=== 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: 9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0 ################# 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, SIRIUS 7.4.3, SPGLIB 1.16.2, # LIBGRPP 20231215 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 15.12.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/01 job id: 50548028 --- 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/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/02 job id: 50548030 --- Point --- name: 20 plot: h2o_32_ri_rpa_mp2 regex: Total MP2 Time= label: RI-MP2 (8n/6r/2t) --- Point --- name: 21 plot: h2o_32_ri_rpa_mp2_mem regex: Estimated peak process memory label: RI-MP2 (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/03 job id: 50548034 --- Point --- name: 100 plot: h2o_64_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 101 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/04 job id: 50548036 --- Point --- name: 102 plot: h2o_64_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 103 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/05 job id: 50548038 --- Point --- name: 104 plot: h2o_64_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 105 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/06 job id: 50548040 --- Point --- name: 106 plot: h2o_64_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 107 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/07 job id: 50548042 --- Point --- name: 108 plot: h2o_64_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 109 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/08 job id: 50548043 --- 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/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/09 job id: 50548044 --- 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/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/10 job id: 50548045 --- 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/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/11 job id: 50548046 --- 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/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/12 job id: 50548047 --- 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/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/13 job id: 50548049 --- 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/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/14 job id: 50548050 --- 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/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/15 job id: 50548051 --- 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/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/16 job id: 50548052 --- 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/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/17 job id: 50548053 --- 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/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/18 job id: 50548054 --- 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/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/19 job id: 50548055 --- 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/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/20 job id: 50548056 --- Point --- name: 410 plot: h2o_256_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 411 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/21 job id: 50548057 --- Point --- name: 500 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/12r/1t) --- Point --- name: 501 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/22 job id: 50548058 --- Point --- name: 502 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/6r/2t) --- Point --- name: 503 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/23 job id: 50548059 --- Point --- name: 504 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/4r/3t) --- Point --- name: 505 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/24 job id: 50548060 --- Point --- name: 506 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/3r/4t) --- Point --- name: 507 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/25 job id: 50548061 --- Point --- name: 508 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/2r/6t) --- Point --- name: 509 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/26 job id: 50548062 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/27 job id: 50548063 --- 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/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_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 0.022 0.033 134.705 134.706 farming_run 1 2.0 133.908 133.909 134.672 134.674 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.490067E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.027 115.108 115.108 qs_energies 1 2.0 0.000 0.000 114.896 114.897 mp2_main 1 3.0 0.000 0.000 112.838 112.839 mp2_gpw_main 1 4.0 0.032 0.038 111.846 111.847 mp2_ri_gpw_compute_in 1 5.0 0.173 0.183 92.876 93.192 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.725 55.044 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.163 41.134 46.177 get_2c_integrals 1 6.0 0.008 0.009 36.861 37.978 integrate_v_rspace 273 8.0 0.440 0.455 24.685 29.497 pw_transfer 6555 10.6 0.380 0.390 26.901 27.401 fft_wrap_pw1pw2 5465 11.4 0.046 0.047 25.555 26.103 grid_integrate_task_list 273 9.0 20.575 25.867 20.575 25.867 fft_wrap_pw1pw2_100 2178 12.4 0.030 0.034 23.097 23.632 rpa_ri_compute_en 1 5.0 0.019 0.020 18.860 19.093 compute_2c_integrals 1 7.0 0.002 0.002 18.966 18.967 cp_fm_cholesky_decompose 12 8.2 17.821 18.942 17.821 18.942 compute_2c_integrals_loop_lm 1 8.0 0.001 0.003 18.585 18.691 mp2_eri_2c_integrate_gpw 1 9.0 2.390 2.422 18.584 18.690 cholesky_decomp 1 7.0 0.000 0.000 16.737 17.861 fft3d_s 5443 13.4 16.141 16.508 16.162 16.528 ao_to_mo_and_store_B_mult_1 272 7.0 10.764 15.314 10.764 15.314 calculate_wavefunction 272 8.0 5.400 5.549 12.354 12.993 rpa_num_int 1 6.0 0.000 0.001 10.602 10.611 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.581 10.594 calc_mat_Q 8 8.0 0.000 0.000 9.449 9.541 contract_S_to_Q 8 9.0 0.000 0.000 8.873 8.961 calc_potential_gpw 544 9.5 0.004 0.005 8.237 8.633 parallel_gemm_fm 14 9.1 0.000 0.000 8.467 8.563 parallel_gemm_fm_cosma 14 10.1 8.467 8.563 8.467 8.563 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.062 8.279 potential_pw2rs 545 10.0 0.107 0.108 7.488 8.128 create_integ_mat 1 6.0 0.084 0.098 7.848 7.857 collocate_single_gaussian 272 10.0 0.039 0.042 7.326 7.601 array2fm 1 7.0 0.000 0.000 6.767 7.199 pw_scatter_s 2720 13.7 4.377 4.549 4.377 4.549 pw_gather_s 2722 13.2 3.489 3.830 3.489 3.830 array2fm_buffer_send 1 8.0 2.997 3.159 2.997 3.159 pw_poisson_solve 545 10.5 1.114 1.173 2.162 2.320 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.844619, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2807.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 200775. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.072 422.298 422.299 farming_run 1 2.0 421.161 421.189 422.211 422.213 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.245061E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 25151988. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 217.742 217.743 qs_energies 1 2.0 0.000 0.000 217.497 217.509 scf_env_do_scf 1 3.0 0.000 0.000 115.335 115.336 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.425 114.432 rebuild_ks_matrix 4 6.0 0.000 0.000 114.424 114.431 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 114.424 114.431 hfx_ks_matrix 4 8.0 0.001 0.001 114.076 114.080 integrate_four_center 4 9.0 0.152 0.466 114.075 114.080 integrate_four_center_main 4 10.0 0.119 0.453 101.739 105.366 integrate_four_center_bin 269 11.0 101.621 105.067 101.621 105.067 mp2_main 1 3.0 0.002 0.014 101.867 101.880 mp2_gpw_main 1 4.0 0.053 0.100 100.984 100.995 init_scf_loop 1 4.0 0.000 0.000 96.692 96.693 mp2_ri_gpw_compute_in 1 5.0 0.069 0.106 74.041 75.139 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.852 54.948 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.159 41.637 46.509 integrate_v_rspace 95 8.0 0.397 0.564 28.115 32.811 pw_transfer 2240 10.6 0.145 0.162 29.510 29.943 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.516 28.943 mp2_ri_gpw_compute_en 1 5.0 0.056 0.066 26.787 28.876 grid_integrate_task_list 95 9.0 23.529 28.428 23.529 28.428 ao_to_mo_and_store_B_mult_1 91 7.0 10.541 28.255 10.541 28.255 fft_wrap_pw1pw2_100 730 12.4 0.013 0.014 26.261 26.720 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.831 1.882 25.044 25.053 get_2c_integrals 1 6.0 0.000 0.000 20.098 20.126 compute_2c_integrals 1 7.0 0.002 0.003 19.077 19.089 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.641 18.959 mp2_eri_2c_integrate_gpw 1 9.0 1.734 1.869 18.640 18.958 fft3d_s 1823 13.4 18.482 18.881 18.496 18.894 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 18.641 18.641 calculate_wavefunction 91 8.0 2.013 2.050 9.617 9.830 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.583 8.789 9.387 potential_pw2rs 186 10.0 0.033 0.034 8.424 8.998 local_gemm 172 8.0 8.231 8.808 8.231 8.808 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.164 8.530 mp2_ri_gpw_compute_en_comm 22 7.0 0.497 0.516 8.032 8.466 calc_potential_gpw 182 9.5 0.002 0.002 7.914 8.238 collocate_single_gaussian 91 10.0 0.017 0.022 7.758 8.042 mp_sync 37 10.5 4.284 7.396 4.284 7.396 integrate_four_center_load 4 10.0 0.000 0.000 6.784 6.789 hfx_load_balance 1 11.0 0.000 0.000 6.784 6.789 mp_sendrecv_dm3 2068 8.0 6.074 6.508 6.074 6.508 mp2_ri_gpw_compute_en_ener 172 7.0 6.347 6.414 6.347 6.414 pw_gather_s 912 13.2 4.482 4.994 4.482 4.994 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.977861, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1504.