=== 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: 5f6c7e94cf466f5c887cb267571013f7bc5f1114 ################# 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, # SPFFT 1.0.6, SPLA 1.5.5 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 09.03.2024 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-hdf5 --with-libvdwxc --with-pexsi --with-plumed --with-sirius=no --with-spfft --with-spla; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 #USE_DEEPMD := 2.2.7 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 #USE_SIRIUS := 7.5.2 USE_SPFFT := 1.0.6 USE_SPGLIB := 1.16.2 USE_SPLA := 1.5.5 # Only needed for SIRIUS #LIBVDWXC_VER := 0.4.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_DEEPMD),) USE_DEEPMD := $(strip $(USE_DEEPMD)) DEEPMD_INC := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/include DEEPMD_LIB := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/lib CFLAGS += -I$(DEEPMD_INC) DFLAGS += -D__DEEPMD LIBS += $(DEEPMD_LIB)/libdeepmd.so LIBS += $(DEEPMD_LIB)/libdeepmd_c.so LIBS += $(DEEPMD_LIB)/libdeepmd_cc.so LIBS += $(DEEPMD_LIB)/libdeepmd_dyn_cudart.so LIBS += $(DEEPMD_LIB)/libdeepmd_op.so LIBS += $(DEEPMD_LIB)/libdeepmd_op_cuda.so LIBS += $(DEEPMD_LIB)/libtensorflow_cc.so.2 LIBS += $(DEEPMD_LIB)/libtensorflow_framework.so.2 endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib ifeq ($(USE_ACC), yes) SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_SPFFT),) USE_SPFFT := $(strip $(USE_SPFFT)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/include ifeq ($(USE_ACC), yes) SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(USE_SPFFT)/lib endif CFLAGS += -I$(SPFFT_INC) DFLAGS += -D__SPFFT LIBS += $(SPFFT_LIB)/libspfft.a endif ifneq ($(USE_SPLA),) USE_SPLA := $(strip $(USE_SPLA)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPLA_LIB := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/lib/cuda else SPLA_LIB := $(INSTALL_PATH)/SpLA-$(USE_SPLA)/lib endif CFLAGS += -I$(SPLA_INC) DFLAGS += -D__SPLA LIBS += $(SPLA_LIB)/libspla.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/01 job id: 52201882 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/02 job id: 52201883 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/03 job id: 52201884 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/04 job id: 52201885 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/05 job id: 52201886 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/06 job id: 52201887 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/07 job id: 52201888 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/08 job id: 52201889 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/09 job id: 52201890 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/10 job id: 52201891 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/11 job id: 52201892 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/12 job id: 52201893 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/13 job id: 52201894 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/14 job id: 52201895 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/15 job id: 52201896 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/16 job id: 52201897 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/17 job id: 52201898 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/18 job id: 52201899 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/19 job id: 52201900 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/20 job id: 52201901 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/21 job id: 52201902 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/22 job id: 52201903 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/23 job id: 52201904 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/24 job id: 52201905 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/25 job id: 52201906 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/26 job id: 52201907 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/27 job id: 52201908 --- 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/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 0.034 127.963 127.964 farming_run 1 2.0 127.348 127.349 127.931 127.934 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.549681E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822089. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 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 0.007 0.028 110.013 110.014 qs_energies 1 2.0 0.001 0.001 109.823 109.824 mp2_main 1 3.0 0.000 0.000 108.008 108.010 mp2_gpw_main 1 4.0 0.029 0.039 107.115 107.116 mp2_ri_gpw_compute_in 1 5.0 0.174 0.176 88.409 88.623 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 57.157 57.372 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.173 43.596 48.568 get_2c_integrals 1 6.0 0.008 0.009 30.834 31.075 integrate_v_rspace 273 8.0 0.442 0.458 25.543 30.129 fft_wrap_pw1pw2 5465 10.4 0.067 0.073 28.013 28.639 grid_integrate_task_list 273 9.0 20.789 26.325 20.789 26.325 fft_wrap_pw1pw2_100 2178 11.4 1.321 1.448 25.450 26.113 compute_2c_integrals 1 7.0 0.002 0.002 21.434 21.435 compute_2c_integrals_loop_lm 1 8.0 0.002 0.004 20.806 21.165 mp2_eri_2c_integrate_gpw 1 9.0 2.406 2.465 20.805 21.161 rpa_ri_compute_en 1 5.0 0.019 0.020 18.585 18.724 fft3d_s 5443 12.4 16.839 17.410 16.859 17.430 ao_to_mo_and_store_B_mult_1 272 7.0 10.767 15.303 10.767 15.303 calculate_wavefunction 272 8.0 5.267 5.411 12.731 13.662 calc_potential_gpw 544 9.5 0.004 0.004 10.866 11.595 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 9.770 10.512 rpa_num_int 1 6.0 0.000 0.001 10.440 10.441 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.429 10.432 potential_pw2rs 545 10.0 0.106 0.108 8.434 10.290 cp_fm_cholesky_decompose 12 8.2 9.543 9.789 9.543 9.789 calc_mat_Q 8 8.0 0.000 0.000 9.429 9.506 contract_S_to_Q 8 9.0 0.000 0.000 8.852 8.930 cholesky_decomp 1 7.0 0.000 0.000 8.588 8.832 collocate_single_gaussian 272 10.0 0.039 0.041 7.825 8.599 parallel_gemm_fm 14 9.1 0.000 0.000 8.466 8.569 parallel_gemm_fm_cosma 14 10.1 8.466 8.569 8.466 8.569 create_integ_mat 1 6.0 0.023 0.028 7.840 7.840 array2fm 1 7.0 0.000 0.000 6.802 7.249 pw_poisson_solve 545 10.5 0.010 0.010 4.641 4.875 pw_scatter_s 2720 12.7 4.487 4.853 4.487 4.853 pw_gather_s 2722 12.2 3.887 4.837 3.887 4.837 pw_poisson_set 548 11.5 0.017 0.019 3.330 3.502 array2fm_buffer_send 1 8.0 2.970 3.194 2.970 3.194 pw_copy 4911 11.6 2.609 2.699 2.609 2.699 ao_to_mo_and_store_B_E_Ex_1 272 7.0 1.083 1.232 2.029 2.234 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=107.113499, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2817.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 0.023 0.038 426.265 426.266 farming_run 1 2.0 425.469 425.473 426.230 426.233 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.278202E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713945. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171443. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.030 222.846 222.846 qs_energies 1 2.0 0.003 0.023 222.607 222.614 scf_env_do_scf 1 3.0 0.000 0.000 114.895 114.895 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.985 113.992 rebuild_ks_matrix 4 6.0 0.000 0.000 113.984 113.991 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.062 113.984 113.991 hfx_ks_matrix 4 8.0 0.001 0.001 113.618 113.622 integrate_four_center 4 9.0 0.153 0.467 113.618 113.621 mp2_main 1 3.0 0.000 0.000 107.421 107.431 mp2_gpw_main 1 4.0 0.048 0.074 106.531 106.544 integrate_four_center_main 4 10.0 0.124 0.519 102.017 104.959 integrate_four_center_bin 269 11.0 101.892 104.764 101.892 104.764 init_scf_loop 1 4.0 0.000 0.000 96.990 96.990 mp2_ri_gpw_compute_in 1 5.0 0.066 0.078 79.798 80.909 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 56.767 57.875 mp2_eri_3c_integrate_gpw 91 7.0 0.147 0.163 44.642 49.535 integrate_v_rspace 95 8.0 0.398 0.565 28.334 33.063 fft_wrap_pw1pw2 1868 10.4 0.029 0.036 30.459 31.079 mp2_ri_gpw_compute_en 1 5.0 0.054 0.063 26.581 28.804 fft_wrap_pw1pw2_100 730 11.4 0.646 0.749 28.145 28.782 grid_integrate_task_list 95 9.0 23.617 28.545 23.617 28.545 ao_to_mo_and_store_B_mult_1 91 7.0 10.448 27.694 10.448 27.694 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.849 1.909 24.822 24.832 get_2c_integrals 1 6.0 0.000 0.000 22.939 22.968 compute_2c_integrals 1 7.0 0.002 0.003 22.107 22.115 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 21.758 21.955 mp2_eri_2c_integrate_gpw 1 9.0 1.735 1.837 21.757 21.954 fft3d_s 1823 12.4 19.198 20.230 19.212 20.243 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.903 17.903 calc_potential_gpw 182 9.5 0.002 0.002 12.414 13.232 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 10.762 11.346 calculate_wavefunction 91 8.0 1.978 2.029 9.992 10.797 potential_pw2rs 186 10.0 0.033 0.035 9.067 10.551 mp2_ri_gpw_compute_en_expansio 172 7.0 0.553 0.578 8.759 9.219 collocate_single_gaussian 91 10.0 0.018 0.025 8.277 8.878 local_gemm 172 8.0 8.205 8.659 8.205 8.659 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.520 7.832 8.152 mp_sync 37 10.5 3.579 7.037 3.579 7.037 integrate_four_center_load 4 10.0 0.000 0.000 6.756 6.760 hfx_load_balance 1 11.0 0.000 0.000 6.756 6.760 mp2_ri_gpw_compute_en_ener 172 7.0 6.337 6.400 6.337 6.400 mp_sendrecv_dm3 2068 8.0 5.868 6.221 5.868 6.221 pw_poisson_solve 186 10.4 0.005 0.006 5.633 6.185 pw_gather_s 912 12.2 4.775 5.460 4.775 5.460 pw_scatter_s 910 12.7 4.085 4.781 4.085 4.781 pw_poisson_set 189 11.4 0.009 0.010 4.168 4.655 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=106.531487, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1531.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 445.022208E+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 0.013 0.034 54.212 54.214 qs_mol_dyn_low 1 2.0 0.005 0.011 53.850 53.860 qs_forces 11 3.9 0.002 0.002 53.670 53.673 qs_energies 11 4.9 0.001 0.002 52.176 52.189 scf_env_do_scf 11 5.9 0.000 0.001 46.713 46.713 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.795 44.795 qs_scf_new_mos 108 7.5 0.000 0.001 34.763 35.049 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.763 35.048 dbcsr_multiply_generic 2286 12.5 0.093 0.097 34.276 34.743 ot_scf_mini 108 9.5 0.002 0.002 33.029 33.212 multiply_cannon 2286 13.5 0.191 0.204 26.505 28.039 multiply_cannon_loop 2286 14.5 1.831 1.934 25.831 27.329 velocity_verlet 10 3.0 0.028 0.111 27.065 27.067 ot_mini 108 10.5 0.001 0.001 20.071 20.329 qs_ot_get_derivative 108 11.5 0.001 0.002 16.976 17.122 mp_waitall_1 245248 16.5 8.784 14.937 8.784 14.937 multiply_cannon_metrocomm3 54864 15.5 0.073 0.078 6.270 13.218 multiply_cannon_multrec 54864 15.5 3.617 5.670 7.606 10.998 qs_ot_get_p 119 10.4 0.001 0.001 8.183 8.500 rebuild_ks_matrix 119 8.3 0.000 0.000 7.893 8.032 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.893 8.032 mp_sum_l 7287 12.8 5.744 7.465 5.744 7.465 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.943 7.082 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.280 6.718 multiply_cannon_sync_h2d 54864 15.5 5.146 6.519 5.146 6.519 dbcsr_mm_accdrv_process 76910 16.1 1.864 2.953 3.903 5.615 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.219 5.317 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.218 5.285 sum_up_and_integrate 119 10.3 0.001 0.002 4.425 4.432 integrate_v_rspace 119 11.3 0.002 0.002 4.414 4.422 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.351 4.351 init_scf_run 11 5.9 0.000 0.001 4.273 4.273 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.273 4.273 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.110 4.218 calculate_rho_elec 119 8.7 0.011 0.016 4.110 4.217 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.142 4.142 cp_fm_redistribute_end 50 14.0 2.115 4.115 2.121 4.117 cp_fm_diag_elpa_base 50 14.0 1.991 4.007 1.995 4.014 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.956 3.154 apply_single 119 13.6 0.000 0.000 2.956 3.153 calculate_dm_sparse 119 9.5 0.000 0.000 2.957 3.085 multiply_cannon_metrocomm1 54864 15.5 0.055 0.060 1.857 3.032 ot_diis_step 108 11.5 0.006 0.006 2.800 2.800 acc_transpose_blocks 54864 15.5 0.215 0.234 2.197 2.716 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.584 2.650 jit_kernel_multiply 13 15.8 1.977 2.569 1.977 2.569 calculate_first_density_matrix 1 7.0 0.001 0.005 2.345 2.357 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.214 2.229 density_rs2pw 119 9.7 0.003 0.004 2.112 2.209 grid_integrate_task_list 119 12.3 2.028 2.129 2.028 2.129 mp_sum_d 4139 12.0 1.417 2.066 1.417 2.066 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.905 1.963 init_scf_loop 11 6.9 0.000 0.000 1.900 1.900 wfi_extrapolate 11 7.9 0.001 0.001 1.849 1.849 potential_pw2rs 119 12.3 0.004 0.004 1.784 1.797 make_m2s 4572 13.5 0.053 0.057 1.721 1.781 make_images 4572 14.5 0.133 0.138 1.639 1.698 fft_wrap_pw1pw2 1201 11.6 0.011 0.012 1.561 1.629 mp_alltoall_d11v 2130 13.8 1.374 1.527 1.374 1.527 transfer_rs2pw 487 10.6 0.006 0.007 1.379 1.476 acc_transpose_blocks_sync 164592 16.5 1.200 1.449 1.200 1.449 grid_collocate_task_list 119 9.7 1.350 1.433 1.350 1.433 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.391 1.412 mp_waitany 12084 13.8 1.273 1.400 1.273 1.400 fft3d_ps 1201 13.6 0.364 0.463 1.336 1.395 transfer_pw2rs 487 13.2 0.005 0.006 1.329 1.338 fft_wrap_pw1pw2_140 487 12.2 0.049 0.054 1.202 1.271 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.816 1.228 mp_alltoall_i22 627 13.8 0.837 1.124 0.837 1.124 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.214000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=424.545455, yerr=1.157084 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 483.905536E+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 1587978. 