop 2527 14.6 1.139 1.214 59.860 62.588 ot_mini 118 10.6 0.001 0.001 42.946 43.318 init_scf_loop 11 6.9 0.001 0.003 35.114 35.116 mp_waitall_1 171878 16.6 25.179 34.215 25.179 34.215 prepare_preconditioner 11 7.9 0.000 0.000 31.081 31.130 make_preconditioner 11 8.9 0.000 0.001 31.081 31.130 rebuild_ks_matrix 129 8.3 0.001 0.001 30.656 31.124 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.021 30.655 31.123 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.709 30.097 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.604 28.032 multiply_cannon_multrec 20216 15.6 13.004 16.153 22.451 25.604 multiply_cannon_metrocomm3 20216 15.6 0.062 0.066 15.887 24.775 qs_ot_get_derivative 118 11.6 0.001 0.002 23.000 23.375 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.050 21.049 apply_single 129 13.6 0.001 0.001 20.050 21.049 ot_diis_step 118 11.6 0.018 0.018 19.827 19.827 qs_ot_get_p 129 10.4 0.001 0.002 19.015 19.489 make_m2s 5054 13.6 0.079 0.085 14.750 15.938 make_images 5054 14.6 1.166 1.249 14.513 15.699 multiply_cannon_sync_h2d 20216 15.6 13.611 15.492 13.611 15.492 qs_ot_p2m_diag 83 11.4 0.265 0.272 14.724 14.734 cp_fm_cholesky_invert 11 10.9 14.499 14.508 14.499 14.508 sum_up_and_integrate 129 10.3 0.002 0.003 14.162 14.196 integrate_v_rspace 129 11.3 0.003 0.004 14.102 14.138 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.719 13.756 calculate_rho_elec 129 8.7 0.131 0.145 13.718 13.755 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.665 13.666 init_scf_run 11 5.9 0.000 0.001 10.697 10.698 scf_env_initial_rho_setup 11 6.9 0.002 0.002 10.697 10.698 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.452 10.476 cp_fm_diag_elpa_base 83 14.4 10.038 10.182 10.448 10.472 make_images_data 5054 15.6 0.065 0.074 8.790 10.417 hybrid_alltoall_any 5240 16.5 0.453 2.074 7.668 9.796 multiply_cannon_metrocomm4 17689 15.6 0.066 0.076 3.596 9.758 mp_irecv_dv 50625 16.2 3.463 9.488 3.463 9.488 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.001 9.272 pw_transfer 1559 11.6 0.085 0.101 8.994 9.111 dbcsr_mm_accdrv_process 41830 16.2 5.684 6.011 8.903 9.042 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 8.768 8.892 fft_wrap_pw1pw2_140 527 13.2 1.086 1.126 7.786 7.915 grid_integrate_task_list 129 12.3 7.371 7.699 7.371 7.699 wfi_extrapolate 11 7.9 0.003 0.020 7.569 7.570 density_rs2pw 129 9.7 0.006 0.006 7.061 7.521 cp_fm_cholesky_decompose 22 10.9 7.405 7.428 7.405 7.428 cp_fm_upper_to_full 105 14.8 5.734 7.272 5.734 7.272 dbcsr_complete_redistribute 395 12.7 1.169 1.213 4.643 6.387 fft3d_ps 1301 14.7 2.863 3.094 6.038 6.139 calculate_dm_sparse 129 9.5 0.001 0.001 5.928 6.018 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 5.524 5.530 grid_collocate_task_list 129 9.7 5.203 5.520 5.203 5.520 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.671 5.390 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.466 5.204 mp_alltoall_d11v 2423 14.1 4.334 4.750 4.334 4.750 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.594 4.739 mp_allgather_i34 2527 14.6 1.539 4.699 1.539 4.699 mp_sum_l 8010 12.9 3.256 4.691 3.256 4.691 potential_pw2rs 129 12.3 0.020 0.022 4.501 4.515 transfer_fm_to_dbcsr 11 9.9 0.019 0.024 2.352 4.053 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.020 4.052 qs_energies_init_hamiltonians 11 5.9 0.004 0.011 4.025 4.031 mp_alltoall_i22 718 14.1 1.953 3.830 1.953 3.830 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.556 3.598 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.563000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=876.545455, yerr=9.680004 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/18/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 4.353791E+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 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.155641E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57949. MP_Allreduce 11127 1081. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.037 187.978 187.979 qs_mol_dyn_low 1 2.0 0.003 0.003 187.576 187.602 qs_forces 11 3.9 0.004 0.004 187.467 187.474 qs_energies 11 4.9 0.011 0.014 180.353 180.360 scf_env_do_scf 11 5.9 0.001 0.001 163.641 163.654 velocity_verlet 10 3.0 0.001 0.002 124.715 124.717 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 117.150 117.151 dbcsr_multiply_generic 2507 12.6 0.192 0.199 80.527 81.275 qs_scf_new_mos 117 7.6 0.001 0.001 80.767 81.101 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.767 81.100 ot_scf_mini 117 9.6 0.003 0.004 76.216 76.567 multiply_cannon 2507 13.6 0.549 0.580 55.816 59.534 multiply_cannon_loop 2507 14.6 1.844 1.923 52.170 54.058 init_scf_loop 11 6.9 0.000 0.000 46.364 46.365 ot_mini 117 10.6 0.001 0.001 42.938 43.263 