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.497408E+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 0.034 0.054 55.414 55.416 qs_mol_dyn_low 1 2.0 0.003 0.004 55.034 55.063 qs_forces 11 3.9 0.002 0.002 54.970 54.971 qs_energies 11 4.9 0.026 0.027 53.431 53.442 scf_env_do_scf 11 5.9 0.002 0.017 46.479 46.480 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 44.461 44.461 dbcsr_multiply_generic 2286 12.5 0.094 0.098 34.438 34.888 qs_scf_new_mos 108 7.5 0.000 0.001 34.224 34.520 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.223 34.520 ot_scf_mini 108 9.5 0.002 0.002 32.519 32.701 multiply_cannon 2286 13.5 0.188 0.194 27.071 28.587 multiply_cannon_loop 2286 14.5 1.820 1.933 26.407 27.924 velocity_verlet 10 3.0 0.001 0.001 26.599 26.599 ot_mini 108 10.5 0.001 0.001 19.561 19.801 qs_ot_get_derivative 108 11.5 0.001 0.001 16.543 16.747 mp_waitall_1 245248 16.5 8.744 15.151 8.744 15.151 multiply_cannon_metrocomm3 54864 15.5 0.073 0.078 6.277 13.545 multiply_cannon_multrec 54864 15.5 3.640 5.731 7.962 11.527 qs_ot_get_p 119 10.4 0.001 0.001 8.160 8.418 rebuild_ks_matrix 119 8.3 0.000 0.000 7.906 8.041 qs_ks_build_kohn_sham_matrix 119 9.3 0.021 0.065 7.905 8.040 mp_sum_l 7287 12.8 5.414 7.150 5.414 7.150 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.919 7.044 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.932 6.394 multiply_cannon_sync_h2d 54864 15.5 5.152 6.342 5.152 6.342 dbcsr_mm_accdrv_process 76910 16.1 1.836 2.903 4.235 6.052 init_scf_run 11 5.9 0.000 0.001 5.583 5.583 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.583 5.583 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.278 5.315 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.004 5.097 sum_up_and_integrate 119 10.3 0.001 0.002 4.435 4.490 integrate_v_rspace 119 11.3 0.002 0.003 4.424 4.480 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.469 4.470 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.254 4.254 qs_rho_update_rho_low 119 7.7 0.000 0.000 4.087 4.246 calculate_rho_elec 119 8.7 0.011 0.016 4.087 4.246 cp_fm_redistribute_end 50 14.0 2.170 4.227 2.179 4.231 cp_fm_diag_elpa_base 50 14.0 2.045 4.112 2.049 4.121 calculate_first_density_matrix 1 7.0 0.000 0.001 3.403 3.406 calculate_dm_sparse 119 9.5 0.000 0.000 3.101 3.264 multiply_cannon_metrocomm1 54864 15.5 0.055 0.060 1.810 3.232 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.951 3.178 apply_single 119 13.6 0.000 0.000 2.951 3.178 jit_kernel_multiply 13 15.8 2.336 3.149 2.336 3.149 acc_transpose_blocks 54864 15.5 0.222 0.244 2.468 3.110 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.769 2.771 ot_diis_step 108 11.5 0.006 0.006 2.740 2.741 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.540 2.619 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.396 2.471 density_rs2pw 119 9.7 0.004 0.004 2.083 2.161 grid_integrate_task_list 119 12.3 2.011 2.115 2.011 2.115 wfi_extrapolate 11 7.9 0.001 0.001 2.111 2.112 init_scf_loop 11 6.9 0.000 0.000 1.999 2.001 mp_sum_d 4137 12.0 1.386 1.982 1.386 1.982 potential_pw2rs 119 12.3 0.004 0.004 1.814 1.871 pw_transfer 1439 11.6 0.051 0.056 1.621 1.732 make_m2s 4572 13.5 0.053 0.056 1.647 1.699 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.546 1.661 make_images 4572 14.5 0.133 0.140 1.565 1.617 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.559 1.592 mp_alltoall_d11v 2130 13.8 1.325 1.498 1.325 1.498 mp_waitany 12084 13.8 1.291 1.464 1.291 1.464 acc_transpose_blocks_sync 164592 16.5 1.198 1.444 1.198 1.444 transfer_rs2pw 487 10.6 0.005 0.006 1.325 1.417 grid_collocate_task_list 119 9.7 1.350 1.402 1.350 1.402 transfer_pw2rs 487 13.2 0.006 0.007 1.365 1.395 acc_transpose_blocks_kernels 54864 16.5 0.246 0.379 1.002 1.390 fft3d_ps 1201 14.6 0.365 0.474 1.256 1.366 fft_wrap_pw1pw2_140 487 13.2 0.110 0.116 1.202 1.319 parallel_gemm_fm 81 9.0 0.000 0.000 1.190 1.194 parallel_gemm_fm_cosma 81 10.0 1.190 1.194 1.190 1.194 dbcsr_dot_sd 1205 11.9 0.048 0.060 0.744 1.168 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.416000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.727273, yerr=0.616575 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_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 490.246144E+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 567685. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.026 42.450 42.451 qs_mol_dyn_low 1 2.0 0.003 0.004 42.149 42.258 qs_forces 11 3.9 0.002 0.003 42.026 42.026 qs_energies 11 4.9 0.001 0.002 40.354 40.359 scf_env_do_scf 11 5.9 0.001 0.001 35.076 35.077 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 32.506 32.506 dbcsr_multiply_generic 2286 12.5 0.102 0.106 22.993 23.417 qs_scf_new_mos 108 7.5 0.001 0.001 23.099 23.329 qs_scf_loop_do_ot 108 8.5 0.001 0.001 23.098 23.329 ot_scf_mini 108 9.5 0.002 0.003 22.176 22.344 velocity_verlet 10 3.0 0.001 0.002 19.238 19.240 multiply_cannon 2286 13.5 0.208 0.216 17.539 19.084 multiply_cannon_loop 2286 14.5 1.188 1.247 16.294 17.906 ot_mini 108 10.5 0.001 0.001 13.763 13.996 mp_waitall_1 200699 16.5 6.057 11.356 6.057 11.356 qs_ot_get_derivative 108 11.5 0.001 0.001 11.181 11.346 multiply_cannon_metrocomm3 27432 15.5 0.072 0.074 4.410 10.071 multiply_cannon_multrec 27432 15.5 1.827 4.233 6.190 9.054 rebuild_ks_matrix 119 8.3 0.000 0.000 6.769 6.908 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.018 6.769 6.908 dbcsr_mm_accdrv_process 47894 16.0 3.563 5.859 4.282 6.367 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.951 6.078 qs_ot_get_p 119 10.4 0.019 0.049 5.575 5.797 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.789 4.638 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.406 4.439 calculate_rho_elec 119 8.7 0.021 0.024 4.406 4.439 mp_sum_l 7287 12.8 2.234 4.395 2.234 4.395 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.155 4.249 apply_single 119 13.6 0.000 0.000 3.154 4.249 init_scf_run 11 5.9 0.000 0.001 4.080 4.080 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.079 4.080 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.825 3.843 sum_up_and_integrate 119 10.3 0.001 0.003 3.695 3.700 integrate_v_rspace 119 11.3 0.002 0.003 3.681 3.686 cp_dbcsr_syevd 50 12.0 0.003 0.003 3.424 3.424 make_m2s 4572 13.5 0.052 0.054 2.769 3.102 make_images 4572 14.5 0.205 0.243 2.680 3.016 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.868 2.879 cp_fm_redistribute_end 50 14.0 1.491 2.852 1.528 2.856 density_rs2pw 119 9.7 0.004 0.004 2.706 2.785 cp_fm_diag_elpa_base 50 14.0 1.284 2.600 1.312 2.640 calculate_first_density_matrix 1 7.0 0.000 0.001 2.548 2.550 init_scf_loop 11 6.9 0.000 0.000 2.548 2.548 ot_diis_step 108 11.5 0.010 0.011 2.509 2.511 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.339 2.426 calculate_dm_sparse 119 9.5 0.000 0.001 2.100 2.182 acc_transpose_blocks 27432 15.5 0.113 0.119 1.712 2.169 multiply_cannon_sync_h2d 27432 15.5 1.682 2.168 1.682 2.168 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.084 2.086 jit_kernel_multiply 10 16.4 0.661 1.944 0.661 1.944 grid_integrate_task_list 119 12.3 1.844 1.941 1.844 1.941 pw_transfer 1439 11.6 0.065 0.069 1.902 1.938 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.847 1.887 transfer_rs2pw 487 10.6 0.005 0.005 1.813 1.874 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.810 1.849 make_images_data 4572 15.5 0.048 0.054 1.294 1.660 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.215 1.591 mp_alltoall_d11v 2130 13.8 1.465 1.578 1.465 1.578 prepare_preconditioner 11 7.9 0.000 0.000 1.536 1.564 make_preconditioner 11 8.9 0.000 0.000 1.536 1.564 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.432 1.486 fft3d_ps 1201 14.6 0.522 0.577 1.440 1.474 wfi_extrapolate 11 7.9 0.001 0.001 1.469 1.469 potential_pw2rs 119 12.3 0.006 0.006 1.437 1.444 fft_wrap_pw1pw2_140 487 13.2 0.128 0.132 1.406 1.443 grid_collocate_task_list 119 9.7 1.288 1.361 1.288 1.361 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.285 1.296 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.215 1.260 mp_sum_d 4137 12.0 0.657 1.146 0.657 1.146 acc_transpose_blocks_kernels 27432 16.5 0.187 0.276 0.761 1.101 mp_allgather_i34 2286 14.5 0.669 1.097 0.669 1.097 mp_sendrecv_dv 16779 12.7 1.033 1.055 1.033 1.055 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.962 1.050 dbcsr_complete_redistribute 329 12.2 0.095 0.133 0.956 1.022 acc_transpose_blocks_sync 82296 16.5 0.813 0.939 0.813 0.939 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.933 0.933 transfer_pw2rs 487 13.2 0.004 0.005 0.902 0.906 transfer_rs2pw_50 119 11.7 0.193 0.199 0.818 0.899 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.876 0.889 cp_dbcsr_cholesky_decompose 11 12.9 0.292 0.297 0.853 0.866 make_images_sizes 4572 15.5 0.005 0.005 0.607 0.864 mp_alltoall_i44 4572 16.5 0.603 0.859 0.603 0.859 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=42.451000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=466.181818, yerr=2.208174 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_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 521.842688E+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 0.017 0.046 33.163 33.165 qs_mol_dyn_low 1 2.0 0.003 0.005 32.830 32.865 qs_forces 11 3.9 0.003 0.003 32.305 32.306 qs_energies 11 4.9 0.006 0.035 30.756 30.757 scf_env_do_scf 11 5.9 0.002 0.009 25.792 25.792 scf_env_do_scf_inner_loop 108 6.5 0.008 0.052 23.271 23.274 dbcsr_multiply_generic 2286 12.5 0.099 0.122 17.281 17.365 velocity_verlet 10 3.0 0.001 0.002 15.652 15.668 qs_scf_new_mos 108 7.5 0.001 0.001 15.588 15.612 qs_scf_loop_do_ot 108 8.5 0.001 0.007 15.587 15.611 ot_scf_mini 108 9.5 0.003 0.011 14.834 14.860 multiply_cannon 2286 13.5 0.195 0.198 13.960 14.666 multiply_cannon_loop 2286 14.5 0.860 0.900 13.165 13.891 ot_mini 108 10.5 0.001 0.001 9.181 9.213 qs_ot_get_derivative 108 11.5 0.001 0.001 7.671 7.699 multiply_cannon_multrec 18288 15.5 1.895 2.946 6.977 7.361 rebuild_ks_matrix 119 8.3 0.000 0.000 5.974 5.988 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 