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 0.013 0.040 39.238 39.265 qs_mol_dyn_low 1 2.0 0.005 0.009 38.986 39.014 qs_forces 11 3.9 0.003 0.009 38.914 38.941 qs_energies 11 4.9 0.004 0.018 37.238 37.267 scf_env_do_scf 11 5.9 0.001 0.002 31.726 31.749 scf_env_do_scf_inner_loop 108 6.5 0.002 0.009 29.463 29.484 dbcsr_multiply_generic 2286 12.5 0.103 0.126 22.173 22.527 qs_scf_new_mos 108 7.5 0.001 0.001 20.767 20.997 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.767 20.996 ot_scf_mini 108 9.5 0.003 0.003 19.848 20.017 multiply_cannon 2286 13.5 0.208 0.217 17.032 18.397 velocity_verlet 10 3.0 0.002 0.005 18.060 18.072 multiply_cannon_loop 2286 14.5 1.190 1.243 15.874 17.375 ot_mini 108 10.5 0.001 0.001 12.220 12.463 mp_waitall_1 200699 16.5 5.523 10.656 5.523 10.656 qs_ot_get_derivative 108 11.5 0.001 0.001 9.864 10.040 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.046 9.416 multiply_cannon_multrec 27432 15.5 1.824 4.145 6.262 9.112 rebuild_ks_matrix 119 8.3 0.000 0.000 6.782 6.932 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.017 6.782 6.932 dbcsr_mm_accdrv_process 47894 16.0 3.489 5.773 4.356 6.478 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.964 6.100 qs_ot_get_p 119 10.4 0.001 0.002 4.715 4.939 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.701 4.634 init_scf_run 11 5.9 0.000 0.001 4.249 4.253 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.249 4.253 mp_sum_l 7287 12.8 2.137 4.143 2.137 4.143 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.084 4.065 apply_single 119 13.6 0.000 0.000 3.084 4.064 sum_up_and_integrate 119 10.3 0.001 0.003 3.650 3.657 integrate_v_rspace 119 11.3 0.002 0.003 3.636 3.642 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.543 3.564 calculate_rho_elec 119 8.7 0.021 0.025 3.543 3.563 qs_ot_p2m_diag 50 11.0 0.009 0.014 3.071 3.090 make_m2s 4572 13.5 0.052 0.054 2.561 2.888 make_images 4572 14.5 0.205 0.243 2.472 2.802 calculate_first_density_matrix 1 7.0 0.001 0.002 2.760 2.763 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.670 2.672 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.325 2.414 cp_fm_diag_elpa 50 13.0 0.000 0.001 2.339 2.340 cp_fm_redistribute_end 50 14.0 1.184 2.309 1.188 2.313 ot_diis_step 108 11.5 0.011 0.011 2.308 2.310 multiply_cannon_sync_h2d 27432 15.5 1.721 2.256 1.721 2.256 cp_fm_diag_elpa_base 50 14.0 1.091 2.214 1.120 2.249 init_scf_loop 11 6.9 0.001 0.004 2.243 2.245 calculate_dm_sparse 119 9.5 0.000 0.000 2.134 2.212 jit_kernel_multiply 10 16.3 0.810 2.059 0.810 2.059 acc_transpose_blocks 27432 15.5 0.111 0.118 1.625 1.967 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.954 1.956 grid_integrate_task_list 119 12.3 1.847 1.936 1.847 1.936 density_rs2pw 119 9.7 0.004 0.004 1.838 1.929 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.718 1.757 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.658 1.690 make_images_data 4572 15.5 0.047 0.054 1.213 1.514 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.080 1.428 wfi_extrapolate 11 7.9 0.001 0.002 1.425 1.426 fft3d_ps 1201 13.6 0.511 0.565 1.376 1.406 potential_pw2rs 119 12.3 0.006 0.006 1.389 1.398 mp_alltoall_d11v 2130 13.8 1.224 1.373 1.224 1.373 grid_collocate_task_list 119 9.7 1.290 1.364 1.290 1.364 prepare_preconditioner 11 7.9 0.000 0.000 1.320 1.347 make_preconditioner 11 8.9 0.000 0.001 1.320 1.347 fft_wrap_pw1pw2_140 487 12.2 0.048 0.051 1.309 1.341 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.322 1.329 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.223 1.267 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.210 1.255 transfer_rs2pw 487 10.6 0.005 0.006 1.117 1.218 mp_sum_d 4139 12.0 0.563 1.073 0.563 1.073 mp_allgather_i34 2286 14.5 0.583 0.981 0.583 0.981 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.972 0.974 acc_transpose_blocks_sync 82296 16.5 0.812 0.932 0.812 0.932 transfer_pw2rs 487 13.2 0.004 0.005 0.904 0.912 acc_transpose_blocks_kernels 27432 16.5 0.188 0.277 0.677 0.894 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.875 0.889 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.265000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=460.818182, yerr=1.641406 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 515.547136E+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 0.033 0.048 32.822 32.824 qs_mol_dyn_low 1 2.0 0.005 0.012 32.479 32.488 qs_forces 11 3.9 0.008 0.012 32.414 32.418 qs_energies 11 4.9 0.019 0.032 30.841 30.850 scf_env_do_scf 11 5.9 0.001 0.002 25.861 25.861 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.553 23.555 dbcsr_multiply_generic 2286 12.5 0.092 0.094 17.095 17.174 qs_scf_new_mos 108 7.5 0.001 0.001 15.724 15.737 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.724 15.737 velocity_verlet 10 3.0 0.002 0.005 15.171 15.173 ot_scf_mini 108 9.5 0.002 0.003 14.962 14.977 multiply_cannon 2286 13.5 0.193 0.197 13.743 14.406 multiply_cannon_loop 2286 14.5 0.862 0.905 12.954 13.638 ot_mini 108 10.5 0.001 0.001 9.177 9.193 qs_ot_get_derivative 108 11.5 0.001 0.001 7.689 7.704 multiply_cannon_multrec 18288 15.5 1.879 2.918 6.980 7.378 rebuild_ks_matrix 119 8.3 0.000 0.000 6.115 6.129 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.114 6.129 dbcsr_mm_accdrv_process 38222 16.0 4.936 5.730 5.004 5.834 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.380 5.394 mp_waitall_1 158411 16.6 2.763 3.925 2.763 3.925 init_scf_run 11 5.9 0.000 0.001 3.703 3.703 scf_env_initial_rho_setup 11 6.9 0.004 0.029 3.702 3.702 sum_up_and_integrate 119 10.3 0.001 0.001 3.639 3.645 integrate_v_rspace 119 11.3 0.003 0.003 3.626 3.634 qs_ot_get_p 119 10.4 0.001 0.002 3.538 3.558 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.880 3.509 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.339 3.341 calculate_rho_elec 119 8.7 0.031 0.032 3.338 3.341 calculate_first_density_matrix 1 7.0 0.000 0.003 2.444 2.449 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.414 2.444 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.033 2.393 apply_single 119 13.6 0.000 0.000 2.033 2.392 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.375 2.380 init_scf_loop 11 6.9 0.001 0.004 2.290 2.292 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.068 2.069 make_m2s 4572 13.5 0.044 0.045 1.890 2.039 density_rs2pw 119 9.7 0.004 0.005 1.900 2.013 make_images 4572 14.5 0.191 0.203 1.804 1.952 grid_integrate_task_list 119 12.3 1.795 1.912 1.795 1.912 calculate_dm_sparse 119 9.5 0.000 0.001 1.856 1.865 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.797 1.809 cp_fm_diag_elpa_base 50 14.0 1.773 1.787 1.796 1.807 acc_transpose_blocks 18288 15.5 0.077 0.079 1.694 1.763 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.756 1.761 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.755 1.756 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.702 1.720 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.580 1.595 prepare_preconditioner 11 7.9 0.000 0.000 1.496 1.498 make_preconditioner 11 8.9 0.000 0.001 1.495 1.498 ot_diis_step 108 11.5 0.011 0.012 1.475 1.475 mp_sum_l 7287 12.8 1.088 1.470 1.088 1.470 potential_pw2rs 119 12.3 0.007 0.007 1.428 1.442 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.374 1.435 fft3d_ps 1201 13.6 0.536 0.555 1.388 1.405 fft_wrap_pw1pw2_140 487 12.2 0.060 0.063 1.346 1.362 grid_collocate_task_list 119 9.7 1.234 1.314 1.234 1.314 transfer_rs2pw 487 10.6 0.005 0.005 1.143 1.261 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.192 1.197 wfi_extrapolate 11 7.9 0.001 0.001 1.197 1.197 multiply_cannon_sync_h2d 18288 15.5 1.046 1.176 1.046 1.176 qs_energies_init_hamiltonians 11 5.9 0.007 0.011 1.004 1.013 make_images_data 4572 15.5 0.047 0.051 0.818 0.991 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.953 0.971 transfer_pw2rs 487 13.2 0.004 0.004 0.934 0.945 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.715 0.916 mp_alltoall_d11v 2130 13.8 0.729 0.889 0.729 0.889 acc_transpose_blocks_kernels 18288 16.5 0.217 0.225 0.859 0.871 acc_transpose_blocks_sync 54864 16.5 0.739 0.808 0.739 0.808 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.805 0.807 mp_alltoall_z22v 1201 15.6 0.698 0.770 0.698 0.770 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.674 0.736 mp_waitany 9880 13.7 0.555 0.676 0.555 0.676 transfer_rs2pw_140 130 11.5 0.116 0.124 0.549 0.663 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.824000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=490.727273, yerr=1.710444 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 553.009152E+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 3672 62653. MP_Allreduce 10304 342. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.028 35.186 35.186 qs_mol_dyn_low 1 2.0 0.003 0.004 35.020 35.029 qs_forces 11 3.9 0.002 0.002 34.959 34.960 qs_energies 11 4.9 0.001 0.002 33.264 33.270 scf_env_do_scf 11 5.9 0.000 0.001 28.120 28.120 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 25.363 25.364 dbcsr_multiply_generic 2286 12.5 0.099 0.102 19.634 19.755 qs_scf_new_mos 108 7.5 0.001 0.001 17.557 17.617 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.556 17.616 velocity_verlet 10 3.0 0.001 0.002 17.568 17.570 ot_scf_mini 108 9.5 0.002 0.003 16.532 16.590 multiply_cannon 2286 13.5 0.217 0.223 15.964 16.471 multiply_cannon_loop 2286 14.5 1.522 1.603 15.045 15.436 ot_mini 108 10.5 0.001 0.001 10.167 10.237 multiply_cannon_multrec 27432 15.5 2.462 3.117 9.043 9.351 qs_ot_get_derivative 108 11.5 0.001 0.001 8.317 8.376 dbcsr_mm_accdrv_process 47916 15.9 6.093 7.419 6.475 7.833 rebuild_ks_matrix 119 8.3 0.000 0.000 6.197 6.255 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.196 6.254 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.484 5.535 init_scf_run 11 5.9 0.000 0.001 3.821 3.822 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.821 3.821 qs_ot_get_p 119 10.4 0.001 0.001 3.530 3.610 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.996 3.492 sum_up_and_integrate 119 10.3 0.001 0.001 3.356 3.365 integrate_v_rspace 119 11.3 0.002 0.003 3.345 3.353 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.178 3.209 calculate_rho_elec 119 8.7 0.040 0.046 3.177 3.209 init_scf_loop 11 6.9 0.000 0.000 2.736 2.736 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.194 2.650 apply_single 119 13.6 0.000 0.000 2.194 2.650 acc_transpose_blocks 27432 15.5 0.114 0.117 2.381 2.501 calculate_first_density_matrix 1 7.0 0.000 0.000 2.459 2.460 make_m2s 4572 13.5 0.053 0.055 2.215 2.326 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.285 2.318 make_images 4572 14.5 0.272 0.332 2.108 2.216 calculate_dm_sparse 119 9.5 0.000 0.000 2.132 2.184 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.150 2.160 mp_waitall_1 137007 16.6 1.594 2.143 1.594 2.143 prepare_preconditioner 11 7.9 0.000 0.000 1.922 1.930 make_preconditioner 11 8.9 0.000 0.000 1.922 1.930 grid_integrate_task_list 119 12.3 1.820 1.916 1.820 1.916 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.664 1.868 ot_diis_step 108 11.5 0.012 0.012 1.814 1.814 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.806 1.807 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.753 1.754 density_rs2pw 119 9.7 0.003 0.004 1.637 1.738 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.667 1.700 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.634 1.647 mp_sum_l 7287 12.8 1.031 1.573 1.031 1.573 acc_transpose_blocks_sync 82296 16.5 1.417 1.533 1.417 1.533 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.523 1.524 cp_fm_redistribute_end 50 14.0 0.764 1.503 0.768 1.505 cp_fm_diag_elpa_base 50 14.0 0.711 1.438 0.735 1.479 fft_wrap_pw1pw2_140 487 12.2 0.071 0.072 1.384 1.418 fft3d_ps 1201 13.6 0.575 0.639 1.329 1.356 grid_collocate_task_list 119 9.7 1.247 1.341 1.247 1.341 wfi_extrapolate 11 7.9 0.001 0.001 1.311 1.311 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.273 1.287 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.236 1.246 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.683 1.173 potential_pw2rs 119 12.3 0.008 0.009 1.169 1.172 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.069 1.070 jit_kernel_multiply 6 16.1 0.318 1.069 0.318 1.069 transfer_rs2pw 487 10.6 0.005 0.005 0.859 0.957 make_images_data 4572 15.5 0.047 0.051 0.822 0.930 hybrid_alltoall_any 4725 16.4 0.065 0.156 0.701 0.879 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.798 0.878 mp_alltoall_d11v 2130 13.8 0.728 0.844 0.728 0.844 acc_transpose_blocks_kernels 27432 16.5 0.272 0.279 0.822 0.839 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.799 0.805 dbcsr_complete_redistribute 329 12.2 0.109 0.124 0.639 0.798 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.186000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=525.363636, yerr=2.672171 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 625.082368E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 62672. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 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 0.047 0.095 28.269 28.269 qs_mol_dyn_low 1 2.0 0.003 0.004 28.037 28.045 qs_forces 11 3.9 0.002 0.002 27.973 27.974 qs_energies 11 4.9 0.002 0.002 26.277 26.279 scf_env_do_scf 11 5.9 0.000 0.001 21.669 21.669 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.407 19.407 velocity_verlet 10 3.0 0.002 0.002 14.368 14.371 dbcsr_multiply_generic 2286 12.5 0.092 0.095 13.148 13.279 qs_scf_new_mos 108 7.5 0.001 0.001 11.952 11.974 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.951 11.973 ot_scf_mini 108 9.5 0.002 0.002 11.241 11.260 multiply_cannon 2286 13.5 0.223 0.231 10.552 11.145 multiply_cannon_loop 2286 14.5 0.644 0.658 9.633 9.884 ot_mini 108 10.5 0.001 0.001 6.530 6.556 multiply_cannon_multrec 9144 15.5 1.684 1.938 5.965 6.160 rebuild_ks_matrix 119 8.3 0.000 0.000 5.684 5.707 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 5.684 5.707 qs_ot_get_derivative 108 11.5 0.001 0.001 5.247 5.265 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.050 5.070 dbcsr_mm_accdrv_process 12550 15.8 3.639 4.191 4.169 4.234 sum_up_and_integrate 119 10.3 0.001 0.001 3.348 3.352 integrate_v_rspace 119 11.3 0.003 0.003 3.338 3.342 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.214 3.218 calculate_rho_elec 119 8.7 0.060 0.061 3.213 3.218 init_scf_run 11 5.9 0.000 0.001 3.172 3.172 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.172 3.172 qs_ot_get_p 119 10.4 0.001 0.001 2.826 2.853 init_scf_loop 11 6.9 0.000 0.000 2.241 2.242 calculate_first_density_matrix 1 7.0 0.000 0.000 2.058 2.059 mp_waitall_1 115863 16.7 1.509 2.034 1.509 2.034 make_m2s 4572 13.5 0.034 0.035 1.796 1.954 grid_integrate_task_list 119 12.3 1.869 1.943 1.869 1.943 make_images 4572 14.5 0.269 0.300 1.707 1.864 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.831 1.833 calculate_dm_sparse 119 9.5 0.000 0.000 1.740 1.756 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.711 1.728 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.720 1.728 density_rs2pw 119 9.7 0.003 0.003 1.597 1.700 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.590 1.591 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.538 1.539 prepare_preconditioner 11 7.9 0.000 0.000 1.516 1.519 make_preconditioner 11 8.9 0.000 0.000 1.516 1.519 acc_transpose_blocks 9144 15.5 0.041 0.042 1.420 1.452 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.412 1.437 fft_wrap_pw1pw2_140 487 12.2 0.094 0.096 1.424 1.433 grid_collocate_task_list 119 9.7 1.297 1.392 1.297 1.392 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.373 1.382 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.341 1.351 jit_kernel_multiply 8 15.7 0.492 1.345 0.492 1.345 fft3d_ps 1201 13.6 0.636 0.651 1.332 1.340 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.289 1.297 cp_fm_diag_elpa_base 50 14.0 1.262 1.280 1.288 1.296 ot_diis_step 108 11.5 0.013 0.013 1.268 1.268 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.220 1.260 apply_single 119 13.6 0.000 0.000 1.220 1.260 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.225 1.226 potential_pw2rs 119 12.3 0.010 0.010 1.159 1.161 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.128 1.134 wfi_extrapolate 11 7.9 0.001 0.001 1.063 1.063 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.768 0.999 make_images_data 4572 15.5 0.042 0.047 0.782 0.960 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.926 mp_alltoall_d11v 2130 13.8 0.810 0.895 0.810 0.895 transfer_rs2pw 487 10.6 0.004 0.005 0.781 0.872 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.817 0.824 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.754 0.757 cp_fm_cholesky_invert 11 10.9 0.743 0.748 0.743 0.748 acc_transpose_blocks_sync 27432 16.5 0.717 0.737 0.717 0.737 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.674 0.724 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.456 0.711 transfer_pw2rs 487 13.2 0.003 0.004 0.672 0.674 mp_allgather_i34 2286 14.5 0.237 0.663 0.237 0.663 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.647 0.658 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.288 0.612 mp_alltoall_z22v 1201 15.6 0.571 0.607 0.571 0.607 qs_create_task_list 11 7.9 0.000 0.001 0.550 0.575 generate_qs_task_list 11 8.9 0.199 0.223 0.550 0.575 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.269000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=592.818182, yerr=5.966851 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 781.762560E+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 0.038 0.045 41.058 41.058 qs_mol_dyn_low 1 2.0 0.003 0.004 40.737 40.747 qs_forces 11 3.9 0.002 0.002 40.670 40.671 qs_energies 11 4.9 0.002 0.002 38.681 38.687 scf_env_do_scf 11 5.9 0.001 0.001 32.899 32.899 scf_env_do_scf_inner_loop 108 6.5 0.004 0.008 26.907 26.909 velocity_verlet 10 3.0 0.002 0.002 22.352 22.358 dbcsr_multiply_generic 2286 12.5 0.101 0.104 19.286 19.489 qs_scf_new_mos 108 7.5 0.001 0.001 17.598 17.698 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.597 17.697 ot_scf_mini 108 9.5 0.002 0.002 16.418 16.525 multiply_cannon 2286 13.5 0.300 0.304 15.194 16.062 multiply_cannon_loop 2286 14.5 0.862 0.880 13.914 14.810 ot_mini 108 10.5 0.001 0.001 9.951 10.076 multiply_cannon_multrec 9144 15.5 3.490 5.088 8.878 8.955 qs_ot_get_derivative 108 11.5 0.001 0.001 7.872 7.979 rebuild_ks_matrix 119 8.3 0.000 0.000 7.066 7.220 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.014 7.065 7.220 dbcsr_mm_accdrv_process 12550 15.8 4.499 6.222 5.243 6.609 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.366 6.507 init_scf_loop 11 6.9 0.000 0.000 5.960 5.962 prepare_preconditioner 11 7.9 0.000 0.000 4.949 4.963 make_preconditioner 11 8.9 0.000 0.000 4.949 4.963 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.982 4.850 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.027 4.086 calculate_rho_elec 119 8.7 0.118 0.121 4.026 4.086 sum_up_and_integrate 119 10.3 0.001 0.001 3.731 3.736 integrate_v_rspace 119 11.3 0.003 0.003 3.720 3.726 init_scf_run 11 5.9 0.000 0.001 3.702 3.702 scf_env_initial_rho_setup 11 6.9 0.002 0.002 3.702 3.702 qs_ot_get_p 119 10.4 0.038 0.038 3.524 3.671 mp_waitall_1 94719 16.7 2.388 3.372 2.388 3.372 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.854 3.241 cp_fm_upper_to_full 72 14.2 2.058 2.932 2.058 2.932 make_m2s 4572 13.5 0.038 0.038 2.425 2.580 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.201 2.491 apply_single 119 13.6 0.000 0.000 2.201 2.491 make_images 4572 14.5 0.352 0.383 2.304 2.458 fft_wrap_pw1pw2 1201 11.6 0.014 0.014 2.415 2.421 calculate_dm_sparse 119 9.5 0.000 0.000 2.321 2.379 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.393 2.286 calculate_first_density_matrix 1 7.0 0.000 0.000 2.161 2.224 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.107 2.160 grid_integrate_task_list 119 12.3 2.096 2.122 2.096 2.122 dbcsr_complete_redistribute 329 12.2 0.258 0.265 1.556 2.086 ot_diis_step 108 11.5 0.014 0.015 2.049 2.050 fft_wrap_pw1pw2_140 487 12.2 0.179 0.180 2.041 2.048 qs_ot_p2m_diag 50 11.0 0.044 0.044 2.043 2.045 density_rs2pw 119 9.7 0.003 0.003 1.991 2.020 mp_sum_l 7287 12.8 1.144 1.925 1.144 1.925 fft3d_ps 1201 13.6 0.859 0.880 1.868 1.874 acc_transpose_blocks 9144 15.5 0.043 0.044 1.811 1.841 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.832 1.833 qs_energies_init_hamiltonians 11 5.9 0.014 0.028 1.798 1.799 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.225 1.761 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.682 1.683 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.565 1.610 grid_collocate_task_list 119 9.7 1.533 1.562 1.533 1.562 mp_alltoall_i22 627 13.8 0.943 1.522 0.943 1.522 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 0.961 1.496 wfi_extrapolate 11 7.9 0.001 0.001 1.412 1.412 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.341 1.363 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.361 1.361 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.303 1.360 cp_fm_diag_elpa_base 50 14.0 1.274 1.287 1.359 1.359 mp_alltoall_d11v 2130 13.8 1.254 1.332 1.254 1.332 hybrid_alltoall_any 4725 16.4 0.090 0.150 1.081 1.264 make_images_data 4572 15.5 0.046 0.048 1.034 1.234 potential_pw2rs 119 12.3 0.013 0.014 1.227 1.229 cp_fm_cholesky_invert 11 10.9 1.153 1.160 1.153 1.160 acc_transpose_blocks_sync 27432 16.5 1.097 1.130 1.097 1.130 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.102 1.119 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.994 1.030 jit_kernel_multiply 6 15.6 0.716 1.012 0.716 1.012 qs_create_task_list 11 7.9 0.000 0.001 0.935 0.944 generate_qs_task_list 11 8.9 0.372 0.392 0.935 0.944 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.910 0.924 mp_alltoall_z22v 1201 15.6 0.874 0.897 0.874 0.897 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.058000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=735.181818, yerr=14.179487 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 497.840128E+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 3521 65365. MP_Allreduce 9920 483. MP_Sync 52 MP_Alltoall 1938 1032136. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.057 84.843 84.845 qs_mol_dyn_low 1 2.0 0.006 0.017 84.468 84.480 qs_forces 11 3.9 0.003 0.003 84.331 84.333 qs_energies 11 4.9 0.002 0.003 81.447 81.477 scf_env_do_scf 11 5.9 0.000 0.001 72.716 72.717 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 67.407 67.408 dbcsr_multiply_generic 2055 12.4 0.106 0.109 53.305 53.593 qs_scf_new_mos 99 7.5 0.000 0.001 50.294 50.415 qs_scf_loop_do_ot 99 8.5 0.000 0.001 50.293 50.414 ot_scf_mini 99 9.5 0.002 0.002 47.790 47.864 multiply_cannon 2055 13.4 0.183 0.188 43.341 44.812 multiply_cannon_loop 2055 14.4 1.817 1.885 42.325 43.781 velocity_verlet 10 3.0 0.003 0.008 43.434 43.436 ot_mini 99 10.5 0.001 0.001 28.615 28.688 qs_ot_get_derivative 99 11.5 0.003 0.010 21.687 21.782 multiply_cannon_multrec 49320 15.4 11.052 11.859 17.211 18.045 rebuild_ks_matrix 110 8.3 0.000 0.000 14.563 14.667 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.016 14.562 14.667 mp_waitall_1 220248 16.4 12.020 13.659 12.020 13.659 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.773 12.869 qs_ot_get_p 110 10.4 0.001 0.001 10.452 10.594 multiply_cannon_sync_h2d 49320 15.4 9.317 9.907 9.317 9.907 multiply_cannon_metrocomm3 49320 15.4 0.081 0.087 7.277 8.831 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.661 8.239 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.213 7.636 apply_single 110 13.6 0.000 0.000 7.212 7.636 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.983 7.034 sum_up_and_integrate 110 10.3 0.002 0.003 6.968 6.982 integrate_v_rspace 110 11.3 0.003 0.003 6.943 6.962 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.738 6.788 init_scf_run 11 5.9 0.000 0.001 6.706 6.707 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.706 6.707 ot_diis_step 99 11.5 0.005 0.006 6.587 6.587 mp_sum_l 6594 12.7 5.491 6.569 5.491 6.569 dbcsr_mm_accdrv_process 87628 16.1 3.217 3.438 6.032 6.442 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.301 6.441 calculate_rho_elec 110 8.6 0.020 0.024 6.301 6.440 cp_dbcsr_syevd 48 12.0 0.002 0.003 6.013 6.013 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.514 5.515 cp_fm_redistribute_end 48 14.0 2.797 5.482 2.810 5.487 cp_fm_diag_elpa_base 48 14.0 2.664 5.350 2.673 5.366 init_scf_loop 11 6.9 0.000 0.000 5.276 5.276 make_m2s 4110 13.4 0.061 0.065 4.019 4.156 make_images 4110 14.4 0.178 0.192 3.923 4.063 wfi_extrapolate 11 7.9 0.001 0.001 4.033 4.033 calculate_dm_sparse 110 9.5 0.001 0.001 3.879 3.974 multiply_cannon_metrocomm1 49320 15.4 0.060 0.064 2.790 3.918 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 3.584 3.586 density_rs2pw 110 9.6 0.004 0.004 3.288 3.522 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.414 3.464 grid_integrate_task_list 110 12.3 3.264 3.405 3.264 3.405 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.186 3.238 prepare_preconditioner 11 7.9 0.000 0.000 3.094 3.120 make_preconditioner 11 8.9 0.000 0.000 3.094 3.120 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.858 2.904 mp_alltoall_d11v 2046 13.8 2.283 2.831 2.283 2.831 fft_wrap_pw1pw2 1111 11.6 0.014 0.016 2.728 2.792 acc_transpose_blocks 49320 15.4 0.209 0.222 2.623 2.767 potential_pw2rs 110 12.3 0.006 0.006 2.561 2.582 calculate_first_density_matrix 1 7.0 0.000 0.001 2.562 2.571 jit_kernel_multiply 13 15.9 2.536 2.561 2.536 2.561 fft_wrap_pw1pw2_140 451 12.1 0.092 0.097 2.258 2.328 fft3d_ps 1111 13.6 0.732 0.812 2.268 2.317 grid_collocate_task_list 110 9.6 2.157 2.285 2.157 2.285 transfer_rs2pw 451 10.6 0.006 0.006 1.975 2.220 mp_waitany 14300 13.8 1.818 2.147 1.818 2.147 mp_sum_d 3893 11.9 1.558 2.110 1.558 2.110 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.931 1.948 make_images_data 4110 15.4 0.043 0.047 1.798 1.926 hybrid_alltoall_any 4261 16.3 0.084 0.485 1.590 1.854 transfer_pw2rs 451 13.1 0.006 0.007 1.812 1.823 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.660 1.702 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.845000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=472.090909, yerr=3.117532 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 589.262848E+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 3521 65581. MP_Allreduce 9919 558. MP_Sync 52 MP_Alltoall 1717 2630535. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.028 68.013 68.014 qs_mol_dyn_low 1 2.0 0.004 0.007 67.794 67.806 qs_forces 11 3.9 0.003 0.003 67.718 67.719 qs_energies 11 4.9 0.002 0.002 64.383 64.388 scf_env_do_scf 11 5.9 0.000 0.001 55.913 55.913 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.000 49.000 dbcsr_multiply_generic 2055 12.4 0.112 0.115 37.960 38.181 velocity_verlet 10 3.0 0.002 0.007 35.373 35.375 qs_scf_new_mos 99 7.5 0.001 0.001 33.226 33.370 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.225 33.369 multiply_cannon 2055 13.4 0.224 0.243 31.257 32.593 ot_scf_mini 99 9.5 0.003 0.003 31.554 31.673 multiply_cannon_loop 2055 14.4 1.158 1.187 29.982 30.949 ot_mini 99 10.5 0.001 0.001 18.393 18.529 multiply_cannon_multrec 24660 15.4 6.956 8.341 14.013 15.507 rebuild_ks_matrix 110 8.3 0.000 0.001 13.594 13.718 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 13.593 13.718 qs_ot_get_derivative 99 11.5 0.001 0.001 12.639 12.762 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.959 12.071 mp_waitall_1 176588 16.5 7.538 10.181 7.538 10.181 multiply_cannon_metrocomm3 24660 15.4 0.071 0.072 5.099 7.909 multiply_cannon_sync_h2d 24660 15.4 6.340 7.393 6.340 7.393 dbcsr_mm_accdrv_process 52282 16.1 5.491 6.317 6.886 7.231 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.433 7.148 apply_single 110 13.6 0.000 0.001 6.433 7.148 init_scf_loop 11 6.9 0.000 0.000 6.872 6.872 qs_ot_get_p 110 10.4 0.001 0.001 6.508 6.673 sum_up_and_integrate 110 10.3 0.001 0.002 6.305 6.317 integrate_v_rspace 110 11.3 0.003 0.006 6.278 6.291 init_scf_run 11 5.9 0.000 0.001 6.122 6.123 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.122 6.123 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.741 5.753 calculate_rho_elec 110 8.6 0.039 0.047 5.741 5.753 ot_diis_step 99 11.5 0.010 0.010 5.708 5.708 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.742 5.433 prepare_preconditioner 11 7.9 0.000 0.000 4.854 4.877 make_preconditioner 11 8.9 0.000 0.000 4.854 4.877 make_m2s 4110 13.4 0.057 0.059 4.277 4.758 make_images 4110 14.4 0.404 0.457 4.165 4.642 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.484 4.590 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.558 4.580 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.095 4.096 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.541 3.542 cp_fm_redistribute_end 48 14.0 1.786 3.512 1.790 3.514 wfi_extrapolate 11 7.9 0.001 0.001 3.478 3.478 cp_fm_diag_elpa_base 48 14.0 1.662 3.376 1.717 3.460 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.310 3.378 grid_integrate_task_list 110 12.3 3.166 3.361 3.166 3.361 fft_wrap_pw1pw2 1111 11.6 0.014 0.017 3.144 3.295 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.250 3.252 density_rs2pw 110 9.6 0.004 0.004 2.985 3.183 calculate_dm_sparse 110 9.5 0.001 0.001 3.014 3.044 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.902 2.944 make_images_data 4110 15.4 0.048 0.052 2.324 2.834 hybrid_alltoall_any 4261 16.3 0.105 0.457 2.017 2.831 fft_wrap_pw1pw2_140 451 12.1 0.110 0.117 2.642 2.789 fft3d_ps 1111 13.6 1.070 1.278 2.523 2.658 calculate_first_density_matrix 1 7.0 0.000 0.000 2.547 2.550 mp_sum_l 6594 12.7 1.796 2.452 1.796 2.452 cp_fm_cholesky_invert 11 10.9 2.376 2.391 2.376 2.391 grid_collocate_task_list 110 9.6 2.171 2.314 2.171 2.314 potential_pw2rs 110 12.3 0.008 0.009 2.174 2.189 jit_kernel_multiply 11 16.2 1.032 2.106 1.032 2.106 acc_transpose_blocks 24660 15.4 0.113 0.117 1.991 2.030 mp_alltoall_d11v 2046 13.8 1.829 1.999 1.829 1.999 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.964 1.986 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.848 1.849 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.795 1.807 mp_allgather_i34 2055 14.4 0.586 1.641 0.586 1.641 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.581 1.596 multiply_cannon_metrocomm4 22605 15.4 0.078 0.082 0.799 1.593 transfer_rs2pw 451 10.6 0.007 0.007 1.451 1.584 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.386 1.490 mp_irecv_dv 57340 16.2 0.667 1.471 0.667 1.471 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.014000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=557.454545, yerr=6.733412 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 666.316800E+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 0.009 0.031 58.910 58.911 qs_mol_dyn_low 1 2.0 0.003 0.004 58.625 58.637 qs_forces 11 3.9 0.003 0.003 58.550 58.552 qs_energies 11 4.9 0.002 0.002 55.374 55.377 scf_env_do_scf 11 5.9 0.001 0.001 47.524 47.525 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.955 39.955 velocity_verlet 10 3.0 0.001 0.002 31.727 31.729 dbcsr_multiply_generic 2055 12.4 0.108 0.110 28.974 29.250 qs_scf_new_mos 99 7.5 0.001 0.001 25.517 25.622 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.516 25.621 ot_scf_mini 99 9.5 0.002 0.002 24.257 24.374 multiply_cannon 2055 13.4 0.212 0.221 22.586 23.728 multiply_cannon_loop 2055 14.4 0.818 0.849 21.362 22.528 ot_mini 99 10.5 0.001 0.001 13.923 14.043 rebuild_ks_matrix 110 8.3 0.000 0.000 12.156 12.316 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.156 12.316 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.700 10.845 multiply_cannon_multrec 16440 15.4 3.549 4.682 9.766 10.841 mp_waitall_1 139946 16.5 7.102 10.438 7.102 10.438 qs_ot_get_derivative 99 11.5 0.001 0.001 9.371 9.489 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.459 7.553 init_scf_loop 11 6.9 0.000 0.000 7.532 7.532 dbcsr_mm_accdrv_process 34862 16.1 5.382 5.809 6.061 6.195 sum_up_and_integrate 110 10.3 0.001 0.002 6.167 6.183 integrate_v_rspace 110 11.3 0.003 0.003 6.141 6.158 prepare_preconditioner 11 7.9 0.000 0.000 5.768 5.784 make_preconditioner 11 8.9 0.000 0.000 5.768 5.784 qs_ot_get_p 110 10.4 0.001 0.001 5.449 5.587 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.513 5.522 calculate_rho_elec 110 8.6 0.058 0.058 5.512 5.522 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.995 5.508 apply_single 110 13.6 0.000 0.000 4.995 5.508 init_scf_run 11 5.9 0.000 0.001 5.460 5.460 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.460 5.460 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.214 5.435 make_m2s 4110 13.4 0.050 0.051 4.259 4.676 make_images 4110 14.4 0.396 0.520 4.142 4.558 ot_diis_step 99 11.5 0.011 0.016 4.517 4.518 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.225 3.854 multiply_cannon_sync_h2d 16440 15.4 3.221 3.825 3.221 3.825 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.803 3.809 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.421 3.422 grid_integrate_task_list 110 12.3 3.213 3.368 3.213 3.368 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.065 3.074 wfi_extrapolate 11 7.9 0.001 0.001 2.956 2.956 make_images_data 4110 15.4 0.045 0.050 2.410 2.913 density_rs2pw 110 9.6 0.004 0.004 2.764 2.905 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.850 2.859 cp_fm_diag_elpa_base 48 14.0 2.809 2.827 2.848 2.858 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.855 2.856 hybrid_alltoall_any 4261 16.3 0.108 0.382 2.141 2.811 fft_wrap_pw1pw2_140 451 12.1 0.126 0.128 2.626 2.637 calculate_dm_sparse 110 9.5 0.001 0.001 2.539 2.576 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.488 2.553 multiply_cannon_metrocomm4 14385 15.4 0.048 0.052 0.906 2.529 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.446 2.505 cp_fm_cholesky_invert 11 10.9 2.407 2.420 2.407 2.420 calculate_first_density_matrix 1 7.0 0.000 0.000 2.412 2.414 grid_collocate_task_list 110 9.6 2.218 2.407 2.218 2.407 mp_irecv_dv 48980 15.7 0.830 2.393 0.830 2.393 fft3d_ps 1111 13.6 1.084 1.093 2.379 2.391 mp_sum_l 6594 12.7 1.509 2.158 1.509 2.158 mp_alltoall_d11v 2046 13.8 1.872 2.028 1.872 2.028 potential_pw2rs 110 12.3 0.009 0.010 2.009 2.014 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.973 1.974 dbcsr_complete_redistribute 325 12.2 0.317 0.356 1.496 1.799 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.655 1.671 acc_transpose_blocks 16440 15.4 0.075 0.077 1.582 1.659 mp_allgather_i34 2055 14.4 0.539 1.650 0.539 1.650 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.471 1.481 transfer_rs2pw 451 10.6 0.005 0.006 1.293 1.469 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.360 1.461 cp_fm_cholesky_decompose 22 10.9 1.349 1.375 1.349 1.375 mp_waitany 17072 13.8 1.134 1.333 1.134 1.333 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.291 1.299 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.938 1.236 transfer_pw2rs 451 13.1 0.005 0.005 1.175 1.181 cp_fm_upper_to_full 70 14.2 0.937 1.179 0.937 1.179 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=58.911000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=629.818182, yerr=9.504240 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 741.060608E+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 0.020 0.029 62.871 62.871 qs_mol_dyn_low 1 2.0 0.003 0.004 62.646 62.658 qs_forces 11 3.9 0.003 0.003 62.570 62.573 qs_energies 11 4.9 0.001 0.002 59.175 59.183 scf_env_do_scf 11 5.9 0.000 0.001 50.991 50.997 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.723 41.724 velocity_verlet 10 3.0 0.002 0.002 34.951 34.953 dbcsr_multiply_generic 2055 12.4 0.114 0.118 30.644 30.840 qs_scf_new_mos 99 7.5 0.001 0.001 27.303 27.406 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.303 27.405 ot_scf_mini 99 9.5 0.002 0.003 25.620 25.725 multiply_cannon 2055 13.4 0.237 0.257 23.683 24.893 multiply_cannon_loop 2055 14.4 1.411 1.445 22.099 22.643 ot_mini 99 10.5 0.001 0.001 14.731 14.867 multiply_cannon_multrec 24660 15.4 4.094 6.642 13.063 14.332 rebuild_ks_matrix 110 8.3 0.000 0.000 11.996 12.112 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.996 12.112 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.604 10.700 qs_ot_get_derivative 99 11.5 0.001 0.001 10.547 10.664 dbcsr_mm_accdrv_process 52304 16.0 7.766 9.082 8.810 9.737 init_scf_loop 11 6.9 0.000 0.000 9.226 9.226 prepare_preconditioner 11 7.9 0.000 0.000 7.465 7.483 make_preconditioner 11 8.9 0.000 0.000 7.465 7.483 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.333 7.128 sum_up_and_integrate 110 10.3 0.001 0.002 6.139 6.152 integrate_v_rspace 110 11.3 0.003 0.003 6.112 6.125 mp_waitall_1 121746 16.5 4.095 6.042 4.095 6.042 qs_ot_get_p 110 10.4 0.001 0.001 5.734 5.905 make_m2s 4110 13.4 0.059 0.061 5.361 5.872 make_images 4110 14.4 0.573 0.693 5.220 5.726 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.621 5.633 calculate_rho_elec 110 8.6 0.077 0.081 5.620 5.633 init_scf_run 11 5.9 0.000 0.001 5.444 5.445 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.444 5.445 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.062 4.150 apply_single 110 13.6 0.000 0.000 4.061 4.150 ot_diis_step 99 11.5 0.011 0.011 4.148 4.148 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.869 3.884 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.503 3.555 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.349 3.411 grid_integrate_task_list 110 12.3 3.296 3.401 3.296 3.401 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.304 3.305 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.195 3.218 make_images_data 4110 15.4 0.048 0.052 2.635 3.133 hybrid_alltoall_any 4261 16.3 0.121 0.457 2.259 3.122 calculate_dm_sparse 110 9.5 0.001 0.001 2.984 3.024 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.366 2.971 wfi_extrapolate 11 7.9 0.001 0.001 2.959 2.959 cp_fm_upper_to_full 70 14.2 2.059 2.894 2.059 2.894 density_rs2pw 110 9.6 0.004 0.004 2.734 2.874 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 2.840 2.842 dbcsr_complete_redistribute 325 12.2 0.380 0.446 2.094 2.792 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.749 2.760 cp_fm_diag_elpa_base 48 14.0 2.682 2.706 2.745 2.756 fft_wrap_pw1pw2_140 451 12.1 0.145 0.147 2.713 2.743 multiply_cannon_sync_h2d 24660 15.4 2.380 2.496 2.380 2.496 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.438 2.488 fft3d_ps 1111 13.6 1.124 1.162 2.466 2.488 acc_transpose_blocks 24660 15.4 0.106 0.109 2.407 2.484 cp_fm_cholesky_invert 11 10.9 2.440 2.451 2.440 2.451 grid_collocate_task_list 110 9.6 2.264 2.426 2.264 2.426 calculate_first_density_matrix 1 7.0 0.000 0.000 2.385 2.389 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.253 2.256 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.535 2.211 mp_alltoall_d11v 2046 13.8 1.843 2.121 1.843 2.121 potential_pw2rs 110 12.3 0.012 0.013 1.937 1.944 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.800 1.832 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.124 1.782 mp_alltoall_i22 605 13.7 1.008 1.717 1.008 1.717 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.610 1.711 mp_allgather_i34 2055 14.4 0.672 1.688 0.672 1.688 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.638 1.650 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.552 1.575 multiply_cannon_metrocomm4 20550 15.4 0.058 0.061 0.849 1.485 mp_sum_l 6594 12.7 0.914 1.472 0.914 1.472 acc_transpose_blocks_sync 73980 16.4 1.397 1.472 1.397 1.472 jit_kernel_multiply 8 15.9 0.703 1.446 0.703 1.446 mp_irecv_dv 62702 16.1 0.745 1.406 0.745 1.406 cp_fm_cholesky_decompose 22 10.9 1.354 1.400 1.354 1.400 transfer_rs2pw 451 10.6 0.005 0.006 1.202 1.353 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.196 1.278 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=62.871000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=699.000000, yerr=11.200649 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 856.006656E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 258767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 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 0.011 0.031 54.598 54.598 qs_mol_dyn_low 1 2.0 0.003 0.004 54.313 54.325 qs_forces 11 3.9 0.003 0.003 54.243 54.243 qs_energies 11 4.9 0.002 0.002 50.589 50.594 scf_env_do_scf 11 5.9 0.000 0.001 42.514 42.515 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.334 35.335 velocity_verlet 10 3.0 0.004 0.004 30.452 30.455 dbcsr_multiply_generic 2055 12.4 0.107 0.108 23.480 23.601 qs_scf_new_mos 99 7.5 0.001 0.001 21.088 21.155 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.087 21.154 ot_scf_mini 99 9.5 0.002 0.002 19.825 19.878 multiply_cannon 2055 13.4 0.237 0.245 17.800 19.127 multiply_cannon_loop 2055 14.4 0.605 0.629 16.527 16.769 rebuild_ks_matrix 110 8.3 0.000 0.000 11.548 11.597 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.547 11.597 ot_mini 99 10.5 0.001 0.001 10.929 10.987 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.270 10.318 multiply_cannon_multrec 8220 15.4 