prepare_preconditioner 11 7.9 0.000 0.000 42.236 42.263 make_preconditioner 11 8.9 0.000 0.000 42.236 42.263 make_full_inverse_cholesky 11 9.9 0.012 0.022 35.824 40.870 multiply_cannon_multrec 30084 15.6 13.689 19.045 26.667 31.680 rebuild_ks_matrix 128 8.3 0.001 0.001 29.427 29.775 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 29.427 29.775 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.553 26.872 mp_waitall_1 147882 16.7 17.055 26.409 17.055 26.409 qs_ot_get_derivative 117 11.6 0.001 0.002 23.263 23.605 make_m2s 5014 13.6 0.091 0.094 20.384 21.271 make_images 5014 14.6 1.964 2.237 20.087 20.974 qs_ot_get_p 128 10.4 0.001 0.002 19.407 19.739 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.152 19.716 apply_single 128 13.6 0.001 0.001 19.151 19.716 ot_diis_step 117 11.6 0.017 0.019 19.541 19.543 cp_fm_upper_to_full 105 14.8 11.531 16.972 11.531 16.972 cp_fm_cholesky_invert 11 10.9 16.091 16.100 16.091 16.100 qs_ot_p2m_diag 83 11.4 0.343 0.390 15.256 15.306 multiply_cannon_metrocomm3 30084 15.6 0.049 0.051 6.332 14.731 sum_up_and_integrate 128 10.3 0.002 0.003 14.062 14.088 integrate_v_rspace 128 11.3 0.003 0.004 14.001 14.028 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.848 13.850 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.784 13.809 calculate_rho_elec 128 8.7 0.172 0.187 13.783 13.808 dbcsr_mm_accdrv_process 62264 16.2 8.426 9.222 12.548 13.080 dbcsr_complete_redistribute 395 12.7 1.485 1.598 9.100 12.919 make_images_data 5014 15.6 0.066 0.071 10.909 12.561 multiply_cannon_sync_h2d 30084 15.6 10.462 11.901 10.462 11.901 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.727 11.542 hybrid_alltoall_any 5200 16.5 0.533 2.219 9.771 11.164 init_scf_run 11 5.9 0.000 0.001 10.730 10.731 scf_env_initial_rho_setup 11 6.9 0.009 0.023 10.729 10.731 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.639 10.654 cp_fm_diag_elpa_base 83 14.4 9.653 9.967 10.632 10.646 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.390 10.096 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.745 10.010 pw_transfer 1547 11.6 0.085 0.102 9.259 9.324 mp_alltoall_i22 716 14.1 5.570 9.272 5.570 9.272 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.034 9.103 fft_wrap_pw1pw2_140 523 13.2 1.251 1.282 8.043 8.128 grid_integrate_task_list 128 12.3 7.531 7.898 7.531 7.898 wfi_extrapolate 11 7.9 0.001 0.001 7.655 7.655 cp_fm_cholesky_decompose 22 10.9 7.484 7.579 7.484 7.579 density_rs2pw 128 9.7 0.006 0.006 6.850 7.274 multiply_cannon_metrocomm4 25070 15.6 0.080 0.089 2.776 6.942 mp_irecv_dv 76098 16.2 2.619 6.661 2.619 6.661 calculate_dm_sparse 128 9.5 0.001 0.001 6.247 6.324 fft3d_ps 1291 14.7 2.990 3.047 6.077 6.104 grid_collocate_task_list 128 9.7 5.277 5.764 5.277 5.764 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.441 5.517 mp_alltoall_d11v 2415 14.1 4.942 5.473 4.942 5.473 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.461 4.548 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.467 4.467 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.376 4.457 potential_pw2rs 128 12.3 0.022 0.023 4.333 4.341 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.248 4.314 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.979000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1091.818182, yerr=18.219960 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.520001E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532237E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58346. MP_Allreduce 11057 1167. MP_Sync 87 MP_Alltoall 1712 18838210. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.037 170.637 170.637 qs_mol_dyn_low 1 2.0 0.003 0.003 170.155 170.176 qs_forces 11 3.9 0.005 0.008 170.042 170.044 qs_energies 11 4.9 0.002 0.002 162.598 162.607 scf_env_do_scf 11 5.9 0.001 0.001 145.178 145.185 velocity_verlet 10 3.0 0.021 0.022 111.999 112.003 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 109.255 109.257 dbcsr_multiply_generic 2507 12.6 0.184 0.192 72.432 72.814 qs_scf_new_mos 117 7.6 0.001 0.001 72.095 72.224 qs_scf_loop_do_ot 117 8.6 0.001 0.001 72.095 72.223 ot_scf_mini 117 9.6 0.003 0.004 67.688 67.824 multiply_cannon 2507 13.6 0.557 0.583 53.603 56.796 multiply_cannon_loop 2507 14.6 0.804 0.835 50.621 51.543 ot_mini 117 10.6 0.001 0.001 37.720 37.858 init_scf_loop 11 6.9 0.000 0.000 35.773 35.775 prepare_preconditioner 11 7.9 0.000 0.000 31.879 31.907 make_preconditioner 11 8.9 0.000 0.000 31.879 31.907 mp_waitall_1 125778 16.7 25.222 31.745 25.222 31.745 make_full_inverse_cholesky 11 9.9 0.017 0.028 29.747 30.004 rebuild_ks_matrix 128 8.3 0.001 0.001 29.311 29.448 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 29.311 29.447 