5.973 5.987 dbcsr_mm_accdrv_process 38222 16.0 4.901 5.669 4.988 5.733 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.258 5.271 mp_waitall_1 158411 16.6 2.687 3.750 2.687 3.750 init_scf_run 11 5.9 0.000 0.001 3.734 3.734 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.733 3.734 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.004 3.561 sum_up_and_integrate 119 10.3 0.001 0.002 3.534 3.541 integrate_v_rspace 119 11.3 0.003 0.003 3.521 3.530 qs_ot_get_p 119 10.4 0.001 0.002 3.433 3.457 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.282 3.300 calculate_rho_elec 119 8.7 0.031 0.032 3.282 3.300 calculate_first_density_matrix 1 7.0 0.000 0.003 2.516 2.518 init_scf_loop 11 6.9 0.005 0.036 2.495 2.501 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.515 2.400 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.971 2.371 apply_single 119 13.6 0.000 0.000 1.971 2.371 qs_ot_p2m_diag 50 11.0 0.012 0.014 2.263 2.268 make_m2s 4572 13.5 0.044 0.045 1.847 2.002 density_rs2pw 119 9.7 0.004 0.004 1.843 1.969 cp_dbcsr_syevd 50 12.0 0.003 0.006 1.967 1.968 grid_integrate_task_list 119 12.3 1.814 1.930 1.814 1.930 make_images 4572 14.5 0.192 0.204 1.761 1.915 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.912 1.914 calculate_dm_sparse 119 9.5 0.000 0.001 1.881 1.892 pw_transfer 1439 11.6 0.065 0.069 1.855 1.866 acc_transpose_blocks 18288 15.5 0.078 0.080 1.775 1.840 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.762 1.774 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.723 1.728 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.716 1.725 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.714 1.723 cp_fm_diag_elpa_base 50 14.0 1.690 1.702 1.712 1.721 prepare_preconditioner 11 7.9 0.000 0.000 1.657 1.660 make_preconditioner 11 8.9 0.000 0.001 1.657 1.660 make_full_inverse_cholesky 11 9.9 0.001 0.006 1.513 1.603 mp_sum_l 7287 12.8 1.098 1.516 1.098 1.516 ot_diis_step 108 11.5 0.011 0.011 1.495 1.495 fft_wrap_pw1pw2_140 487 13.2 0.178 0.182 1.420 1.432 potential_pw2rs 119 12.3 0.007 0.008 1.331 1.335 fft3d_ps 1201 14.6 0.546 0.568 1.316 1.327 grid_collocate_task_list 119 9.7 1.233 1.320 1.233 1.320 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.223 1.229 multiply_cannon_sync_h2d 18288 15.5 1.068 1.209 1.068 1.209 wfi_extrapolate 11 7.9 0.001 0.001 1.163 1.163 transfer_rs2pw 487 10.6 0.005 0.005 1.004 1.132 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.446 1.047 acc_transpose_blocks_kernels 18288 16.5 0.215 0.223 0.941 0.980 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.969 0.970 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.949 0.967 make_images_data 4572 15.5 0.047 0.051 0.773 0.940 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.671 0.855 mp_alltoall_d11v 2130 13.8 0.656 0.830 0.656 0.830 acc_transpose_blocks_sync 54864 16.5 0.736 0.806 0.736 0.806 transfer_pw2rs 487 13.2 0.004 0.004 0.803 0.805 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.787 0.789 jit_kernel_multiply 4 15.4 0.029 0.780 0.029 0.780 jit_kernel_transpose 5 15.6 0.726 0.768 0.726 0.768 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.670 0.732 cp_fm_cholesky_invert 11 10.9 0.690 0.694 0.690 0.694 mp_alltoall_z22v 1201 16.6 0.622 0.691 0.622 0.691 mp_waitany 9880 13.7 0.551 0.681 0.551 0.681 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.165000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_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 556.507136E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.047 37.922 37.924 qs_mol_dyn_low 1 2.0 0.003 0.003 37.664 37.715 qs_forces 11 3.9 0.002 0.002 36.909 36.909 qs_energies 11 4.9 0.002 0.002 35.210 35.217 scf_env_do_scf 11 5.9 0.000 0.001 29.478 29.479 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.987 25.987 dbcsr_multiply_generic 2286 12.5 0.103 0.105 20.663 20.770 velocity_verlet 10 3.0 0.002 0.003 18.996 19.027 qs_scf_new_mos 108 7.5 0.001 0.001 17.943 18.003 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.942 18.003 multiply_cannon 2286 13.5 0.219 0.227 16.952 17.357 ot_scf_mini 108 9.5 0.002 0.003 16.912 16.960 multiply_cannon_loop 2286 14.5 1.520 1.608 16.016 16.268 ot_mini 108 10.5 0.001 0.001 10.503 10.562 multiply_cannon_multrec 27432 15.5 2.487 3.159 9.504 9.844 qs_ot_get_derivative 108 11.5 0.001 0.001 8.647 8.697 dbcsr_mm_accdrv_process 47916 15.9 6.404 8.195 6.916 8.257 rebuild_ks_matrix 119 8.3 0.000 0.000 6.384 6.430 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.016 6.383 6.430 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.664 5.705 init_scf_run 11 5.9 0.000 0.001 4.381 4.381 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.381 4.381 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.230 3.678 qs_ot_get_p 119 10.4 0.001 0.001 3.572 3.640 sum_up_and_integrate 119 10.3 0.001 0.001 3.467 3.474 init_scf_loop 11 6.9 0.000 0.000 3.472 3.472 integrate_v_rspace 119 11.3 0.003 0.003 3.456 3.463 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.354 3.392 calculate_rho_elec 119 8.7 0.040 0.046 3.353 3.391 calculate_first_density_matrix 1 7.0 0.000 0.000 3.002 3.003 acc_transpose_blocks 27432 15.5 0.118 0.121 2.619 2.745 prepare_preconditioner 11 7.9 0.000 0.000 2.512 2.545 make_preconditioner 11 8.9 0.000 0.000 2.512 2.545 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.117 2.452 calculate_dm_sparse 119 9.5 0.000 0.001 2.396 2.449 mp_waitall_1 137007 16.6 1.884 2.401 1.884 2.401 make_m2s 4572 13.5 0.054 0.056 2.268 2.368 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.303 2.329 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.147 2.320 apply_single 119 13.6 0.000 0.000 2.147 2.319 make_images 4572 14.5 0.273 0.335 2.160 2.257 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.243 2.244 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.181 2.192 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.071 2.086 pw_transfer 1439 11.6 0.065 0.069 1.974 2.008 grid_integrate_task_list 119 12.3 1.831 1.920 1.831 1.920 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.881 1.918 density_rs2pw 119 9.7 0.004 0.004 1.808 1.887 ot_diis_step 108 11.5 0.012 0.013 1.816 1.816 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.786 1.787 fft_wrap_pw1pw2_140 487 13.2 0.228 0.233 1.558 1.596 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.553 1.563 cp_fm_diag_elpa_base 50 14.0 1.518 1.537 1.550 1.561 acc_transpose_blocks_sync 82296 16.5 1.413 1.521 1.413 1.521 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.492 1.504 fft3d_ps 1201 14.6 0.592 0.653 1.369 1.395 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.835 1.394 grid_collocate_task_list 119 9.7 1.249 1.339 1.249 1.339 mp_sum_l 7287 12.8 1.021 1.328 1.021 1.328 wfi_extrapolate 11 7.9 0.001 0.001 1.324 1.325 jit_kernel_multiply 6 16.0 0.446 1.323 0.446 1.323 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.284 1.300 potential_pw2rs 119 12.3 0.009 0.009 1.270 1.275 cp_fm_upper_to_full 72 14.2 0.832 1.183 0.832 1.183 qs_energies_init_hamiltonians 11 5.9 0.023 0.026 1.090 1.090 acc_transpose_blocks_kernels 27432 16.5 0.270 0.279 1.058 1.080 dbcsr_complete_redistribute 329 12.2 0.124 0.152 0.773 1.051 transfer_rs2pw 487 10.6 0.004 0.005 0.885 0.966 make_images_data 4572 15.5 0.049 0.052 0.836 0.949 hybrid_alltoall_any 4725 16.4 0.066 0.156 0.710 0.887 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.874 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.594 0.871 mp_alltoall_d11v 2130 13.8 0.715 0.835 0.715 0.835 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.808 0.812 jit_kernel_transpose 5 15.6 0.789 0.809 0.789 0.809 md_output 10 3.0 0.000 0.000 0.035 0.786 mp_sum_dm 438 4.9 0.723 0.762 0.723 0.762 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.924000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.818182, yerr=2.289032 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_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 633.966592E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.008 0.026 32.609 32.610 qs_mol_dyn_low 1 2.0 0.003 0.003 32.286 32.294 qs_forces 11 3.9 0.002 0.002 32.150 32.151 qs_energies 11 4.9 0.001 0.001 30.446 30.448 scf_env_do_scf 11 5.9 0.002 0.009 25.235 25.235 scf_env_do_scf_inner_loop 108 6.5 0.008 0.020 22.752 22.753 velocity_verlet 10 3.0 0.002 0.002 14.728 14.735 dbcsr_multiply_generic 2286 12.5 0.096 0.099 14.483 14.613 qs_scf_new_mos 108 7.5 0.001 0.001 12.859 12.958 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.859 12.958 ot_scf_mini 108 9.5 0.002 0.002 12.072 12.100 multiply_cannon 2286 13.5 0.225 0.227 11.311 11.898 multiply_cannon_loop 2286 14.5 0.646 0.672 10.389 10.615 rebuild_ks_matrix 119 8.3 0.000 0.000 7.309 7.339 qs_ks_build_kohn_sham_matrix 119 9.3 0.028 0.040 7.308 7.338 ot_mini 108 10.5 0.001 0.001 7.291 7.325 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.670 6.698 multiply_cannon_multrec 9144 15.5 1.736 1.923 6.397 6.665 qs_ot_get_derivative 108 11.5 0.001 0.001 5.998 6.027 dbcsr_mm_accdrv_process 12550 15.8 3.681 4.615 4.550 4.716 sum_up_and_integrate 119 10.3 0.001 0.002 4.181 4.188 integrate_v_rspace 119 11.3 0.003 0.003 4.171 4.178 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.988 4.007 calculate_rho_elec 119 8.7 0.060 0.061 3.988 4.007 init_scf_run 11 5.9 0.000 0.001 3.778 3.778 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.778 3.778 mp_waitall_1 115863 16.7 2.586 3.168 2.586 3.168 qs_ot_get_p 119 10.4 0.002 0.002 2.901 2.941 pw_transfer 1439 11.6 0.065 0.068 2.831 2.841 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.737 2.750 calculate_first_density_matrix 1 7.0 0.000 0.001 2.640 2.641 init_scf_loop 11 6.9 0.000 0.000 2.462 2.464 fft_wrap_pw1pw2_140 487 13.2 0.325 0.331 2.383 2.398 calculate_dm_sparse 119 9.5 0.000 0.001 2.150 2.252 make_m2s 4572 13.5 0.033 0.034 