3.190 4.447 7.612 8.722 mp_waitall_1 103326 16.6 6.033 7.734 6.033 7.734 qs_ot_get_derivative 99 11.5 0.001 0.001 7.182 7.236 init_scf_loop 11 6.9 0.000 0.000 7.127 7.128 sum_up_and_integrate 110 10.3 0.001 0.002 6.084 6.096 integrate_v_rspace 110 11.3 0.003 0.003 6.057 6.069 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.700 5.713 calculate_rho_elec 110 8.6 0.115 0.116 5.699 5.713 prepare_preconditioner 11 7.9 0.000 0.000 5.486 5.498 make_preconditioner 11 8.9 0.000 0.000 5.486 5.498 dbcsr_mm_accdrv_process 17442 15.9 3.141 4.147 4.282 5.201 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.078 5.154 qs_ot_get_p 110 10.4 0.001 0.002 4.917 4.992 init_scf_run 11 5.9 0.000 0.001 4.896 4.896 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.896 4.896 make_m2s 4110 13.4 0.038 0.039 4.259 4.503 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 3.119 4.479 make_images 4110 14.4 0.643 0.704 4.129 4.368 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.782 3.813 apply_single 110 13.6 0.000 0.000 3.782 3.813 ot_diis_step 99 11.5 0.012 0.012 3.724 3.724 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.504 3.509 grid_integrate_task_list 110 12.3 3.387 3.494 3.387 3.494 fft_wrap_pw1pw2 1111 11.6 0.014 0.015 3.371 3.376 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.159 3.159 fft_wrap_pw1pw2_140 451 12.1 0.192 0.194 2.890 2.895 density_rs2pw 110 9.6 0.004 0.004 2.713 2.839 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.731 2.731 cp_fm_cholesky_invert 11 10.9 2.609 2.618 2.609 2.618 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.608 2.615 cp_fm_diag_elpa_base 48 14.0 2.551 2.577 2.606 2.613 make_images_data 4110 15.4 0.041 0.047 2.273 2.600 hybrid_alltoall_any 4261 16.3 0.200 0.846 2.199 2.599 grid_collocate_task_list 110 9.6 2.370 2.555 2.370 2.555 wfi_extrapolate 11 7.9 0.001 0.001 2.548 2.548 fft3d_ps 1111 13.6 1.270 1.285 2.530 2.542 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.532 2.533 calculate_dm_sparse 110 9.5 0.001 0.001 2.492 2.527 multiply_cannon_sync_h2d 8220 15.4 2.341 2.500 2.341 2.500 calculate_first_density_matrix 1 7.0 0.000 0.000 2.253 2.254 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.150 2.169 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.064 2.108 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.783 2.009 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.958 1.986 potential_pw2rs 110 12.3 0.015 0.015 1.866 1.869 mp_alltoall_d11v 2046 13.8 1.711 1.820 1.711 1.820 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.558 1.674 dbcsr_complete_redistribute 325 12.2 0.520 0.542 1.524 1.627 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.606 1.610 mp_allgather_i34 2055 14.4 0.448 1.543 0.448 1.543 cp_fm_cholesky_decompose 22 10.9 1.446 1.466 1.446 1.466 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.439 1.449 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.479 1.416 acc_transpose_blocks 8220 15.4 0.038 0.040 1.361 1.394 qs_create_task_list 11 7.9 0.036 0.037 1.271 1.370 mp_irecv_dv 24056 15.7 0.452 1.368 0.452 1.368 generate_qs_task_list 11 8.9 0.396 0.467 1.235 1.334 multiply_cannon_metrocomm1 8220 15.4 0.022 0.022 0.797 1.331 transfer_rs2pw 451 10.6 0.005 0.005 1.120 1.274 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.182 1.214 jit_kernel_multiply 7 15.5 0.827 1.188 0.827 1.188 mp_waitany 9240 13.8 1.034 1.184 1.034 1.184 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.598000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=806.636364, yerr=11.702935 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 1.387491E+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 0.014 0.032 78.966 78.967 qs_mol_dyn_low 1 2.0 0.003 0.004 78.674 78.684 qs_forces 11 3.9 0.003 0.003 78.598 78.599 qs_energies 11 4.9 0.002 0.002 74.509 74.512 scf_env_do_scf 11 5.9 0.000 0.001 64.674 64.674 velocity_verlet 10 3.0 0.002 0.004 48.530 48.537 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.951 43.953 dbcsr_multiply_generic 2055 12.4 0.121 0.125 29.812 29.866 qs_scf_new_mos 99 7.5 0.001 0.001 26.917 26.959 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.917 26.958 ot_scf_mini 99 9.5 0.002 0.002 25.092 25.128 multiply_cannon 2055 13.4 0.335 0.373 22.506 23.204 multiply_cannon_loop 2055 14.4 0.822 0.837 20.708 21.067 init_scf_loop 11 6.9 0.000 0.000 20.651 20.653 prepare_preconditioner 11 7.9 0.000 0.000 18.657 18.664 make_preconditioner 11 8.9 0.000 0.000 18.657 18.664 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.682 18.159 ot_mini 99 10.5 0.001 0.001 14.214 14.250 rebuild_ks_matrix 110 8.3 0.001 0.001 13.509 13.553 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.508 13.553 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.255 12.295 cp_fm_upper_to_full 70 14.2 8.273 11.851 8.273 11.851 multiply_cannon_multrec 8220 15.4 4.067 4.243 9.782 9.886 qs_ot_get_derivative 99 11.5 0.001 0.001 9.522 9.560 mp_waitall_1 84994 16.7 7.751 8.558 7.751 8.558 dbcsr_complete_redistribute 325 12.2 0.945 0.971 5.580 7.789 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.891 6.928 calculate_rho_elec 110 8.6 0.227 0.227 6.890 6.927 copy_fm_to_dbcsr 174 11.2 0.001 0.001 4.532 6.736 sum_up_and_integrate 110 10.3 0.002 0.002 6.611 6.625 integrate_v_rspace 110 11.3 0.004 0.004 6.583 6.596 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 3.959 6.140 make_m2s 4110 13.4 0.044 0.044 5.542 6.043 make_images 4110 14.4 0.874 0.917 5.352 5.852 mp_alltoall_i22 605 13.7 3.578 5.795 3.578 5.795 dbcsr_mm_accdrv_process 11614 15.7 3.888 4.147 5.566 5.752 init_scf_run 11 5.9 0.000 0.001 5.587 5.587 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.587 5.587 qs_ot_get_p 110 10.4 0.002 0.002 5.522 5.566 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.893 5.311 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.826 5.266 apply_single 110 13.6 0.000 0.000 4.825 5.266 ot_diis_step 99 11.5 0.015 0.015 4.668 4.668 fft_wrap_pw1pw2 1111 11.6 0.016 0.016 4.506 4.513 fft_wrap_pw1pw2_140 451 12.1 0.363 0.365 3.927 3.938 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.890 3.896 cp_fm_cholesky_invert 11 10.9 3.870 3.880 3.870 3.880 hybrid_alltoall_any 4261 16.3 0.265 0.576 2.987 3.736 grid_integrate_task_list 110 12.3 3.673 3.734 3.673 3.734 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.656 3.656 make_images_data 4110 15.4 0.044 0.047 3.008 3.638 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.050 3.504 fft3d_ps 1111 13.6 1.828 1.848 3.365 3.371 density_rs2pw 110 9.6 0.004 0.004 3.354 3.368 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.364 3.365 wfi_extrapolate 11 7.9 0.001 0.001 3.242 3.242 calculate_dm_sparse 110 9.5 0.001 0.001 3.198 3.218 multiply_cannon_sync_h2d 8220 15.4 3.135 3.159 3.135 3.159 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.893 2.896 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.744 2.744 cp_fm_diag_elpa_base 48 14.0 2.446 2.522 2.742 2.743 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.704 2.725 grid_collocate_task_list 110 9.6 2.664 2.682 2.664 2.682 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.381 2.396 calculate_first_density_matrix 1 7.0 0.000 0.000 2.232 2.233 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.172 2.230 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.092 2.189 mp_alltoall_d11v 2046 13.8 2.109 2.168 2.109 2.168 potential_pw2rs 110 12.3 0.021 0.021 2.140 2.146 qs_create_task_list 11 7.9 0.001 0.001 1.884 1.932 generate_qs_task_list 11 8.9 0.739 0.790 1.884 1.931 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.856 1.861 cp_fm_cholesky_decompose 22 10.9 1.788 1.816 1.788 1.816 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.772 1.792 acc_transpose_blocks 8220 15.4 0.040 0.041 1.713 1.738 jit_kernel_multiply 10 15.2 1.483 1.726 1.483 1.726 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.678 1.695 qs_ot_get_orbitals 99 10.5 0.000 0.000 1.645 1.666 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=78.967000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1254.363636, yerr=51.865688 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 629.420032E+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 4001 57774. MP_Allreduce 11082 797. MP_Sync 86 MP_Alltoall 2210 2528380. 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 0.014 0.035 203.049 203.050 qs_mol_dyn_low 1 2.0 0.005 0.016 202.604 202.620 qs_forces 11 3.9 0.005 0.005 202.494 202.497 qs_energies 11 4.9 0.003 0.004 196.873 196.895 scf_env_do_scf 11 5.9 0.001 0.001 180.621 180.638 scf_env_do_scf_inner_loop 116 6.6 0.003 0.007 161.733 161.735 dbcsr_multiply_generic 2485 12.5 0.175 0.178 124.727 125.352 qs_scf_new_mos 116 7.6 0.001 0.001 122.921 123.139 qs_scf_loop_do_ot 116 8.6 0.001 0.001 122.921 123.138 velocity_verlet 10 3.0 0.003 0.013 120.203 120.205 ot_scf_mini 116 9.6 0.003 0.003 116.242 116.449 multiply_cannon 2485 13.5 0.236 0.245 100.911 103.175 multiply_cannon_loop 2485 14.5 2.369 2.433 98.743 101.062 ot_mini 116 10.6 0.001 0.001 65.200 65.444 multiply_cannon_multrec 59640 15.5 31.609 33.498 41.356 42.840 qs_ot_get_derivative 116 11.6 0.001 0.001 40.523 40.724 rebuild_ks_matrix 127 8.3 0.001 0.001 33.495 33.910 qs_ks_build_kohn_sham_matrix 127 9.3 0.014 0.017 33.494 33.909 mp_waitall_1 264810 16.5 28.607 30.882 28.607 30.882 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.073 30.500 qs_ot_get_p 127 10.4 0.001 0.001 28.851 29.116 multiply_cannon_sync_h2d 59640 15.5 26.272 27.982 26.272 27.982 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.193 24.966 apply_single 127 13.6 0.001 0.001 24.193 24.965 ot_diis_step 116 11.6 0.007 0.008 24.450 24.452 qs_ot_p2m_diag 82 11.4 0.078 0.090 22.237 22.375 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 19.170 19.349 cp_dbcsr_syevd 82 12.4 0.005 0.005 19.213 19.213 init_scf_loop 11 6.9 0.000 0.000 18.800 18.802 multiply_cannon_metrocomm3 59640 15.5 0.121 0.127 15.924 17.563 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.164 16.189 cp_fm_diag_elpa_base 82 14.4 16.114 16.140 16.158 16.182 prepare_preconditioner 11 7.9 0.000 0.000 14.152 14.205 make_preconditioner 11 8.9 0.000 0.000 14.152 14.205 make_m2s 4970 13.5 0.104 0.115 13.769 14.144 make_images 4970 14.5 0.396 0.416 13.585 13.973 sum_up_and_integrate 127 10.3 0.002 0.004 13.698 13.711 integrate_v_rspace 127 11.3 0.003 0.004 13.640 13.656 make_full_inverse_cholesky 11 9.9 0.000 0.000 13.426 13.593 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.565 12.709 calculate_rho_elec 127 8.7 0.044 0.063 12.565 12.708 init_scf_run 11 5.9 0.000 0.001 12.152 12.152 scf_env_initial_rho_setup 11 6.9 0.018 0.032 12.152 12.152 mp_sum_l 7884 12.9 8.806 10.295 8.806 10.295 dbcsr_mm_accdrv_process 123452 16.2 4.722 4.889 9.318 9.802 wfi_extrapolate 11 7.9 0.001 0.001 8.892 8.892 calculate_dm_sparse 127 9.5 0.001 0.001 8.483 8.589 qs_ot_get_orbitals 116 10.6 0.001 0.001 8.158 8.289 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.124 8.264 cp_fm_cholesky_invert 11 10.9 8.096 8.115 8.096 8.115 multiply_cannon_metrocomm1 59640 15.5 0.090 0.094 6.218 7.905 make_images_data 4970 15.5 0.065 0.070 6.723 7.609 grid_integrate_task_list 127 12.3 7.021 7.476 7.021 7.476 hybrid_alltoall_any 5155 16.4 0.293 2.261 5.897 7.267 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 6.742 6.755 fft_wrap_pw1pw2 1281 11.7 0.018 0.025 6.438 6.706 density_rs2pw 127 9.7 0.005 0.006 6.242 6.681 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.815 5.957 fft_wrap_pw1pw2_140 519 12.2 0.230 0.239 5.553 5.731 fft3d_ps 1281 13.7 2.075 2.583 5.274 5.481 mp_alltoall_d11v 2401 14.1 4.456 5.201 4.456 5.201 grid_collocate_task_list 127 9.7 4.809 5.200 4.809 5.200 mp_sum_d 4455 12.1 3.461 4.290 3.461 4.290 potential_pw2rs 127 12.3 0.009 0.010 4.259 4.273 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=203.050000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.000000, yerr=6.508735 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/16/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 2.166472E+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 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 851.255296E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58142. MP_Allreduce 11096 960. MP_Sync 86 MP_Alltoall 1955 4800878. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.032 182.108 182.109 qs_mol_dyn_low 1 2.0 0.004 0.010 181.744 181.761 qs_forces 11 3.9 0.004 0.005 181.652 181.654 qs_energies 11 4.9 0.002 0.003 174.878 174.889 scf_env_do_scf 11 5.9 0.001 0.001 158.673 158.690 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 132.475 132.478 velocity_verlet 10 3.0 0.003 0.010 111.599 111.601 dbcsr_multiply_generic 2485 12.5 0.185 0.192 96.336 97.714 qs_scf_new_mos 116 7.6 0.001 0.001 94.479 95.045 qs_scf_loop_do_ot 116 8.6 0.001 0.001 94.478 95.044 ot_scf_mini 116 9.6 0.004 0.004 89.718 90.352 multiply_cannon 2485 13.5 0.498 0.551 76.336 80.466 multiply_cannon_loop 2485 14.5 1.554 1.619 73.411 76.278 ot_mini 116 10.6 0.001 0.001 50.037 50.593 mp_waitall_1 212858 16.6 23.888 37.775 23.888 37.775 multiply_cannon_multrec 29820 15.5 20.811 25.429 31.388 36.436 rebuild_ks_matrix 127 8.3 0.001 0.001 32.128 32.855 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 32.127 32.854 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.878 29.553 multiply_cannon_metrocomm3 29820 15.5 0.095 0.101 15.521 28.857 qs_ot_get_derivative 116 11.6 0.001 0.002 28.230 28.841 init_scf_loop 11 6.9 0.000 0.000 26.101 26.103 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.905 23.107 apply_single 127 13.6 0.001 0.001 21.905 23.107 qs_ot_get_p 127 10.4 0.001 0.002 21.978 22.623 prepare_preconditioner 11 7.9 0.000 0.000 21.785 21.847 make_preconditioner 11 8.9 0.000 0.000 21.785 21.847 ot_diis_step 116 11.6 0.014 0.015 21.625 21.627 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.640 21.054 multiply_cannon_sync_h2d 29820 15.5 17.875 19.752 17.875 19.752 qs_ot_p2m_diag 82 11.4 0.186 0.214 17.077 17.111 cp_dbcsr_syevd 82 12.4 0.005 0.006 15.870 15.871 make_m2s 4970 13.5 0.086 0.091 14.089 15.280 make_images 4970 14.5 1.150 1.355 13.875 15.065 sum_up_and_integrate 127 10.3 0.002 0.004 13.685 13.715 integrate_v_rspace 127 11.3 0.003 0.004 13.626 13.659 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.941 12.974 calculate_rho_elec 127 8.7 0.087 0.105 12.940 12.973 cp_fm_diag_elpa 82 13.4 0.000 0.000 12.325 12.362 cp_fm_diag_elpa_base 82 14.4 12.148 12.188 12.315 12.347 cp_fm_cholesky_invert 11 10.9 11.962 11.997 11.962 11.997 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.209 11.674 init_scf_run 11 5.9 0.000 0.001 11.424 11.425 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.423 11.425 multiply_cannon_metrocomm4 27335 15.5 0.105 0.119 3.835 10.830 dbcsr_mm_accdrv_process 61726 16.2 5.501 6.179 10.028 10.581 mp_irecv_dv 68888 16.3 3.626 10.414 3.626 10.414 make_images_data 4970 15.5 0.065 0.075 8.300 9.816 hybrid_alltoall_any 5155 16.4 0.346 1.505 7.208 9.452 wfi_extrapolate 11 7.9 0.001 0.001 8.213 8.213 grid_integrate_task_list 127 12.3 7.158 7.506 7.158 7.506 fft_wrap_pw1pw2 1281 11.7 0.019 0.026 7.377 7.462 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.340 7.160 density_rs2pw 127 9.7 0.005 0.006 6.516 6.817 calculate_dm_sparse 127 9.5 0.001 0.001 6.449 6.583 fft_wrap_pw1pw2_140 519 12.2 0.245 0.254 6.456 6.519 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.236 6.244 mp_sum_l 7884 12.9 4.129 6.006 4.129 6.006 fft3d_ps 1281 13.7 2.740 2.919 5.842 5.909 cp_fm_cholesky_decompose 22 10.9 5.619 5.693 5.619 5.693 mp_alltoall_d11v 2401 14.1 5.070 5.622 5.070 5.622 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.354 5.427 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.215 5.400 grid_collocate_task_list 127 9.7 4.963 5.306 4.963 5.306 mp_allgather_i34 2485 14.5 1.459 4.439 1.459 4.439 potential_pw2rs 127 12.3 0.014 0.015 4.227 4.248 mp_sum_d 4458 12.1 2.627 4.111 2.627 4.111 dbcsr_complete_redistribute 393 12.7 0.721 0.891 3.333 4.037 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.660 3.722 dbcsr_dot_sd 1305 12.0 0.763 0.842 2.463 3.704 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=182.109000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=811.090909, yerr=2.874798 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 1.033757E+09 # 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 0.012 0.030 171.340 171.341 qs_mol_dyn_low 1 2.0 0.004 0.004 170.914 170.929 qs_forces 11 3.9 0.004 0.004 170.816 170.817 qs_energies 11 4.9 0.002 0.002 164.275 164.285 scf_env_do_scf 11 5.9 0.001 0.001 148.901 148.903 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 119.909 119.910 velocity_verlet 10 3.0 0.001 0.002 108.080 108.083 dbcsr_multiply_generic 2527 12.6 0.186 0.190 82.599 83.725 qs_scf_new_mos 118 7.6 0.001 0.001 82.494 82.852 qs_scf_loop_do_ot 118 8.6 0.001 0.001 82.493 82.852 ot_scf_mini 118 9.6 0.003 0.004 78.270 78.628 multiply_cannon 2527 13.6 0.509 0.529 62.596 67.623 multiply_cannon_loop 2527 14.6 1.140 1.206 59.933 62.512 ot_mini 118 10.6 0.001 0.001 42.991 43.354 mp_waitall_1 171878 16.6 25.741 35.278 25.741 35.278 rebuild_ks_matrix 129 8.3 0.001 0.001 30.733 31.199 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.017 30.732 31.198 init_scf_loop 11 6.9 0.000 0.000 28.885 28.886 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.773 28.203 multiply_cannon_multrec 20216 15.6 13.071 16.085 22.488 25.438 multiply_cannon_metrocomm3 20216 15.6 0.061 0.065 15.957 25.231 prepare_preconditioner 11 7.9 0.000 0.000 24.877 24.935 make_preconditioner 11 8.9 0.000 0.000 24.877 24.935 make_full_inverse_cholesky 11 9.9 0.000 0.000 22.919 23.858 qs_ot_get_derivative 118 11.6 0.001 0.002 23.194 23.547 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.751 20.957 apply_single 129 13.6 0.001 0.001 19.751 20.957 qs_ot_get_p 129 10.4 0.001 0.002 20.419 20.855 ot_diis_step 118 11.6 0.018 0.018 19.679 19.679 qs_ot_p2m_diag 83 11.4 0.266 0.272 16.101 16.120 make_m2s 5054 13.6 0.079 0.084 14.959 15.684 multiply_cannon_sync_h2d 20216 15.6 13.588 15.596 13.588 15.596 make_images 5054 14.6 1.145 1.231 14.721 15.443 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.021 15.022 sum_up_and_integrate 129 10.3 0.002 0.003 13.966 13.991 integrate_v_rspace 129 11.3 0.003 0.004 13.905 13.929 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.336 13.370 calculate_rho_elec 129 8.7 0.133 0.147 13.336 13.370 cp_fm_cholesky_invert 11 10.9 11.934 11.954 11.934 11.954 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.466 11.493 cp_fm_diag_elpa_base 83 14.4 11.212 11.271 11.463 11.489 init_scf_run 11 5.9 0.000 0.001 10.418 10.418 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.418 10.418 make_images_data 5054 15.6 0.062 0.073 9.040 10.122 multiply_cannon_metrocomm4 17689 15.6 0.067 0.076 3.546 9.789 mp_irecv_dv 50625 16.2 3.414 9.512 3.414 9.512 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.145 9.408 hybrid_alltoall_any 5240 16.5 0.454 2.080 7.876 9.312 dbcsr_mm_accdrv_process 41830 16.2 5.750 6.108 8.884 9.043 grid_integrate_task_list 129 12.3 7.357 7.921 7.357 7.921 fft_wrap_pw1pw2 1301 11.7 0.019 0.025 7.645 7.770 wfi_extrapolate 11 7.9 0.001 0.001 7.405 7.405 density_rs2pw 129 9.7 0.006 0.006 6.735 7.050 fft_wrap_pw1pw2_140 527 12.2 0.284 0.295 6.676 6.829 cp_fm_cholesky_decompose 22 10.9 6.027 6.063 6.027 6.063 fft3d_ps 1301 13.7 2.786 3.014 5.942 6.022 calculate_dm_sparse 129 9.5 0.001 0.001 5.826 5.896 dbcsr_complete_redistribute 395 12.7 1.036 1.103 4.502 5.700 grid_collocate_task_list 129 9.7 5.192 5.647 5.192 5.647 mp_alltoall_d11v 2423 14.1 4.714 5.594 4.714 5.594 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.509 5.517 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.792 5.378 cp_fm_upper_to_full 105 14.8 3.910 4.898 3.910 4.898 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.588 4.711 mp_sum_l 8010 12.9 3.202 4.627 3.202 4.627 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.263 4.450 potential_pw2rs 129 12.3 0.020 0.022 4.357 4.386 mp_allgather_i34 2527 14.6 1.162 4.357 1.162 4.357 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.048 4.075 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.703 3.704 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.474 3.520 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=171.341000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=982.272727, yerr=3.792119 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 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 1.117098E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379345520 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58348. MP_Allreduce 11057 1083. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.045 177.543 177.543 qs_mol_dyn_low 1 2.0 0.003 0.004 177.145 177.167 qs_forces 11 3.9 0.004 0.004 177.024 177.034 qs_energies 11 4.9 0.026 0.031 169.921 169.934 scf_env_do_scf 11 5.9 0.001 0.001 153.123 153.138 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 115.925 115.926 velocity_verlet 10 3.0 0.003 0.003 115.786 115.789 dbcsr_multiply_generic 2507 12.6 0.190 0.195 80.520 81.102 qs_scf_new_mos 117 7.6 0.001 0.001 80.157 80.497 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.156 80.497 ot_scf_mini 117 9.6 0.003 0.003 75.596 75.944 multiply_cannon 2507 13.6 0.547 0.585 55.647 58.470 multiply_cannon_loop 2507 14.6 1.812 1.873 52.141 53.469 ot_mini 117 10.6 0.001 0.001 42.856 43.191 init_scf_loop 11 6.9 0.000 0.000 37.069 37.070 prepare_preconditioner 11 7.9 0.000 0.000 32.990 33.032 make_preconditioner 11 8.9 0.000 0.000 32.990 33.032 make_full_inverse_cholesky 11 9.9 0.017 0.023 28.218 31.795 multiply_cannon_multrec 30084 15.6 13.597 18.974 26.557 31.665 rebuild_ks_matrix 128 8.3 0.001 0.001 29.280 29.537 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 29.280 29.536 mp_waitall_1 147882 16.7 17.138 27.092 17.138 27.092 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.425 26.658 qs_ot_get_derivative 117 11.6 0.001 0.002 23.054 23.399 make_m2s 5014 13.6 0.094 0.099 20.527 21.453 make_images 5014 14.6 1.994 2.285 20.216 21.139 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.279 19.792 apply_single 128 13.6 0.001 0.001 19.279 19.792 ot_diis_step 117 11.6 0.018 0.020 19.682 19.683 qs_ot_get_p 128 10.4 0.002 0.002 18.837 19.211 multiply_cannon_metrocomm3 30084 15.6 0.050 0.052 6.344 14.912 qs_ot_p2m_diag 83 11.4 0.343 0.390 14.689 14.738 sum_up_and_integrate 128 10.3 0.002 0.003 13.697 13.723 integrate_v_rspace 128 11.3 0.003 0.004 13.637 13.664 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.277 13.278 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.209 13.264 calculate_rho_elec 128 8.7 0.174 0.189 13.209 13.264 dbcsr_mm_accdrv_process 62264 16.2 8.416 9.344 12.527 12.978 make_images_data 5014 15.6 0.064 0.071 10.951 12.388 cp_fm_upper_to_full 105 14.8 8.707 12.388 8.707 12.388 cp_fm_cholesky_invert 11 10.9 12.170 12.192 12.170 12.192 hybrid_alltoall_any 5200 16.5 0.528 2.210 9.979 11.875 multiply_cannon_sync_h2d 30084 15.6 10.519 11.218 10.519 11.218 init_scf_run 11 5.9 0.000 0.001 10.720 10.721 scf_env_initial_rho_setup 11 6.9 0.017 0.023 10.720 10.721 dbcsr_complete_redistribute 395 12.7 1.415 1.530 7.637 10.490 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.029 10.050 cp_fm_diag_elpa_base 83 14.4 9.534 9.671 10.021 10.036 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.692 9.952 copy_fm_to_dbcsr 209 11.7 0.001 0.002 6.217 9.040 grid_integrate_task_list 128 12.3 7.546 7.901 7.546 7.901 fft_wrap_pw1pw2 1291 11.7 0.019 0.023 7.763 7.847 wfi_extrapolate 11 7.9 0.001 0.001 7.597 7.597 multiply_cannon_metrocomm4 25070 15.6 0.079 0.088 2.797 7.501 transfer_fm_to_dbcsr 11 9.9 0.002 0.007 4.749 7.460 mp_irecv_dv 76098 16.2 2.642 7.214 2.642 7.214 fft_wrap_pw1pw2_140 523 12.2 0.324 0.333 6.851 6.970 mp_alltoall_i22 716 14.1 4.106 6.921 4.106 6.921 density_rs2pw 128 9.7 0.005 0.005 6.251 6.446 calculate_dm_sparse 128 9.5 0.001 0.001 6.270 6.371 cp_fm_cholesky_decompose 22 10.9 6.080 6.131 6.080 6.131 fft3d_ps 1291 13.7 2.939 3.003 5.957 6.013 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.555 5.623 grid_collocate_task_list 128 9.7 5.292 5.546 5.292 5.546 mp_alltoall_d11v 2415 14.1 4.969 5.343 4.969 5.343 qs_energies_init_hamiltonians 11 5.9 0.012 0.029 4.548 4.549 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.456 4.549 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.389 4.462 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.257 4.312 potential_pw2rs 128 12.3 0.022 0.022 3.982 3.991 dbcsr_special_finalize 12535 15.6 0.065 0.068 3.536 3.697 dbcsr_merge_single_wm 12535 16.6 0.323 0.339 3.447 3.606 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=177.543000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1055.181818, yerr=16.737645 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/19/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 5.820059E+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 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.408831E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108580288 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 759871. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4001 58216. MP_Allreduce 11080 1167. MP_Sync 86 MP_Alltoall 1700 18828153. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.033 163.314 163.315 qs_mol_dyn_low 1 2.0 0.003 0.004 162.901 162.918 qs_forces 11 3.9 0.004 0.006 162.796 162.798 qs_energies 11 4.9 0.002 0.002 155.379 155.388 scf_env_do_scf 11 5.9 0.001 0.001 138.297 138.312 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 107.509 107.510 velocity_verlet 10 3.0 0.002 0.002 105.808 105.813 dbcsr_multiply_generic 2485 12.5 0.182 0.189 71.928 72.375 qs_scf_new_mos 116 7.6 0.001 0.001 71.488 71.634 qs_scf_loop_do_ot 116 8.6 0.001 0.001 71.488 71.634 ot_scf_mini 116 9.6 0.003 0.003 67.114 67.257 multiply_cannon 2485 13.5 0.554 0.578 53.196 56.415 multiply_cannon_loop 2485 14.5 0.803 0.838 50.217 51.239 ot_mini 116 10.6 0.001 0.001 37.435 37.585 mp_waitall_1 124680 16.7 25.093 32.113 25.093 32.113 init_scf_loop 11 6.9 0.000 0.000 30.627 30.628 rebuild_ks_matrix 127 8.3 0.001 0.001 28.912 29.058 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 28.911 29.057 prepare_preconditioner 11 7.9 0.000 0.000 26.776 26.791 make_preconditioner 11 8.9 0.000 0.000 26.776 26.791 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.340 26.478 make_full_inverse_cholesky 11 9.9 0.022 0.028 24.878 25.142 multiply_cannon_multrec 9940 15.5 10.245 14.498 17.809 22.520 multiply_cannon_metrocomm3 9940 15.5 0.024 0.026 12.965 20.764 qs_ot_get_derivative 116 11.6 0.002 0.002 20.378 20.523 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.145 17.332 apply_single 127 13.6 0.001 0.001 17.144 17.332 qs_ot_get_p 127 10.4 0.002 0.002 16.987 17.121 ot_diis_step 116 11.6 0.019 0.020 16.983 16.984 make_m2s 4970 13.5 0.063 0.067 14.853 15.777 make_images 4970 14.5 2.139 2.561 14.549 15.476 cp_fm_cholesky_invert 11 10.9 14.726 14.737 14.726 14.737 sum_up_and_integrate 127 10.3 0.002 0.002 13.710 13.756 integrate_v_rspace 127 11.3 0.004 0.004 13.650 13.698 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.387 13.419 calculate_rho_elec 127 8.7 0.255 0.266 13.386 13.419 qs_ot_p2m_diag 82 11.4 0.490 0.496 13.249 13.267 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.097 12.098 multiply_cannon_sync_h2d 9940 15.5 10.590 11.195 10.590 11.195 init_scf_run 11 5.9 0.000 0.001 10.289 10.289 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.288 10.289 make_images_data 4970 15.5 0.055 0.066 8.609 9.910 hybrid_alltoall_any 5155 16.4 0.837 3.743 8.389 9.754 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.217 9.229 cp_fm_diag_elpa_base 82 14.4 8.983 9.062 9.213 9.225 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.966 8.076 grid_integrate_task_list 127 12.3 7.734 8.070 7.734 8.070 fft_wrap_pw1pw2 1281 11.7 0.017 0.019 8.028 8.052 dbcsr_mm_accdrv_process 20590 16.1 3.186 4.456 7.225 7.889 multiply_cannon_metrocomm1 9940 15.5 0.029 0.030 4.252 7.837 wfi_extrapolate 11 7.9 0.001 0.001 7.240 7.240 fft_wrap_pw1pw2_140 519 12.2 0.432 0.436 7.077 7.096 cp_fm_cholesky_decompose 22 10.9 6.514 6.543 6.514 6.543 density_rs2pw 127 9.7 0.005 0.005 6.126 6.485 calculate_dm_sparse 127 9.5 0.001 0.001 6.066 6.173 fft3d_ps 1281 13.7 3.097 3.208 6.006 6.044 multiply_cannon_metrocomm4 7455 15.5 0.025 0.028 1.855 5.891 grid_collocate_task_list 127 9.7 5.539 5.797 5.539 5.797 mp_irecv_dv 28618 15.9 1.815 5.796 1.815 5.796 dbcsr_complete_redistribute 393 12.7 1.998 2.036 5.106 5.457 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.287 5.335 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.291 5.293 mp_alltoall_d11v 2401 14.1 4.604 5.098 4.604 5.098 mp_allgather_i34 2485 14.5 1.203 4.728 1.203 4.728 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.142 4.203 potential_pw2rs 127 12.3 0.026 0.027 3.937 3.942 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.583 3.887 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.604 3.701 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.605 3.652 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.480 3.577 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.243 3.567 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.469 3.486 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=163.315000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1327.090909, yerr=18.510216 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243758E+12 0.0% 0.0% 100.0% flops max/rank 11.787404E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547744 0.0% 0.0% 100.0% number of processed stacks 1982128 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.0 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 2.915389E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101360 MPI messages size (bytes): total size 1.145137E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.297725E+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 45848 35601252352 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58558. MP_Allreduce 11184 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.055 247.115 247.115 qs_mol_dyn_low 1 2.0 0.003 0.004 246.522 246.541 qs_forces 11 3.9 0.004 0.005 246.425 246.428 qs_energies 11 4.9 0.002 0.002 237.618 237.622 scf_env_do_scf 11 5.9 0.001 0.001 215.153 215.167 velocity_verlet 10 3.0 0.002 0.002 171.791 171.799 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 132.634 132.637 qs_scf_new_mos 118 7.6 0.001 0.001 89.483 89.614 qs_scf_loop_do_ot 118 8.6 0.001 0.001 89.482 89.613 dbcsr_multiply_generic 2534 12.6 0.217 0.227 83.918 84.693 ot_scf_mini 118 9.6 0.003 0.004 84.438 84.535 init_scf_loop 11 6.9 0.000 0.000 82.210 82.216 prepare_preconditioner 11 7.9 0.000 0.000 77.471 77.497 make_preconditioner 11 8.9 0.000 0.000 77.471 77.497 make_full_inverse_cholesky 11 9.9 0.038 0.039 61.414 74.663 multiply_cannon 2534 13.6 0.675 0.703 59.894 60.953 multiply_cannon_loop 2534 14.6 1.054 1.070 55.624 57.680 cp_fm_upper_to_full 106 14.8 33.968 48.416 33.968 48.416 ot_mini 118 10.6 0.001 0.001 45.092 45.213 rebuild_ks_matrix 129 8.3 0.001 0.001 34.095 34.250 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 34.094 34.250 mp_waitall_1 104780 16.8 28.831 34.215 28.831 34.215 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.735 31.877 dbcsr_complete_redistribute 397 12.7 3.815 3.872 22.089 30.595 copy_fm_to_dbcsr 210 11.7 0.001 0.001 18.509 26.968 qs_ot_get_derivative 118 11.6 0.002 0.002 24.709 24.806 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 16.015 24.350 qs_ot_get_p 129 10.4 0.002 0.002 24.009 24.119 mp_alltoall_i22 720 14.1 13.804 22.407 13.804 22.407 multiply_cannon_metrocomm3 10136 15.6 0.024 0.026 19.310 21.290 make_m2s 5068 13.6 0.075 0.077 19.109 20.811 ot_diis_step 118 11.6 0.022 0.023 20.333 20.333 make_images 5068 14.6 3.098 3.291 18.626 20.326 qs_ot_p2m_diag 84 11.4 0.890 0.896 19.838 19.867 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.463 19.816 apply_single 129 13.6 0.001 0.001 19.463 19.815 cp_fm_cholesky_invert 11 10.9 18.374 18.387 18.374 18.387 multiply_cannon_multrec 10136 15.6 10.406 12.239 18.243 18.338 cp_dbcsr_syevd 84 12.4 0.006 0.006 18.020 18.023 sum_up_and_integrate 129 10.3 0.002 0.002 15.839 15.930 integrate_v_rspace 129 11.3 0.004 0.005 15.776 15.867 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.855 15.866 calculate_rho_elec 129 8.7 0.486 0.486 15.854 15.865 cp_fm_diag_elpa 84 13.4 0.000 0.000 14.793 14.793 cp_fm_diag_elpa_base 84 14.4 11.941 13.004 14.788 14.788 multiply_cannon_sync_h2d 10136 15.6 14.421 14.463 14.421 14.463 hybrid_alltoall_any 5255 16.5 1.318 3.068 11.052 13.513 make_images_data 5068 15.6 0.063 0.068 10.805 13.201 init_scf_run 11 5.9 0.000 0.001 12.159 12.160 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.159 12.160 fft_wrap_pw1pw2 1301 11.7 0.021 0.021 10.013 10.027 qs_ot_get_derivative_diag 78 12.4 0.003 0.003 9.944 10.023 dbcsr_mm_accdrv_process 20954 16.1 4.287 6.112 7.586 9.443 fft_wrap_pw1pw2_140 527 12.2 0.830 0.832 8.862 8.877 wfi_extrapolate 11 7.9 0.001 0.001 8.871 8.871 grid_integrate_task_list 129 12.3 8.619 8.810 8.619 8.810 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.155 8.155 cp_fm_cholesky_decompose 22 10.9 7.344 7.401 7.344 7.401 fft3d_ps 1301 13.7 4.016 4.029 7.304 7.318 mp_alltoall_d11v 2429 14.1 7.167 7.296 7.167 7.296 calculate_dm_sparse 129 9.5 0.001 0.001 6.920 6.974 density_rs2pw 129 9.7 0.005 0.005 6.889 6.956 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.638 6.708 grid_collocate_task_list 129 9.7 6.459 6.533 6.459 6.533 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.254 6.371 qs_env_update_s_mstruct 11 6.9 0.000 0.000 5.011 5.208 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.601 5.013 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=247.115000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2606.090909, yerr=151.386594 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 1.270301E+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 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 0.037 0.095 85.034 85.036 qs_energies 1 2.0 0.000 0.001 84.329 84.342 ls_scf 1 3.0 0.000 0.002 83.399 83.412 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.457 72.644 multiply_cannon 111 7.7 0.017 0.020 55.863 56.963 multiply_cannon_loop 111 8.7 0.228 0.247 52.506 53.801 ls_scf_main 1 4.0 0.000 0.002 52.285 52.286 density_matrix_trs4 2 5.0 0.002 0.003 46.746 46.818 ls_scf_init_scf 1 4.0 0.000 0.001 28.068 28.069 ls_scf_init_matrix_S 1 5.0 0.000 0.000 26.902 26.976 mp_waitall_1 11031 10.9 22.431 25.868 22.431 25.868 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 24.834 24.851 multiply_cannon_multrec 2664 9.7 8.183 8.990 15.562 17.412 multiply_cannon_sync_h2d 2664 9.7 13.569 15.876 13.569 15.876 make_m2s 222 7.7 0.009 0.011 13.064 13.455 make_images 222 8.7 0.099 0.109 13.042 13.434 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.741 12.779 make_images_data 222 9.7 0.004 0.005 7.623 8.107 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.413 8.030 dbcsr_mm_accdrv_process 4760 10.4 0.595 0.722 6.997 8.006 hybrid_alltoall_any 227 10.6 0.216 1.834 6.585 7.625 dbcsr_mm_accdrv_process_sort 4760 11.4 6.199 7.102 6.199 7.102 calculate_norms 4752 9.8 5.500 6.163 5.500 6.163 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.032 5.191 mp_sum_l 887 5.1 2.991 4.683 2.991 4.683 make_images_sizes 222 9.7 0.000 0.000 0.690 3.686 mp_alltoall_i44 222 10.7 0.690 3.685 0.690 3.685 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.261 3.499 arnoldi_extremal 4 6.8 0.000 0.000 3.171 3.201 arnoldi_normal_ev 4 7.8 0.001 0.002 3.171 3.201 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.049 3.195 mp_irecv_dv 6231 10.9 2.032 3.176 2.032 3.176 build_subspace 16 8.4 0.009 0.012 3.082 3.083 ls_scf_post 1 4.0 0.000 0.002 3.046 3.059 ls_scf_store_result 1 5.0 0.000 0.000 2.849 2.888 dbcsr_special_finalize 555 9.7 0.005 0.006 2.393 2.778 dbcsr_merge_single_wm 555 10.7 0.455 0.576 2.384 2.769 make_images_pack 222 9.7 2.212 2.634 2.213 2.636 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.316 2.550 dbcsr_sort_data 658 11.4 2.188 2.507 2.188 2.507 dbcsr_matrix_vector_mult_local 304 10.0 2.071 2.456 2.074 2.457 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.295 2.391 buffer_matrices_ensure_size 222 8.7 1.742 2.101 1.742 2.101 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.815 1.816 rebuild_ks_matrix 3 7.3 0.001 0.008 1.803 1.805 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.008 1.802 1.804 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.036000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1137.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 2.168582E+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 0.027 0.040 90.424 90.425 qs_energies 1 2.0 0.000 0.000 89.976 89.980 ls_scf 1 3.0 0.000 0.000 88.650 88.654 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.829 75.135 multiply_cannon 111 7.7 0.028 0.042 52.991 57.907 ls_scf_main 1 4.0 0.000 0.000 55.072 55.079 multiply_cannon_loop 111 8.7 0.135 0.147 50.175 53.873 density_matrix_trs4 2 5.0 0.002 0.003 49.389 49.599 mp_waitall_1 9105 10.9 21.213 30.604 21.213 30.604 ls_scf_init_scf 1 4.0 0.000 0.001 29.857 29.859 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.666 28.770 multiply_cannon_multrec 1332 9.7 13.292 17.268 22.564 27.643 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.300 26.310 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.895 21.143 make_m2s 222 7.7 0.006 0.007 15.238 15.772 make_images 222 8.7 1.367 1.686 15.208 15.741 dbcsr_mm_accdrv_process 4041 10.4 0.351 0.545 8.871 10.394 dbcsr_mm_accdrv_process_sort 4041 11.4 8.384 9.849 8.384 9.849 make_images_data 222 9.7 0.004 0.004 8.833 9.763 hybrid_alltoall_any 227 10.6 0.540 2.548 8.195 9.525 mp_sum_l 887 5.1 5.118 8.679 5.118 8.679 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.233 7.862 mp_irecv_dv 3311 11.0 3.212 7.811 3.212 7.811 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.901 6.767 calculate_norms 2376 9.8 6.050 6.682 6.050 6.682 multiply_cannon_sync_h2d 1332 9.7 4.815 6.172 4.815 6.172 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.942 5.139 arnoldi_extremal 4 6.8 0.000 0.000 4.672 4.689 arnoldi_normal_ev 4 7.8 0.001 0.004 4.672 4.689 build_subspace 16 8.4 0.014 0.021 4.425 4.428 ls_scf_post 1 4.0 0.000 0.000 3.721 3.725 ls_scf_store_result 1 5.0 0.000 0.000 3.440 3.545 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.173 3.404 dbcsr_matrix_vector_mult_local 304 10.0 2.775 3.255 2.778 3.257 mp_allgather_i34 111 8.7 0.812 3.089 0.812 3.089 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.596 2.696 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.128 2.587 dbcsr_data_new 4174 10.1 2.112 2.405 2.112 2.405 make_images_pack 222 9.7 1.821 2.120 1.824 2.122 dbcsr_sort_data 436 11.2 1.867 2.092 1.867 2.092 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.882 1.884 rebuild_ks_matrix 3 7.3 0.000 0.000 1.869 1.871 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.869 1.871 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.425000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1830.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 2.890871E+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 0.025 0.045 92.721 92.722 qs_energies 1 2.0 0.000 0.000 92.164 92.167 ls_scf 1 3.0 0.000 0.000 90.752 90.755 dbcsr_multiply_generic 111 6.7 0.016 0.016 75.419 75.775 ls_scf_main 1 4.0 0.000 0.000 56.845 56.850 multiply_cannon 111 7.7 0.031 0.044 52.145 56.473 multiply_cannon_loop 111 8.7 0.116 0.129 49.370 53.780 density_matrix_trs4 2 5.0 0.002 0.003 50.896 51.129 mp_waitall_1 7281 11.0 23.725 34.144 23.725 34.144 ls_scf_init_scf 1 4.0 0.000 0.001 30.250 30.252 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.066 29.135 matrix_sqrt_Newton_Schulz 2 6.5 0.004 0.010 26.739 26.750 multiply_cannon_multrec 888 9.7 12.660 15.003 21.246 24.324 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.262 23.734 make_m2s 222 7.7 0.006 0.008 16.447 17.183 make_images 222 8.7 1.581 1.858 16.409 17.142 make_images_data 222 9.7 0.004 0.004 9.601 10.664 hybrid_alltoall_any 227 10.6 0.639 2.942 9.125 10.336 dbcsr_mm_accdrv_process 3754 10.4 0.326 0.498 8.091 9.354 dbcsr_mm_accdrv_process_sort 3754 11.4 7.653 8.856 7.653 8.856 mp_sum_l 887 5.1 5.117 8.624 5.117 8.624 multiply_cannon_sync_h2d 888 9.7 6.079 7.502 6.079 7.502 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.469 7.198 mp_irecv_dv 2335 11.1 2.453 7.156 2.453 7.156 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.505 7.083 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.952 6.834 arnoldi_extremal 4 6.8 0.000 0.000 5.166 5.183 arnoldi_normal_ev 4 7.8 0.001 0.004 5.166 5.183 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.778 5.013 build_subspace 16 8.4 0.014 0.020 4.852 4.857 calculate_norms 1584 9.8 4.347 4.756 4.347 4.756 mp_allgather_i34 111 8.7 0.894 3.833 0.894 3.833 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.476 3.822 ls_scf_post 1 4.0 0.000 0.000 3.656 3.662 dbcsr_matrix_vector_mult_local 304 10.0 3.050 3.648 3.052 3.650 ls_scf_store_result 1 5.0 0.000 0.000 3.400 3.483 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.834 2.930 dbcsr_data_new 4116 9.9 2.116 2.457 2.116 2.457 dbcsr_sort_data 325 11.1 1.888 2.123 1.888 2.123 make_images_sizes 222 9.7 0.000 0.000 1.003 2.109 mp_alltoall_i44 222 10.7 1.003 2.109 1.003 2.109 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.913 1.915 rebuild_ks_matrix 3 7.3 0.000 0.000 1.895 1.897 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.895 1.897 make_images_pack 222 9.7 1.626 1.877 1.629 1.880 dbcsr_finalize 304 7.8 0.026 0.032 1.609 1.858 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.722000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2304.