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.715 26.838 multiply_cannon_multrec 10028 15.6 10.315 14.654 17.984 22.596 multiply_cannon_metrocomm3 10028 15.6 0.024 0.026 12.981 20.719 qs_ot_get_derivative 117 11.6 0.001 0.002 20.540 20.671 cp_fm_cholesky_invert 11 10.9 18.204 18.210 18.204 18.210 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 17.246 17.406 apply_single 128 13.6 0.001 0.001 17.246 17.405 qs_ot_get_p 128 10.4 0.002 0.002 17.164 17.311 ot_diis_step 117 11.6 0.020 0.021 17.109 17.110 make_m2s 5014 13.6 0.063 0.067 14.980 15.824 make_images 5014 14.6 2.166 2.590 14.672 15.522 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.257 14.286 calculate_rho_elec 128 8.7 0.255 0.267 14.257 14.285 sum_up_and_integrate 128 10.3 0.002 0.002 14.178 14.220 integrate_v_rspace 128 11.3 0.004 0.004 14.118 14.161 qs_ot_p2m_diag 83 11.4 0.495 0.501 13.368 13.385 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.203 12.204 multiply_cannon_sync_h2d 10028 15.6 10.694 11.259 10.694 11.259 init_scf_run 11 5.9 0.000 0.001 10.577 10.577 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.576 10.577 pw_transfer 1547 11.6 0.085 0.095 9.916 9.950 hybrid_alltoall_any 5200 16.5 0.840 3.770 8.455 9.907 make_images_data 5014 15.6 0.055 0.065 8.669 9.889 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 9.694 9.734 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.312 9.323 cp_fm_diag_elpa_base 83 14.4 9.074 9.146 9.306 9.318 fft_wrap_pw1pw2_140 523 13.2 1.678 1.717 8.595 8.637 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.051 8.143 grid_integrate_task_list 128 12.3 7.763 8.110 7.763 8.110 cp_fm_cholesky_decompose 22 10.9 7.962 8.095 7.962 8.095 dbcsr_mm_accdrv_process 20762 16.1 3.224 4.186 7.324 8.070 multiply_cannon_metrocomm1 10028 15.6 0.030 0.030 4.336 7.525 wfi_extrapolate 11 7.9 0.001 0.001 7.383 7.383 density_rs2pw 128 9.7 0.005 0.005 6.964 7.340 calculate_dm_sparse 128 9.5 0.001 0.001 6.108 6.188 fft3d_ps 1291 14.7 3.142 3.240 6.140 6.182 grid_collocate_task_list 128 9.7 5.561 5.837 5.561 5.837 multiply_cannon_metrocomm4 7521 15.6 0.026 0.029 1.864 5.635 dbcsr_complete_redistribute 395 12.7 2.103 2.168 5.203 5.581 mp_irecv_dv 28860 15.9 1.822 5.539 1.822 5.539 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.293 5.294 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.209 5.226 mp_alltoall_d11v 2415 14.1 4.633 5.140 4.633 5.140 mp_allgather_i34 2507 14.6 1.168 4.649 1.168 4.649 potential_pw2rs 128 12.3 0.026 0.026 4.350 4.356 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.103 4.158 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.588 3.881 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.390 3.698 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.603 3.695 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.627 3.664 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.411 3.523 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.500 3.513 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=170.637000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1427.090909, yerr=16.527963 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 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 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 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 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 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 3.204473E+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 11.342272E+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 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3955 59814. MP_Allreduce 10965 1506. 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 0.022 0.061 290.658 290.658 qs_mol_dyn_low 1 2.0 0.003 0.003 290.033 290.052 qs_forces 11 3.9 0.027 0.029 289.937 289.939 qs_energies 11 4.9 0.004 0.005 280.968 280.975 scf_env_do_scf 11 5.9 0.001 0.001 258.397 258.410 velocity_verlet 10 3.0 0.003 0.005 209.778 209.786 scf_env_do_scf_inner_loop 116 6.6 0.011 0.015 133.930 133.932 init_scf_loop 11 6.9 0.000 0.000 124.197 124.200 prepare_preconditioner 11 7.9 0.000 0.000 119.320 119.340 make_preconditioner 11 8.9 0.000 0.000 119.320 119.340 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.213 116.459 qs_scf_new_mos 116 7.6 0.001 0.001 89.388 89.488 qs_scf_loop_do_ot 116 8.6 0.001 0.001 89.387 89.487 ot_scf_mini 116 9.6 0.004 0.004 84.586 84.616 dbcsr_multiply_generic 2485 12.5 0.214 0.227 81.815 82.466 cp_fm_upper_to_full 104 14.8 52.915 75.996 52.915 75.996 multiply_cannon 2485 13.5 0.670 0.693 58.572 59.724 multiply_cannon_loop 2485 14.5 1.042 1.053 54.511 56.022 ot_mini 116 10.6 0.001 0.001 44.590 44.640 dbcsr_complete_redistribute 393 12.7 3.982 4.018 