2.086 2.226 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.134 2.147 make_images 4572 14.5 0.269 0.300 1.996 2.136 fft3d_ps 1201 14.6 0.651 0.665 2.089 2.098 potential_pw2rs 119 12.3 0.010 0.011 1.957 1.960 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.914 1.950 grid_integrate_task_list 119 12.3 1.870 1.917 1.870 1.917 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.900 1.904 density_rs2pw 119 9.7 0.003 0.003 1.765 1.855 prepare_preconditioner 11 7.9 0.000 0.000 1.715 1.724 make_preconditioner 11 8.9 0.000 0.000 1.715 1.724 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.715 1.717 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.653 1.654 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.608 1.638 jit_kernel_multiply 7 15.6 0.831 1.523 0.831 1.523 acc_transpose_blocks 9144 15.5 0.042 0.044 1.480 1.515 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.495 1.502 mp_alltoall_d11v 2130 13.8 1.419 1.489 1.419 1.489 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.459 1.464 grid_collocate_task_list 119 9.7 1.296 1.407 1.296 1.407 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.363 1.371 cp_fm_diag_elpa_base 50 14.0 1.336 1.351 1.361 1.370 mp_alltoall_z22v 1201 16.6 1.312 1.347 1.312 1.347 transfer_pw2rs 487 13.2 0.003 0.004 1.340 1.342 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.231 1.285 apply_single 119 13.6 0.000 0.000 1.230 1.285 ot_diis_step 108 11.5 0.013 0.013 1.279 1.279 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.220 1.221 make_images_data 4572 15.5 0.043 0.047 0.919 1.129 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.891 1.093 wfi_extrapolate 11 7.9 0.001 0.001 1.087 1.087 x_to_yz 595 16.2 0.062 0.065 1.057 1.065 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.876 0.925 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.440 0.916 transfer_pw2rs_50 119 14.3 0.119 0.124 0.890 0.904 rs_scatter_matrices 130 9.7 0.031 0.035 0.872 0.891 cp_fm_cholesky_invert 11 10.9 0.859 0.861 0.859 0.861 transfer_rs2pw 487 10.6 0.004 0.005 0.768 0.855 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.573 0.844 qs_ot_get_orbitals 108 10.5 0.000 0.001 0.821 0.829 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.759 0.761 mp_allgather_i34 2286 14.5 0.236 0.760 0.236 0.760 acc_transpose_blocks_sync 27432 16.5 0.722 0.756 0.722 0.756 mp_sum_l 7287 12.8 0.635 0.753 0.635 0.753 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.669 0.717 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.700 0.703 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=32.610000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=602.454545, yerr=5.246014 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_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 783.679488E+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 0.078 46.478 46.483 qs_mol_dyn_low 1 2.0 0.003 0.003 46.187 46.203 qs_forces 11 3.9 0.002 0.003 46.111 46.112 qs_energies 11 4.9 0.001 0.001 44.089 44.093 scf_env_do_scf 11 5.9 0.001 0.001 37.977 37.977 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 30.108 30.110 velocity_verlet 10 3.0 0.006 0.007 24.214 24.220 dbcsr_multiply_generic 2286 12.5 0.104 0.106 20.780 21.088 qs_scf_new_mos 108 7.5 0.001 0.001 19.906 20.059 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.906 20.058 ot_scf_mini 108 9.5 0.002 0.002 18.513 18.619 multiply_cannon 2286 13.5 0.299 0.302 15.834 16.872 multiply_cannon_loop 2286 14.5 0.863 0.890 14.498 15.289 ot_mini 108 10.5 0.001 0.001 11.384 11.503 qs_ot_get_derivative 108 11.5 0.001 0.001 9.320 9.427 multiply_cannon_multrec 9144 15.5 3.395 4.666 8.802 8.911 init_scf_loop 11 6.9 0.000 0.000 7.841 7.845 rebuild_ks_matrix 119 8.3 0.000 0.000 7.296 7.441 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.295 7.440 prepare_preconditioner 11 7.9 0.000 0.000 6.873 6.888 make_preconditioner 11 8.9 0.000 0.000 6.873 6.888 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.389 6.767 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.567 6.698 dbcsr_mm_accdrv_process 12550 15.8 4.627 6.468 5.269 6.524 cp_fm_upper_to_full 72 14.2 3.222 4.653 3.222 4.653 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.590 4.596 calculate_rho_elec 119 8.7 0.118 0.120 4.590 4.596 mp_waitall_1 94719 16.7 3.594 4.452 3.594 4.452 qs_ot_get_p 119 10.4 0.002 0.002 4.194 4.339 init_scf_run 11 5.9 0.000 0.001 4.049 4.050 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.049 4.049 sum_up_and_integrate 119 10.3 0.001 0.001 3.871 3.878 integrate_v_rspace 119 11.3 0.004 0.004 3.861 3.867 make_m2s 4572 13.5 0.037 0.038 3.227 3.447 make_images 4572 14.5 0.353 0.384 3.106 3.326 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.834 3.237 pw_transfer 1439 11.6 0.068 0.069 3.037 3.045 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.938 2.945 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.842 2.897 dbcsr_complete_redistribute 329 12.2 0.301 0.305 2.047 2.873 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.788 2.791 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.968 2.669 calculate_first_density_matrix 1 7.0 0.000 0.000 2.585 2.587 calculate_dm_sparse 119 9.5 0.000 0.000 2.479 2.581 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.744 2.568 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.208 2.521 apply_single 119 13.6 0.000 0.000 2.208 2.521 fft_wrap_pw1pw2_140 487 13.2 0.651 0.655 2.502 2.512 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.429 2.429 density_rs2pw 119 9.7 0.003 0.003 2.294 2.321 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.479 2.299 mp_alltoall_i22 627 13.8 1.479 2.286 1.479 2.286 grid_integrate_task_list 119 12.3 2.075 2.138 2.075 2.138 mp_sum_l 7287 12.8 1.170 2.125 1.170 2.125 ot_diis_step 108 11.5 0.015 0.015 2.037 2.037 make_images_data 4572 15.5 0.047 0.050 1.629 1.988 acc_transpose_blocks 9144 15.5 0.045 0.046 1.826 1.874 hybrid_alltoall_any 4725 16.4 0.092 0.152 1.565 1.867 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.861 1.862 fft3d_ps 1201 14.6 0.871 0.892 1.857 1.861 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 1.775 1.776 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.585 1.613 grid_collocate_task_list 119 9.7 1.518 1.553 1.518 1.553 mp_alltoall_d11v 2130 13.8 1.509 1.539 1.509 1.539 cp_fm_cholesky_invert 11 10.9 1.429 1.432 1.429 1.432 wfi_extrapolate 11 7.9 0.001 0.001 1.396 1.396 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.370 1.371 cp_fm_diag_elpa_base 50 14.0 1.223 1.276 1.369 1.369 potential_pw2rs 119 12.3 0.014 0.014 1.365 1.366 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.310 1.331 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.246 1.253 acc_transpose_blocks_sync 27432 16.5 1.101 1.142 1.101 1.142 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.103 1.119 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.995 1.033 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.938 0.952 qs_create_task_list 11 7.9 0.001 0.001 0.930 0.944 generate_qs_task_list 11 8.9 0.367 0.387 0.930 0.944 make_images_sizes 4572 15.5 0.005 0.005 0.496 0.936 mp_alltoall_i44 4572 16.5 0.492 0.932 0.492 0.932 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=46.483000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=728.727273, yerr=21.742672 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_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 503.410688E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.031 81.257 81.260 qs_mol_dyn_low 1 2.0 0.005 0.020 80.728 80.768 qs_forces 11 3.9 0.006 0.029 80.582 80.587 qs_energies 11 4.9 0.021 0.095 77.607 77.624 scf_env_do_scf 11 5.9 0.001 0.001 68.507 68.510 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 63.118 63.119 dbcsr_multiply_generic 2055 12.4 0.104 0.106 50.269 50.511 qs_scf_new_mos 99 7.5 0.000 0.001 46.249 46.384 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.248 46.383 ot_scf_mini 99 9.5 0.002 0.003 43.915 43.996 multiply_cannon 2055 13.4 0.183 0.188 42.284 43.274 multiply_cannon_loop 2055 14.4 1.795 1.833 41.354 42.332 velocity_verlet 10 3.0 0.004 0.009 42.081 42.086 ot_mini 99 10.5 0.001 0.001 25.963 26.054 qs_ot_get_derivative 99 11.5 0.001 0.001 19.119 19.189 multiply_cannon_multrec 49320 15.4 11.463 12.144 17.505 18.178 rebuild_ks_matrix 110 8.3 0.000 0.001 14.466 14.567 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.021 14.466 14.567 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.635 12.726 mp_waitall_1 220248 16.4 10.227 11.204 10.227 11.204 multiply_cannon_sync_h2d 49320 15.4 9.729 10.293 9.729 10.293 qs_ot_get_p 110 10.4 0.001 0.002 9.548 9.699 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.260 7.755 apply_single 110 13.6 0.000 0.000 7.260 7.754 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.108 7.672 multiply_cannon_metrocomm3 49320 15.4 0.083 0.086 6.260 7.257 sum_up_and_integrate 110 10.3 0.002 0.004 6.946 6.968 integrate_v_rspace 110 11.3 0.003 0.025 6.920 6.949 init_scf_run 11 5.9 0.000 0.001 6.775 6.775 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.774 6.775 ot_diis_step 99 11.5 0.005 0.006 6.607 6.607 qs_ot_p2m_diag 48 11.0 0.013 0.019 6.473 6.492 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.226 6.344 calculate_rho_elec 110 8.6 0.021 0.025 6.225 6.344 dbcsr_mm_accdrv_process 87628 16.1 3.063 3.138 5.910 6.194 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.650 5.650 init_scf_loop 11 6.9 0.001 0.002 5.363 5.364 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.288 5.354 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.085 5.116 cp_fm_diag_elpa_base 48 14.0 5.073 5.106 5.083 5.114 mp_sum_l 6594 12.7 3.692 4.492 3.692 4.492 wfi_extrapolate 11 7.9 0.001 0.001 3.990 3.990 make_m2s 4110 13.4 0.060 0.065 3.844 3.943 make_images 4110 14.4 0.178 0.189 3.749 3.852 calculate_dm_sparse 110 9.5 