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 3.386728E+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 0.036 0.048 99.476 99.477 qs_energies 1 2.0 0.000 0.001 98.848 98.854 ls_scf 1 3.0 0.000 0.001 97.148 97.153 dbcsr_multiply_generic 111 6.7 0.017 0.017 80.860 81.115 multiply_cannon 111 7.7 0.043 0.081 54.317 58.982 ls_scf_main 1 4.0 0.000 0.001 58.592 58.593 density_matrix_trs4 2 5.0 0.002 0.004 52.443 52.576 multiply_cannon_loop 111 8.7 0.153 0.166 49.222 51.754 ls_scf_init_scf 1 4.0 0.000 0.002 35.307 35.309 ls_scf_init_matrix_S 1 5.0 0.000 0.003 34.087 34.158 mp_waitall_1 6369 11.0 22.773 31.830 22.773 31.830 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 31.601 31.613 multiply_cannon_multrec 1332 9.7 14.228 16.883 22.128 25.126 make_m2s 222 7.7 0.007 0.008 20.967 22.349 make_images 222 8.7 3.140 3.616 20.917 22.301 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.328 18.417 make_images_data 222 9.7 0.004 0.004 11.680 13.414 hybrid_alltoall_any 227 10.6 0.796 3.780 11.067 12.992 dbcsr_mm_accdrv_process 3641 10.4 0.306 0.486 7.542 9.085 dbcsr_mm_accdrv_process_sort 3641 11.4 7.099 8.585 7.099 8.585 mp_sum_l 887 5.1 3.992 7.627 3.992 7.627 multiply_cannon_sync_h2d 1332 9.7 5.514 6.113 5.514 6.113 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.098 6.065 mp_irecv_dv 3229 10.9 2.073 5.989 2.073 5.989 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.026 5.986 arnoldi_extremal 4 6.8 0.000 0.000 5.272 5.285 arnoldi_normal_ev 4 7.8 0.001 0.005 5.272 5.285 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.591 5.124 build_subspace 16 8.4 0.014 0.021 4.946 4.953 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.596 4.794 mp_allgather_i34 111 8.7 2.213 4.642 2.213 4.642 calculate_norms 2376 9.8 4.168 4.505 4.168 4.505 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.633 3.947 dbcsr_matrix_vector_mult_local 304 10.0 3.240 3.735 3.242 3.737 compute_matrix_preconditioner 1 6.0 0.001 0.001 3.490 3.496 dbcsr_sort_data 658 11.4 3.090 3.345 3.090 3.345 ls_scf_post 1 4.0 0.000 0.003 3.250 3.255 dbcsr_special_finalize 555 9.7 0.006 0.007 2.840 3.153 dbcsr_merge_single_wm 555 10.7 0.538 0.655 2.831 3.144 ls_scf_store_result 1 5.0 0.000 0.000 2.973 3.039 acc_transpose_blocks 1332 9.7 0.008 0.008 2.551 3.037 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.993 3.028 acc_transpose_blocks_kernels 1332 10.7 0.018 0.019 2.439 2.924 jit_kernel_transpose 1 13.0 2.421 2.906 2.421 2.906 dbcsr_data_release 10477 10.7 1.578 2.407 1.578 2.407 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=99.477000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2806.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 4.843602E+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 0.038 0.056 91.778 91.779 qs_energies 1 2.0 0.000 0.000 90.965 90.969 ls_scf 1 3.0 0.000 0.000 89.034 89.039 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.556 70.774 ls_scf_main 1 4.0 0.000 0.000 56.381 56.382 multiply_cannon 111 7.7 0.073 0.149 52.531 55.482 multiply_cannon_loop 111 8.7 0.088 0.092 49.996 51.085 density_matrix_trs4 2 5.0 0.002 0.003 49.372 49.446 ls_scf_init_scf 1 4.0 0.000 0.001 29.290 29.295 mp_waitall_1 5436 11.0 24.227 28.649 24.227 28.649 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.053 28.099 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.967 25.975 multiply_cannon_multrec 444 9.7 13.818 16.390 20.877 21.977 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.287 15.488 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.614 14.621 make_m2s 222 7.7 0.005 0.005 13.388 14.311 make_images 222 8.7 2.038 2.476 13.321 14.241 make_images_data 222 9.7 0.003 0.004 8.143 9.582 hybrid_alltoall_any 227 10.6 0.800 3.807 8.092 9.401 multiply_cannon_sync_h2d 444 9.7 6.761 8.067 6.761 8.067 dbcsr_mm_accdrv_process 3003 10.4 0.354 0.404 6.756 7.940 dbcsr_mm_accdrv_process_sort 3003 11.4 6.391 7.536 6.391 7.536 arnoldi_extremal 4 6.8 0.000 0.000 5.880 5.890 arnoldi_normal_ev 4 7.8 0.001 0.004 5.880 5.890 build_subspace 16 8.4 0.015 0.020 5.488 5.496 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.444 4.663 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.233 4.411 dbcsr_matrix_vector_mult_local 304 10.0 3.738 4.211 3.740 4.213 mp_sum_l 887 5.1 2.641 4.072 2.641 4.072 calculate_norms 792 9.8 3.625 3.711 3.625 3.711 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.497 3.679 mp_irecv_dv 1241 11.2 1.483 3.642 1.483 3.642 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.516 3.597 mp_allgather_i34 111 8.7 1.120 3.495 1.120 3.495 ls_scf_post 1 4.0 0.000 0.000 3.363 3.368 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.878 3.342 make_images_sizes 222 9.7 0.000 0.000 0.861 3.280 mp_alltoall_i44 222 10.7 0.861 3.280 0.861 3.280 ls_scf_store_result 1 5.0 0.000 0.000 3.151 3.190 dbcsr_finalize 304 7.8 0.062 0.076 2.199 2.262 dbcsr_data_new 4608 9.7 1.794 2.241 1.794 2.241 dbcsr_merge_all 275 8.9 0.477 0.523 2.055 2.101 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.035 2.036 rebuild_ks_matrix 3 7.3 0.000 0.000 2.001 2.002 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.004 2.001 2.002 qs_energies_init_hamiltonians 1 3.0 0.001 0.004 1.916 1.916 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=91.779000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3797.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_performance_tests/26/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 30.358840E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 339931 0.0% 0.0% 100.0% average stack size 0.0 0.0 28341.7 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 9.060356E+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 0.082 0.135 105.049 105.049 qs_energies 1 2.0 0.000 0.000 103.677 103.683 ls_scf 1 3.0 0.000 0.000 100.714 100.720 dbcsr_multiply_generic 111 6.7 0.024 0.028 74.644 74.786 ls_scf_main 1 4.0 0.000 0.000 63.481 63.481 density_matrix_trs4 2 5.0 0.002 0.003 54.711 54.776 multiply_cannon 111 7.7 0.153 0.188 48.206 50.060 multiply_cannon_loop 111 8.7 0.097 0.098 45.253 46.085 ls_scf_init_scf 1 4.0 0.000 0.001 33.442 33.443 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.048 32.070 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.341 29.349 mp_waitall_1 4527 11.1 21.720 25.565 21.720 25.565 make_m2s 222 7.7 0.005 0.005 22.673 23.742 make_images 222 8.7 3.575 3.864 22.566 23.633 multiply_cannon_multrec 444 9.7 17.832 18.498 22.471 23.118 hybrid_alltoall_any 227 10.6 1.657 3.613 12.853 15.834 make_images_data 222 9.7 0.003 0.004 13.074 15.190 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.053 10.721 multiply_cannon_sync_h2d 444 9.7 8.769 8.866 8.769 8.866 arnoldi_extremal 4 6.8 0.000 0.000 7.471 7.481 arnoldi_normal_ev 4 7.8 0.002 0.008 7.471 7.481 build_subspace 16 8.4 0.026 0.037 6.927 6.943 dbcsr_matrix_vector_mult 304 9.0 0.017 0.035 5.553 5.698 dbcsr_matrix_vector_mult_local 304 10.0 5.116 5.424 5.119 5.427 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.850 5.101 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.905 4.992 dbcsr_mm_accdrv_process 1814 10.4 0.268 0.364 4.452 4.585 dbcsr_mm_accdrv_process_sort 1814 11.4 4.113 4.245 4.113 4.245 ls_scf_post 1 4.0 0.000 0.000 3.791 3.798 make_images_sizes 222 9.7 0.000 0.000 1.466 3.579 mp_alltoall_i44 222 10.7 1.466 3.579 1.466 3.579 ls_scf_store_result 1 5.0 0.000 0.000 3.515 3.555 mp_allgather_i34 111 8.7 1.076 3.436 1.076 3.436 calculate_norms 792 9.8 3.254 3.327 3.254 3.327 dbcsr_finalize 304 7.8 0.082 0.089 3.078 3.163 dbcsr_merge_all 275 8.9 0.881 0.915 2.865 2.943 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.932 2.932 dbcsr_complete_redistribute 5 7.6 1.423 1.466 2.735 2.843 matrix_ls_to_qs 2 6.0 0.000 0.000 2.394 2.509 dbcsr_sort_data 325 11.1 2.443 2.496 2.443 2.496 dbcsr_new_transposed 4 7.5 0.248 0.268 2.308 2.315 dbcsr_data_new 6591 9.6 1.822 2.294 1.822 2.294 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.277 2.278 dbcsr_frobenius_norm 74 6.6 2.056 2.131 2.211 2.249 rebuild_ks_matrix 3 7.3 0.000 0.000 2.210 2.212 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.003 2.210 2.212 dbcsr_add_d 103 6.2 0.000 0.000 2.123 2.205 dbcsr_add_anytype 103 7.2 0.859 0.891 2.123 2.205 dbcsr_data_release 12724 10.6 1.977 2.178 1.977 2.178 mp_sum_l 887 5.1 1.373 2.123 1.373 2.123 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.049000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7211.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5f6c7e94cf466f5c887cb267571013f7bc5f1114_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 586.809344E+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 4215596. 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 0.033 0.115 225.700 225.702 qs_mol_dyn_low 1 2.0 0.005 0.014 224.601 224.624 qs_forces 5 3.8 0.004 0.011 224.523 224.526 qs_energies 5 4.8 0.003 0.033 221.607 221.624 scf_env_do_scf 5 5.8 0.001 0.005 208.852 208.852 scf_env_do_scf_inner_loop 105 6.6 0.003 0.008 184.289 184.290 qs_scf_new_mos 105 7.6 0.000 0.001 144.659 144.825 qs_scf_loop_do_ot 105 8.6 0.001 0.001 144.659 144.824 ot_scf_mini 105 9.6 0.003 0.004 134.644 134.838 dbcsr_multiply_generic 1445 12.2 0.129 0.137 133.345 133.882 multiply_cannon 1445 13.2 0.275 0.285 114.006 115.751 multiply_cannon_loop 1445 14.2 2.860 2.989 112.330 113.643 velocity_verlet 4 3.0 0.001 0.005 105.087 105.087 ot_mini 105 10.6 0.001 0.001 60.493 60.601 multiply_cannon_multrec 69360 15.2 29.653 34.582 39.579 44.564 qs_ot_get_p 112 10.4 0.001 0.001 43.906 44.188 mp_waitall_1 488190 16.1 35.199 42.195 35.199 42.195 qs_ot_get_derivative 55 11.6 0.001 0.001 38.713 38.851 multiply_cannon_sync_h2d 69360 15.2 29.159 33.138 29.159 33.138 multiply_cannon_metrocomm3 69360 15.2 0.203 0.212 26.041 33.078 qs_ot_p2m_diag 40 11.0 0.020 0.031 32.637 32.717 cp_dbcsr_syevd 40 12.0 0.002 0.002 29.369 29.370 rebuild_ks_matrix 110 8.4 0.000 0.000 28.505 28.660 qs_ks_build_kohn_sham_matrix 110 9.4 0.011 0.017 28.504 28.659 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.181 26.329 init_scf_loop 7 6.6 0.000 0.005 24.525 24.526 cp_fm_syevd 40 13.0 0.000 0.000 24.150 24.316 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.002 23.273 apply_single 62 13.6 0.000 0.000 23.002 23.273 ot_new_cg_direction 55 11.6 0.001 0.001 21.048 21.048 prepare_preconditioner 7 7.6 0.000 0.000 19.775 19.803 make_preconditioner 7 8.6 0.000 0.003 19.775 19.803 cp_fm_redistribute_end 40 14.0 9.757 19.444 9.772 19.450 cp_fm_syevd_base 40 14.0 9.665 19.351 9.665 19.351 qs_rho_update_rho_low 110 7.6 0.000 0.001 16.755 17.207 calculate_rho_elec 110 8.6 0.028 0.031 16.754 17.207 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.047 15.247 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.936 14.022 mp_sum_l 4764 12.2 12.416 13.292 12.416 13.292 make_full_inverse_cholesky 7 9.6 0.000 0.000 12.597 12.658 calculate_dm_sparse 110 9.5 0.000 0.001 11.572 11.721 density_rs2pw 110 9.6 0.004 0.005 10.862 11.215 fft_wrap_pw1pw2 1425 12.5 0.019 0.023 10.637 10.809 dbcsr_mm_accdrv_process 154766 15.8 6.319 6.505 9.793 10.589 init_scf_run 5 5.8 0.000 0.001 10.499 10.500 scf_env_initial_rho_setup 5 6.8 0.002 0.003 10.499 10.500 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.334 10.389 qs_vxc_create 110 10.4 0.002 0.004 9.874 9.891 check_diag 80 13.5 8.558 8.872 9.381 9.548 fft_wrap_pw1pw2_240 915 14.0 0.196 0.214 9.285 9.442 cp_fm_cholesky_invert 7 10.6 8.449 8.467 8.449 8.467 fft3d_pb 915 15.0 2.245 2.424 8.257 8.412 sum_up_and_integrate 60 10.3 0.001 0.004 8.395 8.403 integrate_v_rspace 60 11.3 0.002 0.007 8.377 8.386 transfer_rs2pw 445 10.6 0.008 0.009 7.812 8.148 acc_transpose_blocks 69360 15.2 0.343 0.361 6.929 7.356 xc_rho_set_and_dset_create 110 12.4 0.073 0.078 6.952 7.200 make_m2s 2890 13.2 0.079 0.088 6.472 7.057 make_images 2890 14.2 0.239 0.259 6.363 6.949 make_full_single_inverse 7 9.6 0.001 0.002 6.902 6.930 multiply_cannon_metrocomm1 69360 15.2 0.093 0.099 4.690 6.843 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 6.746 6.766 xc_vxc_pw_create 60 11.3 0.036 0.037 6.594 6.612 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.340 6.397 calculate_first_density_matrix 1 7.0 0.000 0.004 6.218 6.236 xc_pw_derive 510 13.4 0.005 0.006 5.736 5.775 mp_alltoall_z22v 2340 16.7 5.436 5.627 5.436 5.627 mp_waitany 7680 13.5 4.486 5.058 4.486 5.058 potential_pw2rs 60 12.3 0.002 0.003 4.613 4.657 acc_transpose_blocks_kernels 69360 16.2 0.856 0.893 4.193 4.542 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=225.702000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=557.600000, yerr=2.332381 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 5f6c7e94cf466f5c887cb267571013f7bc5f1114 Summary: empty Status: OK