30.002 42.821 copy_fm_to_dbcsr 208 11.6 0.001 0.001 26.506 39.232 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.065 36.540 rebuild_ks_matrix 127 8.3 0.001 0.001 34.620 34.686 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 34.619 34.686 mp_alltoall_i22 712 14.1 21.823 34.599 21.823 34.599 cp_fm_cholesky_invert 11 10.9 33.233 33.240 33.233 33.240 mp_waitall_1 102768 16.8 27.912 32.178 27.912 32.178 qs_ks_update_qs_env 127 7.6 0.001 0.001 32.080 32.154 qs_ot_get_p 127 10.4 0.017 0.018 25.262 25.359 qs_ot_get_derivative 116 11.6 0.002 0.002 24.683 24.712 qs_ot_p2m_diag 82 11.4 0.868 0.874 21.177 21.205 make_m2s 4970 13.5 0.073 0.077 18.622 20.159 multiply_cannon_metrocomm3 9940 15.5 0.025 0.025 18.600 20.103 ot_diis_step 116 11.6 0.022 0.022 19.872 19.872 make_images 4970 14.5 3.045 3.232 18.144 19.684 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.405 19.407 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.221 19.386 apply_single 127 13.6 0.001 0.001 19.221 19.385 multiply_cannon_multrec 9940 15.5 10.290 12.126 18.122 18.179 qs_rho_update_rho_low 127 7.7 0.001 0.001 17.192 17.229 calculate_rho_elec 127 8.7 0.476 0.477 17.192 17.229 sum_up_and_integrate 127 10.3 0.002 0.002 16.388 16.476 integrate_v_rspace 127 11.3 0.005 0.005 16.326 16.414 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.270 16.271 cp_fm_diag_elpa_base 82 14.4 11.891 13.517 16.266 16.266 multiply_cannon_sync_h2d 9940 15.5 14.186 14.228 14.186 14.228 pw_transfer 1535 11.6 0.094 0.096 13.035 13.052 hybrid_alltoall_any 5155 16.4 1.300 3.031 10.772 12.961 make_images_data 4970 15.5 0.063 0.067 10.596 12.848 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 12.799 12.816 init_scf_run 11 5.9 0.000 0.001 12.188 12.190 scf_env_initial_rho_setup 11 6.9 0.003 0.005 12.188 12.190 fft_wrap_pw1pw2_140 519 13.2 3.185 3.237 11.388 11.403 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.653 9.676 dbcsr_mm_accdrv_process 20590 16.0 4.240 6.065 7.583 9.488 wfi_extrapolate 11 7.9 0.001 0.001 8.957 8.958 cp_fm_cholesky_decompose 22 10.9 8.876 8.896 8.876 8.896 grid_integrate_task_list 127 12.3 8.497 8.707 8.497 8.707 density_rs2pw 127 9.7 0.005 0.005 8.278 8.337 qs_energies_init_hamiltonians 11 5.9 0.007 0.024 8.219 8.221 fft3d_ps 1281 14.7 3.938 3.951 7.314 7.363 mp_alltoall_d11v 2401 14.1 6.907 6.999 6.907 6.999 calculate_dm_sparse 127 9.5 0.001 0.001 6.681 6.727 grid_collocate_task_list 127 9.7 6.371 6.453 6.371 6.453 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.336 6.408 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.154 6.244 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=290.658000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2880.363636, yerr=166.422940 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 1.251992E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 15114667. 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 0.052 0.126 86.205 86.208 qs_energies 1 2.0 0.001 0.009 85.298 85.318 ls_scf 1 3.0 0.002 0.010 84.369 84.380 dbcsr_multiply_generic 111 6.7 0.014 0.015 73.053 73.197 multiply_cannon 111 7.7 0.017 0.020 56.391 57.532 multiply_cannon_loop 111 8.7 0.231 0.248 53.012 54.405 ls_scf_main 1 4.0 0.002 0.017 52.302 52.316 density_matrix_trs4 2 5.0 0.002 0.011 46.746 46.827 ls_scf_init_scf 1 4.0 0.000 0.002 29.015 29.026 ls_scf_init_matrix_S 1 5.0 0.000 0.002 27.656 27.706 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.440 25.459 mp_waitall_1 11031 10.9 22.229 24.262 22.229 24.262 multiply_cannon_multrec 2664 9.7 8.143 8.908 15.607 17.363 multiply_cannon_sync_h2d 2664 9.7 13.666 15.312 13.666 15.312 make_m2s 222 7.7 0.008 0.011 13.073 13.492 make_images 222 8.7 0.098 0.108 13.050 13.472 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.466 12.229 make_images_data 222 9.7 0.004 0.005 7.625 8.253 dbcsr_mm_accdrv_process 4760 10.4 0.597 0.705 7.078 8.054 hybrid_alltoall_any 227 10.6 0.216 1.856 6.539 7.964 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.470 7.843 dbcsr_mm_accdrv_process_sort 4760 11.4 6.282 7.199 6.282 7.199 calculate_norms 4752 9.8 5.528 6.258 5.528 6.258 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.023 5.164 mp_sum_l 887 5.1 3.098 4.256 3.098 4.256 make_images_sizes 222 9.7 0.000 0.000 0.773 3.439 mp_alltoall_i44 222 10.7 0.772 3.439 0.772 3.439 arnoldi_extremal 4 6.8 0.000 0.000 3.193 3.224 arnoldi_normal_ev 4 7.8 0.001 0.003 3.193 3.224 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.065 3.143 mp_irecv_dv 6231 10.9 2.047 3.112 2.047 3.112 build_subspace 16 8.4 0.009 0.012 3.099 3.102 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.273 3.081 ls_scf_post 1 4.0 0.001 0.018 3.050 3.059 ls_scf_store_result 1 5.0 0.000 0.000 