0.001 0.001 3.725 3.823 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.557 3.561 density_rs2pw 110 9.6 0.004 0.004 3.284 3.483 grid_integrate_task_list 110 12.3 3.247 3.393 3.247 3.393 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.222 3.271 prepare_preconditioner 11 7.9 0.000 0.000 3.204 3.219 make_preconditioner 11 8.9 0.000 0.002 3.204 3.219 multiply_cannon_metrocomm1 49320 15.4 0.064 0.066 2.197 3.191 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.079 3.133 pw_transfer 1331 11.6 0.055 0.064 3.024 3.105 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.936 3.018 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.955 3.010 calculate_first_density_matrix 1 7.0 0.002 0.010 2.675 2.679 acc_transpose_blocks 49320 15.4 0.211 0.218 2.561 2.633 jit_kernel_multiply 13 15.9 2.570 2.632 2.570 2.632 fft_wrap_pw1pw2_140 451 13.1 0.311 0.333 2.533 2.625 potential_pw2rs 110 12.3 0.005 0.006 2.550 2.570 mp_alltoall_d11v 2046 13.8 2.044 2.492 2.044 2.492 fft3d_ps 1111 14.6 0.791 0.886 2.224 2.282 grid_collocate_task_list 110 9.6 2.161 2.270 2.161 2.270 transfer_rs2pw 451 10.6 0.005 0.006 1.948 2.147 mp_waitany 14300 13.8 1.937 2.139 1.937 2.139 mp_sum_d 3891 11.9 1.433 2.023 1.433 2.023 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.961 1.974 make_images_data 4110 15.4 0.044 0.048 1.756 1.891 hybrid_alltoall_any 4261 16.3 0.085 0.485 1.525 1.785 cp_fm_cholesky_invert 11 10.9 1.719 1.722 1.719 1.722 transfer_pw2rs 451 13.1 0.006 0.007 1.701 1.713 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.665 1.691 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.260000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.727273, yerr=3.466487 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 592.474112E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 3156821. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.061 70.475 70.476 qs_mol_dyn_low 1 2.0 0.003 0.005 70.099 70.111 qs_forces 11 3.9 0.003 0.004 70.022 70.023 qs_energies 11 4.9 0.004 0.025 66.691 66.702 scf_env_do_scf 11 5.9 0.001 0.002 57.219 57.223 scf_env_do_scf_inner_loop 99 6.5 0.005 0.025 49.245 49.246 dbcsr_multiply_generic 2055 12.4 0.115 0.119 38.861 39.041 velocity_verlet 10 3.0 0.013 0.018 35.665 35.666 qs_scf_new_mos 99 7.5 0.001 0.001 33.519 33.699 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.518 33.699 multiply_cannon 2055 13.4 0.225 0.245 32.049 33.546 ot_scf_mini 99 9.5 0.003 0.004 31.855 32.014 multiply_cannon_loop 2055 14.4 1.161 1.192 30.797 31.965 ot_mini 99 10.5 0.003 0.018 18.858 19.034 multiply_cannon_multrec 24660 15.4 7.009 8.769 14.488 16.173 rebuild_ks_matrix 110 8.3 0.000 0.001 13.546 13.644 qs_ks_build_kohn_sham_matrix 110 9.3 0.016 0.044 13.546 13.644 qs_ot_get_derivative 99 11.5 0.001 0.001 13.077 13.222 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.913 12.006 mp_waitall_1 176588 16.5 7.669 10.230 7.669 10.230 multiply_cannon_metrocomm3 24660 15.4 0.074 0.077 5.323 8.371 init_scf_loop 11 6.9 0.029 0.234 7.940 7.941 multiply_cannon_sync_h2d 24660 15.4 6.449 7.757 6.449 7.757 dbcsr_mm_accdrv_process 52282 16.1 5.843 6.748 7.308 7.657 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.723 7.455 apply_single 110 13.6 0.000 0.001 6.722 7.455 init_scf_run 11 5.9 0.000 0.001 6.826 6.826 scf_env_initial_rho_setup 11 6.9 0.024 0.060 6.826 6.826 qs_ot_get_p 110 10.4 0.001 0.002 6.354 6.524 sum_up_and_integrate 110 10.3 0.002 0.003 6.292 6.307 integrate_v_rspace 110 11.3 0.003 0.003 6.265 6.279 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.946 5.864 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.821 5.833 calculate_rho_elec 110 8.6 0.039 0.047 5.820 5.833 ot_diis_step 99 11.5 0.010 0.010 5.732 5.733 prepare_preconditioner 11 7.9 0.000 0.000 5.621 5.644 make_preconditioner 11 8.9 0.005 0.036 5.621 5.644 make_full_inverse_cholesky 11 9.9 0.032 0.252 5.081 5.376 make_m2s 4110 13.4 0.057 0.059 4.179 4.638 make_images 4110 14.4 0.406 0.464 4.068 4.523 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.470 4.490 cp_dbcsr_syevd 48 12.0 0.006 0.026 4.013 4.013 pw_transfer 1331 11.6 0.066 0.074 3.490 3.631 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.383 3.526 wfi_extrapolate 11 7.9 0.001 0.001 3.483 3.483 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.442 3.455 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.438 3.440 cp_fm_diag_elpa_base 48 14.0 3.383 3.397 3.426 3.437 calculate_dm_sparse 110 9.5 0.001 0.001 3.333 3.359 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.256 3.350 grid_integrate_task_list 110 12.3 3.174 3.312 3.174 3.312 calculate_first_density_matrix 1 7.0 0.000 0.003 3.222 3.248 density_rs2pw 110 9.6 0.004 0.004 3.052 3.230 fft_wrap_pw1pw2_140 451 13.1 0.358 0.378 2.922 3.066 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.982 3.023 hybrid_alltoall_any 4261 16.3 0.105 0.457 1.973 2.776 make_images_data 4110 15.4 0.049 0.053 2.254 2.761 mp_sum_l 6594 12.7 2.006 2.745 2.006 2.745 fft3d_ps 1111 14.6 1.117 1.349 2.475 2.631 cp_fm_cholesky_invert 11 10.9 2.492 2.498 2.492 2.498 jit_kernel_multiply 10 16.4 1.100 2.321 1.100 2.321 grid_collocate_task_list 110 9.6 2.172 2.311 2.172 2.311 potential_pw2rs 110 12.3 0.008 0.009 2.168 2.173 acc_transpose_blocks 24660 15.4 0.115 0.118 1.991 2.083 qs_energies_init_hamiltonians 11 5.9 0.025 0.044 2.023 2.024 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.948 1.972 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.903 1.911 dbcsr_complete_redistribute 325 12.2 0.329 0.412 1.514 1.908 mp_alltoall_d11v 2046 13.8 1.718 1.886 1.718 1.886 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.794 1.811 transfer_rs2pw 451 10.6 0.006 0.007 1.461 1.674 mp_allgather_i34 2055 14.4 0.558 1.569 0.558 1.569 multiply_cannon_metrocomm4 22605 15.4 0.074 0.077 0.754 1.529 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.126 1.504 mp_waitany 10164 13.8 1.245 1.491 1.245 1.491 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.389 1.491 mp_irecv_dv 57340 16.2 0.629 1.413 0.629 1.413 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.476000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.909091, yerr=5.664154 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_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 670.318592E+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 0.020 0.026 61.128 61.129 qs_mol_dyn_low 1 2.0 0.003 0.003 60.773 60.781 qs_forces 11 3.9 0.003 0.003 60.696 60.696 qs_energies 11 4.9 0.001 0.002 57.533 57.536 scf_env_do_scf 11 5.9 0.001 0.001 49.205 49.206 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 40.725 40.725 velocity_verlet 10 3.0 0.008 0.020 32.221 32.223 dbcsr_multiply_generic 2055 12.4 0.109 0.113 30.286 30.574 qs_scf_new_mos 99 7.5 0.001 0.001 26.208 26.298 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.207 26.297 multiply_cannon 2055 13.4 0.213 0.222 24.166 25.520 ot_scf_mini 99 9.5 0.002 0.003 24.962 25.069 multiply_cannon_loop 2055 14.4 0.817 0.862 23.023 24.186 ot_mini 99 10.5 0.001 0.001 14.861 14.968 rebuild_ks_matrix 110 8.3 0.000 0.000 12.069 12.201 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.024 12.068 12.201 multiply_cannon_multrec 16440 15.4 3.630 4.590 10.951 11.907 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.610 10.731 qs_ot_get_derivative 99 11.5 0.001 0.001 10.441 10.545 mp_waitall_1 139946 16.5 7.333 10.052 7.333 10.052 init_scf_loop 11 6.9 0.000 0.000 8.449 8.449 dbcsr_mm_accdrv_process 34862 16.1 5.895 6.918 7.165 7.413 multiply_cannon_metrocomm3 16440 15.4 0.046 0.047 4.704 7.351 prepare_preconditioner 11 7.9 0.000 0.000 6.697 6.716 make_preconditioner 11 8.9 0.000 0.000 6.697 6.716 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.001 6.380 sum_up_and_integrate 110 10.3 0.002 0.002 6.210 6.223 integrate_v_rspace 110 11.3 0.003 0.003 6.184 6.197 init_scf_run 11 5.9 0.000 0.001 5.934 5.934 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.934 5.934 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.692 5.704 calculate_rho_elec 110 8.6 0.059 0.059 5.692 5.703 qs_ot_get_p 110 10.4 0.002 0.007 5.307 5.440 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.898 5.428 apply_single 110 13.6 0.000 0.000 4.898 5.428 make_m2s 4110 13.4 0.049 0.051 4.106 4.471 ot_diis_step 99 11.5 0.011 0.011 4.390 4.390 make_images 4110 14.4 0.399 0.525 3.989 4.353 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.390 4.017 multiply_cannon_sync_h2d 16440 15.4 3.244 3.799 3.244 3.799 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.706 3.712 pw_transfer 1331 11.6 0.066 0.071 3.550 3.558 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.443 3.453 grid_integrate_task_list 110 12.3 3.206 3.406 3.206 3.406 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.340 3.341 density_rs2pw 110 9.6 0.004 0.004 2.928 3.157 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.147 3.148 fft_wrap_pw1pw2_140 451 13.1 0.460 0.471 2.999 3.010 wfi_extrapolate 11 7.9 0.001 0.001 2.935 2.935 calculate_first_density_matrix 1 7.0 0.000 0.001 2.908 2.909 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.803 2.859 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.822 2.832 cp_fm_diag_elpa_base 48 14.0 2.755 2.786 2.821 2.830 make_images_data 4110 15.4 0.047 0.051 2.307 2.757 hybrid_alltoall_any 4261 16.3 0.109 0.382 2.038 2.716 calculate_dm_sparse 110 9.5 0.001 0.001 2.663 2.703 cp_fm_cholesky_invert 11 10.9 2.545 2.551 2.545 2.551 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.421 2.481 multiply_cannon_metrocomm4 14385 15.4 0.047 0.051 0.894 2.476 grid_collocate_task_list 110 9.6 2.219 2.411 2.219 2.411 fft3d_ps 1111 14.6 1.118 1.130 2.359 2.372 mp_irecv_dv 48980 15.7 0.820 2.338 0.820 2.338 mp_sum_l 6594 12.7 1.386 2.101 1.386 2.101 mp_alltoall_d11v 2046 13.8 1.760 2.075 1.760 2.075 potential_pw2rs 110 12.3 0.011 0.011 2.058 2.064 jit_kernel_multiply 6 16.6 0.866 1.961 0.866 1.961 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.953 1.954 dbcsr_complete_redistribute 325 12.2 0.334 0.391 1.480 1.949 cp_fm_upper_to_full 70 14.2 1.454 1.830 1.454 1.830 acc_transpose_blocks 16440 15.4 0.078 0.080 1.665 1.737 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.650 1.663 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.602 1.615 cp_fm_cholesky_decompose 22 10.9 1.529 1.552 1.529 1.552 transfer_rs2pw 451 10.6 0.005 0.006 1.246 1.495 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.006 1.479 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.357 1.466 mp_allgather_i34 2055 14.4 0.453 1.442 0.453 1.442 mp_waitany 17072 13.8 1.143 1.426 1.143 1.426 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.259 1.268 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.129000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=635.454545, yerr=10.093775 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_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 736.972800E+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 0.008 0.026 66.185 66.186 qs_mol_dyn_low 1 2.0 0.003 0.003 65.975 65.984 qs_forces 11 3.9 0.003 0.003 65.911 65.912 qs_energies 11 4.9 0.001 0.001 62.524 62.528 scf_env_do_scf 11 5.9 0.001 0.002 54.109 54.111 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 42.372 42.372 velocity_verlet 10 3.0 0.004 0.005 36.960 36.962 dbcsr_multiply_generic 2055 12.4 0.117 0.122 31.281 31.495 qs_scf_new_mos 99 7.5 0.001 0.001 27.820 27.919 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.820 27.918 ot_scf_mini 99 9.5 0.003 0.003 26.142 26.242 multiply_cannon 2055 13.4 0.239 0.254 24.017 24.938 multiply_cannon_loop 2055 14.4 1.395 1.460 22.684 23.355 ot_mini 99 10.5 0.001 0.001 15.327 15.451 multiply_cannon_multrec 24660 15.4 4.250 7.232 13.478 14.722 rebuild_ks_matrix 110 8.3 0.000 0.000 11.922 12.019 qs_ks_build_kohn_sham_matrix 110 9.3 0.021 0.028 11.921 12.019 init_scf_loop 11 6.9 0.000 0.000 11.698 11.699 qs_ot_get_derivative 99 11.5 0.001 0.001 11.186 11.291 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.529 10.615 dbcsr_mm_accdrv_process 52304 16.0 7.904 9.832 9.071 10.186 prepare_preconditioner 11 7.9 0.000 0.000 9.778 9.799 make_preconditioner 11 8.9 0.000 0.000 9.778 9.799 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.965 9.476 mp_waitall_1 121746 16.5 4.380 6.298 4.380 6.298 sum_up_and_integrate 110 10.3 0.002 0.002 6.167 6.181 integrate_v_rspace 110 11.3 0.003 0.003 6.141 6.154 qs_ot_get_p 110 10.4 0.008 0.015 5.836 6.025 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.822 5.832 calculate_rho_elec 110 8.6 0.078 0.081 5.822 5.831 make_m2s 4110 13.4 0.059 0.061 5.476 5.780 init_scf_run 11 5.9 0.000 0.001 5.685 5.686 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.685 5.685 make_images 4110 14.4 0.574 0.695 5.335 5.633 cp_fm_upper_to_full 70 14.2 3.410 4.969 3.410 4.969 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.063 4.159 apply_single 110 13.6 0.000 0.000 4.063 4.159 ot_diis_step 99 11.5 0.011 0.011 4.105 4.105 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.861 3.880 dbcsr_complete_redistribute 325 12.2 0.421 0.472 2.661 3.836 pw_transfer 1331 11.6 0.065 0.073 3.729 3.764 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.623 3.660 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.542 3.610 grid_integrate_task_list 110 12.3 3.273 3.405 3.273 3.405 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.321 3.382 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.162 3.329 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.303 3.303 fft_wrap_pw1pw2_140 451 13.1 0.531 0.546 3.136 3.173 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.459 3.102 make_images_data 4110 15.4 0.050 0.053 2.760 3.084 density_rs2pw 110 9.6 0.004 0.004 2.929 3.073 hybrid_alltoall_any 4261 16.3 0.123 0.458 2.314 3.017 calculate_dm_sparse 110 9.5 0.001 0.001 2.972 3.016 wfi_extrapolate 11 7.9 0.001 0.001 2.976 2.977 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.804 2.958 mp_alltoall_i22 605 13.7 1.700 2.905 1.700 2.905 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.838 2.839 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.781 2.790 cp_fm_diag_elpa_base 48 14.0 2.634 2.692 2.779 2.788 multiply_cannon_sync_h2d 24660 15.4 2.378 2.654 2.378 2.654 calculate_first_density_matrix 1 7.0 0.000 0.001 2.614 2.617 cp_fm_cholesky_invert 11 10.9 2.557 2.567 2.557 2.567 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.473 2.505 acc_transpose_blocks 24660 15.4 0.110 0.114 2.397 2.497 fft3d_ps 1111 14.6 1.152 1.192 2.448 2.465 grid_collocate_task_list 110 9.6 2.269 2.409 2.269 2.409 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.231 2.232 mp_alltoall_d11v 2046 13.8 1.736 1.997 1.736 1.997 potential_pw2rs 110 12.3 0.013 0.013 1.988 1.991 mp_sum_l 6594 12.7 1.095 1.824 1.095 1.824 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.780 1.817 jit_kernel_multiply 8 15.9 0.824 1.816 0.824 1.816 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.613 1.717 cp_fm_cholesky_decompose 22 10.9 1.608 1.656 1.608 1.656 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.632 1.644 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.534 1.549 multiply_cannon_metrocomm4 20550 15.4 0.059 0.061 0.832 1.524 acc_transpose_blocks_sync 73980 16.4 1.376 1.468 1.376 1.468 mp_irecv_dv 62702 16.1 0.730 1.441 0.730 1.441 mp_allgather_i34 2055 14.4 0.416 1.376 0.416 1.376 transfer_rs2pw 451 10.6 0.005 0.006 1.135 1.341 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.186000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=696.454545, yerr=9.680004 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_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 869.240832E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.073 0.108 55.857 55.859 qs_mol_dyn_low 1 2.0 0.003 0.003 55.199 55.230 qs_forces 11 3.9 0.003 0.003 55.134 55.135 qs_energies 11 4.9 0.001 0.001 51.487 51.492 scf_env_do_scf 11 5.9 0.000 0.001 43.042 43.042 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.455 35.456 velocity_verlet 10 3.0 0.008 0.009 30.861 30.863 dbcsr_multiply_generic 2055 12.4 0.107 0.109 23.666 23.818 qs_scf_new_mos 99 7.5 0.001 0.001 20.886 20.933 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.885 20.933 ot_scf_mini 99 9.5 0.002 0.002 19.637 19.663 multiply_cannon 2055 13.4 0.240 0.254 18.169 19.397 multiply_cannon_loop 2055 14.4 0.602 0.622 16.902 17.109 rebuild_ks_matrix 110 8.3 0.000 0.000 11.615 11.638 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.615 11.638 ot_mini 99 10.5 0.001 0.001 11.119 11.138 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.333 10.352 multiply_cannon_multrec 8220 15.4 3.224 4.420 8.128 8.987 init_scf_loop 11 6.9 0.000 0.000 7.542 7.543 qs_ot_get_derivative 99 11.5 0.001 0.001 7.436 7.462 mp_waitall_1 103326 16.6 5.772 7.402 5.772 7.402 sum_up_and_integrate 110 10.3 0.001 0.002 6.178 6.192 integrate_v_rspace 110 11.3 0.003 0.003 6.152 6.165 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.005 6.013 calculate_rho_elec 110 8.6 0.113 0.113 6.004 6.012 prepare_preconditioner 11 7.9 0.000 0.000 5.899 5.904 make_preconditioner 11 8.9 0.000 0.000 5.899 5.903 dbcsr_mm_accdrv_process 17442 15.9 3.435 4.749 4.764 5.827 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.495 5.567 init_scf_run 11 5.9 0.000 0.001 5.330 5.330 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.330 5.330 qs_ot_get_p 110 10.4 0.002 0.002 4.589 4.602 make_m2s 4110 13.4 0.038 0.039 4.176 4.469 make_images 4110 14.4 0.645 0.710 4.046 4.338 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 3.000 4.261 pw_transfer 1331 11.6 0.066 0.071 4.052 4.060 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.944 3.956 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.724 3.775 apply_single 110 13.6 0.000 0.000 3.723 3.775 ot_diis_step 99 11.5 0.012 0.012 3.659 3.659 grid_integrate_task_list 110 12.3 3.376 3.515 3.376 3.515 fft_wrap_pw1pw2_140 451 13.1 0.720 0.733 3.433 3.446 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.250 3.254 density_rs2pw 110 9.6 0.004 0.004 2.989 3.098 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.913 2.914 cp_fm_cholesky_invert 11 10.9 2.832 2.835 2.832 2.835 hybrid_alltoall_any 4261 16.3 0.199 0.852 2.250 2.770 wfi_extrapolate 11 7.9 0.001 0.001 2.685 2.685 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.666 2.667 make_images_data 4110 15.4 0.041 0.047 2.236 2.618 calculate_dm_sparse 110 9.5 0.001 0.001 2.566 2.603 calculate_first_density_matrix 1 7.0 0.000 0.000 2.548 2.549 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.529 2.530 grid_collocate_task_list 110 9.6 2.368 2.518 2.368 2.518 fft3d_ps 1111 14.6 1.290 1.297 2.487 2.502 multiply_cannon_sync_h2d 8220 15.4 2.388 2.476 2.388 2.476 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.438 2.474 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.364 2.372 cp_fm_diag_elpa_base 48 14.0 2.308 2.335 2.362 2.370 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.164 2.179 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.794 2.010 potential_pw2rs 110 12.3 0.015 0.016 1.987 1.989 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.886 1.894 mp_alltoall_d11v 2046 13.8 1.675 1.865 1.675 1.865 cp_fm_cholesky_decompose 22 10.9 1.666 1.699 1.666 1.699 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.609 1.612 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.493 1.605 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.519 1.533 mp_allgather_i34 2055 14.4 0.435 1.522 0.435 1.522 jit_kernel_multiply 6 15.5 1.013 1.498 1.013 1.498 dbcsr_complete_redistribute 325 12.2 0.561 0.593 1.385 1.459 acc_transpose_blocks 8220 15.4 0.040 0.041 1.368 1.394 qs_create_task_list 11 7.9 0.000 0.001 1.205 1.303 