2.851 2.900 dbcsr_special_finalize 555 9.7 0.005 0.006 2.321 2.764 dbcsr_merge_single_wm 555 10.7 0.453 0.589 2.313 2.755 make_images_pack 222 9.7 2.207 2.628 2.209 2.630 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.308 2.549 dbcsr_sort_data 658 11.4 2.119 2.537 2.119 2.537 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.458 2.069 2.460 ls_scf_dm_to_ks 2 5.0 0.000 0.009 2.305 2.371 buffer_matrices_ensure_size 222 8.7 1.739 2.023 1.739 2.023 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.853 1.866 rebuild_ks_matrix 3 7.3 0.000 0.000 1.844 1.857 qs_ks_build_kohn_sham_matrix 3 8.3 0.021 0.063 1.844 1.857 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.208000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1124.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 2.122076E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.058 89.844 89.846 qs_energies 1 2.0 0.002 0.013 89.038 89.058 ls_scf 1 3.0 0.001 0.004 87.675 87.686 dbcsr_multiply_generic 111 6.7 0.015 0.016 73.910 74.230 multiply_cannon 111 7.7 0.028 0.043 52.624 56.734 ls_scf_main 1 4.0 0.001 0.006 54.201 54.207 multiply_cannon_loop 111 8.7 0.136 0.147 49.813 52.594 density_matrix_trs4 2 5.0 0.002 0.004 48.539 48.749 ls_scf_init_scf 1 4.0 0.000 0.002 29.778 29.782 mp_waitall_1 9105 10.9 20.506 29.181 20.506 29.181 ls_scf_init_matrix_S 1 5.0 0.000 0.002 28.576 28.689 multiply_cannon_multrec 1332 9.7 13.287 17.115 22.645 27.578 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.204 26.214 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.368 19.712 make_m2s 222 7.7 0.006 0.007 14.938 15.512 make_images 222 8.7 1.371 1.694 14.907 15.482 dbcsr_mm_accdrv_process 4041 10.4 0.323 0.542 8.949 10.527 dbcsr_mm_accdrv_process_sort 4041 11.4 8.466 9.986 8.466 9.986 make_images_data 222 9.7 0.004 0.005 8.646 9.491 hybrid_alltoall_any 227 10.6 0.540 2.561 8.005 9.380 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.190 7.721 mp_irecv_dv 3311 11.0 3.169 7.665 3.169 7.665 mp_sum_l 887 5.1 4.903 7.398 4.903 7.398 calculate_norms 2376 9.8 6.075 6.849 6.075 6.849 multiply_cannon_sync_h2d 1332 9.7 4.891 6.178 4.891 6.178 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.653 5.939 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.911 5.119 arnoldi_extremal 4 6.8 0.000 0.000 4.606 4.619 arnoldi_normal_ev 4 7.8 0.001 0.004 4.606 4.618 build_subspace 16 8.4 0.014 0.021 4.355 4.361 ls_scf_post 1 4.0 0.001 0.012 3.696 3.704 ls_scf_store_result 1 5.0 0.000 0.000 3.413 3.509 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.106 3.335 dbcsr_matrix_vector_mult_local 304 10.0 2.730 3.208 2.732 3.210 mp_allgather_i34 111 8.7 0.799 3.010 0.799 3.010 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.139 2.828 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.590 2.665 dbcsr_data_new 4174 10.1 2.114 2.398 2.114 2.398 make_images_pack 222 9.7 1.821 2.125 1.823 2.127 dbcsr_sort_data 436 11.2 1.858 2.053 1.858 2.053 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.850 1.855 rebuild_ks_matrix 3 7.3 0.000 0.000 1.837 1.842 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.006 1.837 1.842 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.846000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1782.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 2.847949E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.046 92.559 92.560 qs_energies 1 2.0 0.000 0.000 91.982 91.987 ls_scf 1 3.0 0.000 0.000 90.551 90.557 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.374 75.688 ls_scf_main 1 4.0 0.000 0.000 56.530 56.534 multiply_cannon 111 7.7 0.031 0.058 52.238 56.208 multiply_cannon_loop 111 8.7 0.116 0.129 49.452 53.371 density_matrix_trs4 2 5.0 0.002 0.003 50.683 50.901 mp_waitall_1 7281 11.0 23.660 32.916 23.660 32.916 ls_scf_init_scf 1 4.0 0.000 0.000 30.372 30.375 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.181 29.266 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.793 26.803 multiply_cannon_multrec 888 9.7 12.596 15.304 21.226 24.544 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.129 22.886 make_m2s 222 7.7 0.006 0.007 16.393 17.153 make_images 222 8.7 1.587 1.852 16.355 17.113 make_images_data 222 9.7 0.004 0.004 9.444 10.651 hybrid_alltoall_any 227 10.6 0.641 2.945 8.961 10.329 dbcsr_mm_accdrv_process 3754 10.4 0.302 0.494 8.166 9.425 dbcsr_mm_accdrv_process_sort 3754 11.4 7.727 8.932 7.727 8.932 mp_sum_l 887 5.1 5.037 8.463 5.037 8.463 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.732 7.406 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.502 7.235 multiply_cannon_sync_h2d 888 9.7 5.975 7.205 5.975 7.205 mp_irecv_dv 2335 11.1 2.487 7.178 2.487 7.178 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.946 6.900 arnoldi_extremal 4 6.8 0.000 0.000 5.048 5.062 arnoldi_normal_ev 4 7.8 0.001 0.005 5.048 5.062 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.733 4.934 calculate_norms 1584 9.8 4.427 4.839 4.427 4.839 build_subspace 16 8.4 0.014 0.020 4.748 4.753 mp_allgather_i34 111 8.7 0.888 3.813 0.888 3.813 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.409 3.740 ls_scf_post 1 4.0 0.000 0.000 3.648 3.655 dbcsr_matrix_vector_mult_local 304 10.0 3.008 3.580 3.010 3.582 ls_scf_store_result 1 5.0 0.000 0.000 3.395 3.462 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.820 2.916 dbcsr_data_new 4116 9.9 2.119 2.453 2.119 2.453 make_images_sizes 222 9.7 0.000 0.000 1.126 2.273 mp_alltoall_i44 222 10.7 1.125 2.272 1.125 2.272 dbcsr_sort_data 325 11.1 1.862 2.100 1.862 2.100 make_images_pack 222 9.7 1.625 1.881 1.628 1.884 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.878 1.880 rebuild_ks_matrix 3 7.3 0.000 0.000 1.860 1.862 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.860 1.862 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.560000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2279.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 3.392856E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.045 96.793 96.794 qs_energies 1 2.0 0.000 0.000 96.177 96.183 ls_scf 1 3.0 0.000 0.000 94.510 94.515 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.243 78.482 ls_scf_main 1 4.0 0.000 0.000 58.599 58.600 multiply_cannon 111 7.7 0.047 0.107 51.503 55.820 density_matrix_trs4 2 5.0 0.002 0.003 52.443 52.580 multiply_cannon_loop 111 8.7 0.152 0.166 46.444 48.842 ls_scf_init_scf 1 4.0 0.000 0.000 32.651 32.654 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.431 31.513 mp_waitall_1 6369 11.0 22.524 28.976 22.524 28.976 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.947 28.958 multiply_cannon_multrec 1332 9.7 14.109 17.196 22.090 24.719 make_m2s 222 7.7 0.006 0.007 21.121 22.591 make_images 222 8.7 3.150 3.601 21.071 22.544 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.059 17.054 make_images_data 222 9.7 0.004 0.004 11.799 13.361 hybrid_alltoall_any 227 10.6 0.798 3.787 11.083 12.999 dbcsr_mm_accdrv_process 3641 10.4 0.305 0.483 7.617 9.170 dbcsr_mm_accdrv_process_sort 3641 11.4 7.175 8.671 7.175 8.671 mp_sum_l 887 5.1 4.028 7.464 4.028 7.464 multiply_cannon_sync_h2d 1332 9.7 5.379 5.967 5.379 5.967 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.096 5.963 mp_irecv_dv 3229 10.9 2.072 5.897 2.072 5.897 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.093 5.887 arnoldi_extremal 4 6.8 0.000 0.000 5.209 5.220 arnoldi_normal_ev 4 7.8 0.001 0.005 5.209 5.220 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.497 4.918 build_subspace 16 8.4 0.014 0.021 4.873 4.882 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.573 4.771 calculate_norms 2376 9.8 4.221 4.578 4.221 4.578 mp_allgather_i34 111 8.7 2.170 4.535 2.170 4.535 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.571 3.874 dbcsr_matrix_vector_mult_local 304 10.0 3.175 3.673 3.177 3.675 dbcsr_sort_data 658 11.4 3.075 3.441 3.075 3.441 ls_scf_post 1 4.0 0.000 0.000 3.259 3.265 dbcsr_special_finalize 555 9.7 0.006 0.007 2.817 3.210 dbcsr_merge_single_wm 555 10.7 0.542 0.661 2.809 3.202 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.030 3.080 ls_scf_store_result 1 5.0 0.000 0.000 2.994 3.061 dbcsr_data_release 10477 10.7 1.599 2.431 1.599 2.431 dbcsr_finalize 304 7.8 0.049 0.061 1.815 1.988 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.794000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2804.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 4.760183E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.062 91.968 91.969 qs_energies 1 2.0 0.000 0.000 91.137 91.139 ls_scf 1 3.0 0.000 0.000 89.164 89.167 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.752 70.892 ls_scf_main 1 4.0 0.000 0.000 56.331 56.331 multiply_cannon 111 7.7 0.068 0.154 52.739 55.654 multiply_cannon_loop 111 8.7 0.088 0.094 50.183 51.413 density_matrix_trs4 2 5.0 0.002 0.003 49.340 49.426 ls_scf_init_scf 1 4.0 0.000 0.000 29.423 29.425 mp_waitall_1 5436 11.0 24.301 28.441 24.301 28.441 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.123 28.156 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.025 26.037 multiply_cannon_multrec 444 9.7 13.703 16.511 20.851 22.110 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.754 15.573 make_m2s 222 7.7 0.005 0.005 13.333 14.254 make_images 222 8.7 2.042 2.480 13.266 14.185 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.270 13.800 make_images_data 222 9.7 0.003 0.004 8.094 9.593 hybrid_alltoall_any 227 10.6 0.802 3.818 8.038 9.362 multiply_cannon_sync_h2d 444 9.7 6.759 8.208 6.759 8.208 dbcsr_mm_accdrv_process 3003 10.4 0.353 0.409 6.843 7.945 dbcsr_mm_accdrv_process_sort 3003 11.4 6.479 7.547 6.479 7.547 arnoldi_extremal 4 6.8 0.000 0.000 5.770 5.781 arnoldi_normal_ev 4 7.8 0.002 0.005 5.770 5.780 build_subspace 16 8.4 0.015 0.020 5.366 5.374 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.438 4.664 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.146 4.329 mp_sum_l 887 5.1 2.697 4.271 2.697 4.271 dbcsr_matrix_vector_mult_local 304 10.0 3.664 4.130 3.667 4.132 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.513 3.767 mp_irecv_dv 1241 11.2 1.498 3.722 1.498 3.722 calculate_norms 792 9.8 3.609 3.721 3.609 3.721 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.523 3.576 mp_allgather_i34 111 8.7 1.134 3.497 1.134 3.497 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.883 3.448 ls_scf_post 1 4.0 0.000 0.000 3.410 3.412 make_images_sizes 222 9.7 0.000 0.000 0.848 3.225 mp_alltoall_i44 222 10.7 0.848 3.225 0.848 3.225 ls_scf_store_result 1 5.0 0.000 0.000 3.181 3.216 dbcsr_finalize 304 7.8 0.062 0.077 2.196 2.310 dbcsr_data_new 4608 9.7 1.813 2.276 1.813 2.276 dbcsr_merge_all 275 8.9 0.479 0.535 2.054 2.149 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.091 2.092 rebuild_ks_matrix 3 7.3 0.000 0.000 2.059 2.060 qs_ks_build_kohn_sham_matrix 3 8.3 0.011 0.014 2.059 2.060 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 1.956 1.956 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.969000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3802.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_performance_tests/26/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 30.358840E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 339931 0.0% 0.0% 100.0% average stack size 0.0 0.0 28341.7 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 8.821875E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.071 0.098 107.943 107.943 qs_energies 1 2.0 0.000 0.000 106.556 106.559 ls_scf 1 3.0 0.000 0.000 103.592 103.595 dbcsr_multiply_generic 111 6.7 0.024 0.027 77.117 77.254 ls_scf_main 1 4.0 0.000 0.000 63.869 63.869 density_matrix_trs4 2 5.0 0.002 0.003 54.783 54.837 multiply_cannon 111 7.7 0.179 0.274 48.765 50.460 multiply_cannon_loop 111 8.7 0.099 0.099 45.805 48.204 ls_scf_init_scf 1 4.0 0.000 0.000 35.935 35.936 ls_scf_init_matrix_S 1 5.0 0.000 0.000 34.345 34.375 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 31.655 31.663 mp_waitall_1 4527 11.1 21.970 25.641 21.970 25.641 make_m2s 222 7.7 0.005 0.005 22.676 23.730 make_images 222 8.7 3.576 3.880 22.569 23.621 multiply_cannon_multrec 444 9.7 17.825 18.443 22.509 23.104 hybrid_alltoall_any 227 10.6 1.655 3.635 12.887 15.806 make_images_data 222 9.7 0.004 0.004 13.098 15.187 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.275 10.808 multiply_cannon_sync_h2d 444 9.7 8.787 8.854 8.787 8.854 arnoldi_extremal 4 6.8 0.000 0.000 7.368 7.383 arnoldi_normal_ev 4 7.8 0.008 0.027 7.368 7.383 build_subspace 16 8.4 0.026 0.036 6.785 6.795 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.439 5.571 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.205 5.299 dbcsr_matrix_vector_mult_local 304 10.0 4.975 5.264 4.978 5.266 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.862 5.124 dbcsr_mm_accdrv_process 1814 10.4 0.244 0.364 4.503 4.648 mp_sum_l 887 5.1 3.282 4.314 3.282 4.314 dbcsr_mm_accdrv_process_sort 1814 11.4 4.167 4.296 4.167 4.296 ls_scf_post 1 4.0 0.000 0.000 3.789 3.791 ls_scf_store_result 1 5.0 0.000 0.000 3.499 3.537 make_images_sizes 222 9.7 0.000 0.000 1.448 3.515 mp_alltoall_i44 222 10.7 1.448 3.514 1.448 3.514 mp_allgather_i34 111 8.7 1.063 3.386 1.063 3.386 calculate_norms 792 9.8 3.234 3.305 3.234 3.305 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.651 3.274 dbcsr_finalize 304 7.8 0.082 0.089 3.071 3.165 compute_matrix_preconditioner 1 6.0 0.002 0.002 2.990 2.997 dbcsr_merge_all 275 8.9 0.885 0.918 2.856 2.943 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.934 2.934 dbcsr_complete_redistribute 5 7.6 1.435 1.476 2.738 2.860 matrix_ls_to_qs 2 6.0 0.000 0.000 2.395 2.530 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.502 2.504 dbcsr_sort_data 325 11.1 2.436 2.491 2.436 2.491 rebuild_ks_matrix 3 7.3 0.000 0.000 2.436 2.438 qs_ks_build_kohn_sham_matrix 3 8.3 0.006 0.009 2.436 2.438 acc_transpose_blocks 444 9.7 0.003 0.003 0.509 2.430 dbcsr_data_new 6591 9.6 1.824 2.347 1.824 2.347 acc_transpose_blocks_kernels 444 10.7 0.006 0.006 0.426 2.345 jit_kernel_transpose 1 13.0 0.420 2.339 0.420 2.339 dbcsr_new_transposed 4 7.5 0.242 0.254 2.297 2.315 dbcsr_frobenius_norm 74 6.6 2.056 2.139 2.212 2.250 dbcsr_add_d 103 6.2 0.000 0.000 2.125 2.206 dbcsr_add_anytype 103 7.2 0.859 0.891 2.125 2.206 dbcsr_data_release 12724 10.6 1.977 2.193 1.977 2.193 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.943000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6968.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ac0226eb549c7ef1ea50d0597d545f29d4c8fc87_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 584.998912E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 4804590. 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 0.017 0.059 224.145 224.160 qs_mol_dyn_low 1 2.0 0.007 0.093 223.178 223.206 qs_forces 5 3.8 0.004 0.005 222.926 222.943 qs_energies 5 4.8 0.001 0.009 219.970 220.004 scf_env_do_scf 5 5.8 0.000 0.001 206.588 206.600 scf_env_do_scf_inner_loop 105 6.6 0.004 0.036 179.868 179.880 qs_scf_new_mos 105 7.6 0.000 0.001 140.751 140.915 qs_scf_loop_do_ot 105 8.6 0.001 0.001 140.750 140.915 dbcsr_multiply_generic 1445 12.2 0.129 0.138 131.576 132.126 ot_scf_mini 105 9.6 0.003 0.003 130.910 131.077 multiply_cannon 1445 13.2 0.273 0.285 112.772 114.925 multiply_cannon_loop 1445 14.2 2.836 2.973 111.160 112.582 velocity_verlet 4 3.0 0.018 0.095 106.053 106.060 ot_mini 105 10.6 0.001 0.001 59.825 59.984 multiply_cannon_multrec 69360 15.2 29.750 34.545 39.750 44.855 qs_ot_get_p 112 10.4 0.001 0.001 41.246 41.483 mp_waitall_1 488190 16.1 34.225 40.813 34.225 40.813 qs_ot_get_derivative 55 11.6 0.001 0.001 38.171 38.332 multiply_cannon_sync_h2d 69360 15.2 29.123 34.097 29.123 34.097 multiply_cannon_metrocomm3 69360 15.2 0.201 0.211 25.437 31.827 qs_ot_p2m_diag 40 11.0 0.020 0.031 30.210 30.308 rebuild_ks_matrix 110 8.4 0.000 0.000 28.552 28.752 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.032 28.552 28.752 cp_dbcsr_syevd 40 12.0 0.002 0.002 27.126 27.128 init_scf_loop 7 6.6 0.000 0.000 26.687 26.689 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.178 26.355 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.931 23.213 apply_single 62 13.6 0.000 0.000 22.931 23.213 cp_fm_syevd 40 13.0 0.000 0.000 22.126 22.256 prepare_preconditioner 7 7.6 0.000 0.000 21.860 21.893 make_preconditioner 7 8.6 0.000 0.000 21.860 21.893 ot_new_cg_direction 55 11.6 0.001 0.001 20.968 20.969 cp_fm_redistribute_end 40 14.0 8.719 17.396 8.723 17.397 cp_fm_syevd_base 40 14.0 8.667 17.347 8.667 17.347 qs_rho_update_rho_low 110 7.6 0.001 0.001 16.293 16.673 calculate_rho_elec 110 8.6 0.029 0.032 16.293 16.673 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.715 14.873 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.654 14.726 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.778 13.870 mp_sum_l 4764 12.2 12.123 12.978 12.123 12.978 calculate_dm_sparse 110 9.5 0.000 0.001 11.460 11.629 pw_transfer 1645 12.4 0.078 0.093 11.441 11.624 fft_wrap_pw1pw2 1425 13.5 0.012 0.014 11.303 11.489 density_rs2pw 110 9.6 0.005 0.006 10.594 11.000 dbcsr_mm_accdrv_process 154766 15.8 6.249 6.429 9.868 10.743 init_scf_run 5 5.8 0.000 0.000 10.692 10.693 scf_env_initial_rho_setup 5 6.8 0.001 0.002 10.692 10.693 qs_vxc_create 110 10.4 0.004 0.036 10.305 10.339 cp_fm_cholesky_invert 7 10.6 10.309 10.316 10.309 10.316 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.238 10.311 fft_wrap_pw1pw2_240 915 15.0 0.836 0.914 9.933 10.112 check_diag 80 13.5 8.683 8.935 9.420 9.550 fft3d_pb 915 16.0 2.398 2.595 8.112 8.284 sum_up_and_integrate 60 10.3 0.001 0.002 8.268 8.278 integrate_v_rspace 60 11.3 0.002 0.002 8.251 8.260 transfer_rs2pw 445 10.6 0.007 0.008 7.395 7.816 multiply_cannon_metrocomm1 69360 15.2 0.098 0.103 4.591 7.545 xc_rho_set_and_dset_create 110 12.4 0.078 0.125 7.269 7.508 make_full_single_inverse 7 9.6 0.001 0.002 6.930 6.963 xc_vxc_pw_create 60 11.3 0.039 0.049 6.911 6.946 acc_transpose_blocks 69360 15.2 0.356 0.374 6.318 6.853 make_m2s 2890 13.2 0.078 0.087 6.204 6.794 make_images 2890 14.2 0.236 0.256 6.097 6.687 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 6.189 6.205 xc_pw_derive 510 13.4 0.005 0.007 6.084 6.153 calculate_first_density_matrix 1 7.0 0.000 0.001 5.871 5.895 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 5.784 5.860 mp_alltoall_z22v 2340 17.7 5.107 5.393 5.107 5.393 multiply_cannon_metrocomm4 67915 15.2 0.190 0.210 2.030 5.169 mp_waitany 7680 13.5 4.499 4.958 4.499 4.958 mp_irecv_dv 143422 16.1 1.711 4.576 1.711 4.576 wfi_extrapolate 5 7.8 0.000 0.002 4.516 4.516 potential_pw2rs 60 12.3 0.002 0.003 4.474 4.487 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=224.160000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=554.800000, yerr=2.993326 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: ac0226eb549c7ef1ea50d0597d545f29d4c8fc87 Summary: empty Status: OK