generate_qs_task_list 11 8.9 0.375 0.442 1.204 1.302 transfer_rs2pw 451 10.6 0.005 0.005 1.093 1.207 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.160 1.193 mp_waitany 9240 13.8 1.052 1.159 1.052 1.159 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.859000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=820.545455, yerr=12.999046 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_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 1.392845E+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 0.043 0.069 92.135 92.142 qs_mol_dyn_low 1 2.0 0.003 0.003 91.729 91.743 qs_forces 11 3.9 0.050 0.056 91.659 91.659 qs_energies 11 4.9 0.001 0.002 86.701 86.710 scf_env_do_scf 11 5.9 0.000 0.001 76.175 76.175 velocity_verlet 10 3.0 0.011 0.012 59.362 59.368 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 46.873 46.876 dbcsr_multiply_generic 2055 12.4 0.124 0.128 31.714 31.933 init_scf_loop 11 6.9 0.000 0.000 29.231 29.237 qs_scf_new_mos 99 7.5 0.001 0.001 28.667 28.758 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.666 28.757 ot_scf_mini 99 9.5 0.002 0.002 26.846 26.932 prepare_preconditioner 11 7.9 0.000 0.000 26.828 26.842 make_preconditioner 11 8.9 0.000 0.000 26.828 26.842 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.018 26.287 multiply_cannon 2055 13.4 0.334 0.360 23.708 24.429 multiply_cannon_loop 2055 14.4 0.826 0.836 21.735 22.300 cp_fm_upper_to_full 70 14.2 13.076 18.592 13.076 18.592 ot_mini 99 10.5 0.001 0.001 14.903 14.989 rebuild_ks_matrix 110 8.3 0.000 0.001 14.473 14.610 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.472 14.610 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.839 12.934 dbcsr_complete_redistribute 325 12.2 1.026 1.060 7.431 10.698 mp_waitall_1 84994 16.7 8.977 10.363 8.977 10.363 qs_ot_get_derivative 99 11.5 0.001 0.001 10.120 10.209 multiply_cannon_multrec 8220 15.4 4.038 4.185 9.741 9.871 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.416 9.689 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.794 9.031 mp_alltoall_i22 605 13.7 5.452 8.761 5.452 8.761 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.621 7.650 calculate_rho_elec 110 8.6 0.223 0.223 7.621 7.649 sum_up_and_integrate 110 10.3 0.002 0.002 7.045 7.061 integrate_v_rspace 110 11.3 0.004 0.004 7.017 7.033 make_m2s 4110 13.4 0.043 0.044 5.868 6.437 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 5.913 6.432 qs_ot_get_p 110 10.4 0.002 0.002 6.143 6.252 make_images 4110 14.4 0.883 0.935 5.679 6.249 init_scf_run 11 5.9 0.000 0.001 6.009 6.010 scf_env_initial_rho_setup 11 6.9 0.013 0.014 6.009 6.009 pw_transfer 1331 11.6 0.074 0.075 5.896 5.902 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.779 5.785 dbcsr_mm_accdrv_process 11614 15.7 3.945 4.139 5.555 5.756 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.075 5.597 apply_single 110 13.6 0.000 0.000 5.074 5.596 cp_fm_cholesky_invert 11 10.9 5.430 5.435 5.430 5.435 fft_wrap_pw1pw2_140 451 13.1 1.368 1.373 5.069 5.073 ot_diis_step 99 11.5 0.016 0.017 4.760 4.760 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.192 4.200 density_rs2pw 110 9.6 0.004 0.004 4.049 4.071 make_images_data 4110 15.4 0.046 0.049 3.155 3.929 hybrid_alltoall_any 4261 16.3 0.264 0.565 3.035 3.842 qs_energies_init_hamiltonians 11 5.9 0.094 0.138 3.830 3.839 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.285 3.789 grid_integrate_task_list 110 12.3 3.701 3.767 3.701 3.767 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.671 3.672 fft3d_ps 1111 14.6 1.887 1.899 3.483 3.495 wfi_extrapolate 11 7.9 0.001 0.001 3.442 3.442 calculate_dm_sparse 110 9.5 0.001 0.001 3.192 3.219 multiply_cannon_sync_h2d 8220 15.4 3.125 3.145 3.125 3.145 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.073 3.073 cp_fm_diag_elpa_base 48 14.0 2.528 2.730 3.071 3.071 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.984 2.989 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.720 2.768 grid_collocate_task_list 110 9.6 2.686 2.723 2.686 2.723 potential_pw2rs 110 12.3 0.021 0.022 2.507 2.511 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.416 2.451 calculate_first_density_matrix 1 7.0 0.000 0.000 2.437 2.438 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.231 2.280 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.209 2.265 mp_alltoall_d11v 2046 13.8 2.154 2.229 2.154 2.229 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.115 2.201 cp_fm_cholesky_decompose 22 10.9 1.998 2.019 1.998 2.019 qs_create_task_list 11 7.9 0.011 0.013 1.888 1.930 qs_ot_get_orbitals 99 10.5 0.000 0.000 1.918 1.929 generate_qs_task_list 11 8.9 0.731 0.783 1.877 1.917 mp_allgather_i34 2055 14.4 0.724 1.876 0.724 1.876 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.807 1.857 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=92.142000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1254.727273, yerr=61.826871 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 631.734272E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 57647. MP_Allreduce 11104 795. MP_Sync 87 MP_Alltoall 2226 1932876. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.077 207.656 207.687 qs_mol_dyn_low 1 2.0 0.007 0.031 206.956 207.013 qs_forces 11 3.9 0.006 0.017 206.761 206.764 qs_energies 11 4.9 0.011 0.050 201.105 201.152 scf_env_do_scf 11 5.9 0.001 0.002 184.174 184.178 scf_env_do_scf_inner_loop 117 6.6 0.005 0.019 163.803 163.805 dbcsr_multiply_generic 2507 12.6 0.196 0.204 126.181 126.892 qs_scf_new_mos 117 7.6 0.001 0.001 124.463 124.712 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.462 124.711 velocity_verlet 10 3.0 0.003 0.012 123.108 123.111 ot_scf_mini 117 9.6 0.004 0.012 117.832 118.041 multiply_cannon 2507 13.6 0.236 0.242 102.013 104.530 multiply_cannon_loop 2507 14.6 2.393 2.442 99.891 102.653 ot_mini 117 10.6 0.002 0.010 66.341 66.552 multiply_cannon_multrec 60168 15.6 31.957 34.094 41.764 43.805 qs_ot_get_derivative 117 11.6 0.001 0.002 41.346 41.556 rebuild_ks_matrix 128 8.3 0.001 0.001 33.551 34.165 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.034 33.551 34.165 mp_waitall_1 267128 16.5 29.016 32.291 29.016 32.291 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.138 30.667 qs_ot_get_p 128 10.4 0.003 0.011 29.152 29.523 multiply_cannon_sync_h2d 60168 15.6 26.485 28.963 26.485 28.963 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.550 25.337 apply_single 128 13.6 0.001 0.001 24.550 25.337 ot_diis_step 117 11.6 0.008 0.010 24.733 24.735 qs_ot_p2m_diag 83 11.4 0.081 0.099 22.534 22.613 init_scf_loop 11 6.9 0.001 0.004 20.298 20.299 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.777 19.955 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.618 19.622 multiply_cannon_metrocomm3 60168 15.6 0.118 0.125 16.202 18.543 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.670 16.699 cp_fm_diag_elpa_base 83 14.4 16.599 16.642 16.666 16.695 prepare_preconditioner 11 7.9 0.000 0.000 15.623 15.705 make_preconditioner 11 8.9 0.001 0.004 15.623 15.705 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.817 15.050 make_m2s 5014 13.6 0.104 0.112 13.894 14.251 make_images 5014 14.6 0.401 0.421 13.715 14.083 sum_up_and_integrate 128 10.3 0.003 0.006 13.897 13.912 integrate_v_rspace 128 11.3 0.004 0.011 13.838 13.856 qs_rho_update_rho_low 128 7.7 0.001 0.004 13.111 13.319 calculate_rho_elec 128 8.7 0.045 0.064 13.110 13.318 init_scf_run 11 5.9 0.000 0.001 12.629 12.629 scf_env_initial_rho_setup 11 6.9 0.001 0.003 12.629 12.629 mp_sum_l 7950 12.9 9.056 10.359 9.056 10.359 dbcsr_mm_accdrv_process 124484 16.2 4.702 4.843 9.376 9.917 wfi_extrapolate 11 7.9 0.001 0.001 9.058 9.059 cp_fm_cholesky_invert 11 10.9 8.820 8.827 8.820 8.827 calculate_dm_sparse 128 9.5 0.001 0.001 8.494 8.605 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.215 8.364 multiply_cannon_metrocomm1 60168 15.6 0.092 0.097 6.354 8.260 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.139 8.215 make_images_data 5014 15.6 0.069 0.076 6.804 7.745 pw_transfer 1547 11.6 0.074 0.086 7.556 7.735 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.355 7.530 density_rs2pw 128 9.7 0.006 0.006 6.718 7.456 grid_integrate_task_list 128 12.3 7.036 7.415 7.036 7.415 hybrid_alltoall_any 5200 16.5 0.296 2.273 5.976 7.220 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.777 6.790 fft_wrap_pw1pw2_140 523 13.2 0.864 0.903 6.479 6.647 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.896 5.985 fft3d_ps 1291 14.7 2.192 2.837 5.350 5.668 mp_alltoall_d11v 2415 14.1 4.409 5.429 4.409 5.429 grid_collocate_task_list 128 9.7 4.832 5.253 4.832 5.253 cp_fm_cholesky_decompose 22 10.9 4.569 4.582 4.569 4.582 potential_pw2rs 128 12.3 0.009 0.010 4.415 4.429 mp_sum_d 4476 12.1 3.560 4.315 3.560 4.315 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.687000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.272727, yerr=6.165755 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 845.713408E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 5470167. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.057 0.081 191.743 191.744 qs_mol_dyn_low 1 2.0 0.003 0.004 191.269 191.282 qs_forces 11 3.9 0.036 0.037 191.190 191.191 qs_energies 11 4.9 0.002 0.009 184.429 184.442 scf_env_do_scf 11 5.9 0.001 0.002 167.196 167.205 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 132.166 132.169 velocity_verlet 10 3.0 0.001 0.001 121.322 121.324 dbcsr_multiply_generic 2507 12.6 0.188 0.195 97.592 98.848 qs_scf_new_mos 117 7.6 0.001 0.001 93.869 94.478 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.868 94.477 ot_scf_mini 117 9.6 0.004 0.005 89.065 89.759 multiply_cannon 2507 13.6 0.502 0.561 77.431 82.049 multiply_cannon_loop 2507 14.6 1.565 1.640 74.373 76.968 ot_mini 117 10.6 0.001 0.001 50.039 50.663 mp_waitall_1 214728 16.6 24.511 38.740 24.511 38.740 multiply_cannon_multrec 30084 15.6 20.981 24.700 31.595 35.786 init_scf_loop 11 6.9 0.001 0.003 34.941 34.942 rebuild_ks_matrix 128 8.3 0.001 0.001 32.080 32.715 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.079 32.714 prepare_preconditioner 11 7.9 0.000 0.000 30.652 30.731 make_preconditioner 11 8.9 0.000 0.002 30.652 30.731 make_full_inverse_cholesky 11 9.9 0.004 0.034 29.060 29.886 multiply_cannon_metrocomm3 30084 15.6 0.100 0.107 15.935 29.840 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.851 29.433 qs_ot_get_derivative 117 11.6 0.001 0.002 28.089 28.767 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.074 23.159 apply_single 128 13.6 0.001 0.001 22.074 23.158 qs_ot_get_p 128 10.4 0.001 0.002 21.182 21.933 ot_diis_step 117 11.6 0.014 0.015 21.769 21.771 multiply_cannon_sync_h2d 30084 15.6 17.979 19.959 17.979 19.959 cp_fm_cholesky_invert 11 10.9 17.299 17.312 17.299 17.312 qs_ot_p2m_diag 83 11.4 0.188 0.216 16.301 16.335 make_m2s 5014 13.6 0.087 0.092 14.211 15.361 make_images 5014 14.6 1.165 1.367 13.996 15.143 cp_dbcsr_syevd 83 12.4 0.008 0.026 15.100 15.100 sum_up_and_integrate 128 10.3 0.002 0.005 13.926 13.956 integrate_v_rspace 128 11.3 0.003 0.004 13.865 13.899 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.303 13.375 calculate_rho_elec 128 8.7 0.088 0.105 13.302 13.374 init_scf_run 11 5.9 0.000 0.001 12.301 12.302 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.300 12.302 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.898 11.928 cp_fm_diag_elpa_base 83 14.4 11.638 11.733 11.885 11.915 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.100 11.602 multiply_cannon_metrocomm4 27577 15.6 0.099 0.112 3.880 10.673 dbcsr_mm_accdrv_process 62242 16.2 5.459 5.966 10.063 10.663 mp_irecv_dv 69486 16.3 3.680 10.281 3.680 10.281 make_images_data 5014 15.6 0.067 0.077 8.396 9.847 hybrid_alltoall_any 5200 16.5 0.348 1.520 7.072 9.182 wfi_extrapolate 11 7.9 0.001 0.001 9.017 9.017 pw_transfer 1547 11.6 0.085 0.096 8.529 8.581 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.306 8.359 cp_fm_cholesky_decompose 22 10.9 7.609 7.695 7.609 7.695 grid_integrate_task_list 128 12.3 7.190 7.610 7.190 7.610 fft_wrap_pw1pw2_140 523 13.2 0.935 0.966 7.363 7.431 density_rs2pw 128 9.7 0.006 0.006 6.995 7.401 calculate_dm_sparse 128 9.5 0.001 0.001 7.160 7.310 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.253 6.981 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.117 6.128 mp_sum_l 7950 12.9 4.166 6.069 4.166 6.069 fft3d_ps 1291 14.7 2.829 3.005 5.858 5.905 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.365 5.447 grid_collocate_task_list 128 9.7 5.002 5.387 5.002 5.387 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.222 5.380 dbcsr_complete_redistribute 395 12.7 0.769 0.855 3.424 4.783 mp_alltoall_d11v 2415 14.1 4.143 4.716 4.143 4.716 mp_allgather_i34 2507 14.6 1.572 4.695 1.572 4.695 potential_pw2rs 128 12.3 0.016 0.019 4.426 4.441 mp_sum_d 4467 12.1 2.698 4.217 2.698 4.217 cp_fm_upper_to_full 105 14.8 3.353 4.195 3.353 4.195 copy_fm_to_dbcsr 209 11.7 0.002 0.002 2.568 3.898 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.744000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=804.727273, yerr=5.361325 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 2.951022E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 4017216 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.3 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 955.043840E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1052064 MPI messages size (bytes): total size 2.737208E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.601750E+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 284544 37295751168 131072 < size <= 4194304 665856 1004519030784 4194304 < size <= 16777216 66080 937890532432 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57629. MP_Allreduce 11184 996. MP_Sync 88 MP_Alltoall 1724 9394234. 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 0.061 0.078 179.630 179.631 qs_mol_dyn_low 1 2.0 0.003 0.003 178.838 178.900 qs_forces 11 3.9 0.004 0.005 178.271 178.272 qs_energies 11 4.9 0.002 0.003 171.060 171.070 scf_env_do_scf 11 5.9 0.002 0.008 154.454 154.457 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 119.317 119.317 velocity_verlet 10 3.0 0.001 0.002 116.413 116.430 dbcsr_multiply_generic 2529 12.6 0.187 0.191 83.866 85.171 qs_scf_new_mos 118 7.6 0.001 0.001 82.070 82.483 qs_scf_loop_do_ot 118 8.6 0.001 0.001 82.069 82.482 ot_scf_mini 118 9.6 0.003 0.004 77.810 78.271 multiply_cannon 2529 13.6 0.514 0.545 63.967 68.735 multiply_cannon_loop 2529 14.6 1.145 1.230 61.247 63.991 ot_mini 118 10.6 0.001 0.001 44.283 44.721 mp_waitall_1 172006 16.6 26.555 36.232 26.555 36.232 init_scf_loop 11 6.9 0.000 0.000 35.039 35.041 rebuild_ks_matrix 129 8.3 0.001 0.001 30.861 31.489 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.020 30.861 31.489 prepare_preconditioner 11 7.9 0.000 0.000 31.070 31.126 make_preconditioner 11 8.9 0.000 0.000 31.070 31.126 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.555 30.052 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.329 27.893 multiply_cannon_metrocomm3 20232 15.6 0.063 0.066 16.668 26.994 multiply_cannon_multrec 20232 15.6 13.062 16.230 22.476 25.648 qs_ot_get_derivative 118 11.6 0.002 0.002 22.974 23.420 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 21.317 22.436 apply_single 129 13.6 0.001 0.001 21.317 22.436 ot_diis_step 118 11.6 0.018 0.019 21.193 21.193 qs_ot_get_p 129 10.4 0.001 0.002 18.817 19.341 multiply_cannon_sync_h2d 20232 15.6 13.640 15.676 13.640 15.676 make_m2s 5058 13.6 0.079 0.083 14.846 15.551 make_images 5058 14.6 1.154 1.257 14.609 15.311 qs_ot_p2m_diag 84 11.4 0.268 0.275 14.564 14.573 cp_fm_cholesky_invert 11 10.9 14.251 14.260 14.251 14.260 sum_up_and_integrate 129 10.3 0.002 0.004 14.002 14.027 integrate_v_rspace 129 11.3 0.003 0.004 13.941 13.963 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.594 13.622 calculate_rho_elec 129 8.7 0.132 0.147 13.593 13.622 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.495 13.496 init_scf_run 11 5.9 0.000 0.001 11.517 11.517 scf_env_initial_rho_setup 11 6.9 0.023 0.024 11.517 11.517 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.294 10.317 cp_fm_diag_elpa_base 84 14.4 9.869 10.039 10.291 10.314 make_images_data 5058 15.6 0.065 0.076 8.898 9.989 hybrid_alltoall_any 5245 16.5 0.451 2.045 7.774 9.719 multiply_cannon_metrocomm4 17703 15.6 0.065 0.074 3.578 9.550 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.152 9.480 pw_transfer 1559 11.6 0.085 0.103 9.211 9.315 mp_irecv_dv 50659 16.2 3.448 9.292 3.448 9.292 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.987 9.097 dbcsr_mm_accdrv_process 41846 16.2 5.729 6.050 8.868 9.036 wfi_extrapolate 11 7.9 0.001 0.001 8.220 8.220 fft_wrap_pw1pw2_140 527 13.2 1.069 1.106 8.034 8.157 grid_integrate_task_list 129 12.3 7.356 7.737 7.356 7.737 cp_fm_upper_to_full 106 14.8 5.996 7.606 5.996 7.606 density_rs2pw 129 9.7 0.006 0.006 6.936 7.444 cp_fm_cholesky_decompose 22 10.9 7.247 7.277 7.247 7.277 dbcsr_complete_redistribute 397 12.7 1.168 1.195 4.712 6.559 fft3d_ps 1301 14.7 2.847 3.091 6.276 6.383 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.335 6.342 calculate_dm_sparse 129 9.5 0.001 0.001 5.931 6.021 grid_collocate_task_list 129 9.7 5.206 5.731 5.206 5.731 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.556 5.406 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.573 5.260 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.732 5.090 mp_alltoall_d11v 2429 14.1 4.494 4.996 4.494 4.996 mp_sum_l 8016 12.9 3.209 4.682 3.209 4.682 potential_pw2rs 129 12.3 0.021 0.023 4.387 4.394 transfer_fm_to_dbcsr 11 9.9 0.020 0.026 2.496 4.330 mp_alltoall_i22 720 14.1 2.304 4.320 2.304 4.320 mp_allgather_i34 2529 14.6 1.200 4.241 1.200 4.241 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 4.155 4.201 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.998 4.030 qs_energies_init_hamiltonians 11 5.9 0.009 0.019 3.780 3.782 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=179.631000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=908.090909, yerr=8.016512 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 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 1.157509E+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 665379241840 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 12503084. 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 0.179 0.221 189.498 189.498 qs_mol_dyn_low 1 2.0 0.003 0.003 188.472 188.485 qs_forces 11 3.9 0.004 0.005 188.361 188.368 qs_energies 11 4.9 0.002 0.002 181.252 181.265 scf_env_do_scf 11 5.9 0.001 0.001 164.494 164.505 velocity_verlet 10 3.0 0.001 0.002 124.855 124.858 scf_env_do_scf_inner_loop 117 6.6 0.011 0.029 117.982 117.983 qs_scf_new_mos 117 7.6 0.001 0.001 81.509 81.846 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.508 81.846 dbcsr_multiply_generic 2507 12.6 0.194 0.200 80.564 81.344 ot_scf_mini 117 9.6 0.003 0.004 76.921 77.284 multiply_cannon 2507 13.6 0.555 0.589 55.870 59.018 multiply_cannon_loop 2507 14.6 1.822 1.934 52.122 53.951 init_scf_loop 11 6.9 0.000 0.000 46.389 46.390 ot_mini 117 10.6 0.001 0.001 43.028 43.388 prepare_preconditioner 11 7.9 0.000 0.000 42.309 42.345 make_preconditioner 11 8.9 0.000 0.000 42.309 42.345 make_full_inverse_cholesky 11 9.9 0.012 0.023 35.840 40.861 multiply_cannon_multrec 30084 15.6 13.535 19.213 26.450 31.816 rebuild_ks_matrix 128 8.3 0.001 0.001 29.345 29.646 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 29.345 29.646 mp_waitall_1 147882 16.7 17.256 27.672 17.256 27.672 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.476 26.748 qs_ot_get_derivative 117 11.6 0.002 0.002 23.302 23.670 make_m2s 5014 13.6 0.092 0.095 20.317 21.483 make_images 5014 14.6 1.937 2.257 20.009 21.177 qs_ot_get_p 128 10.4 0.001 0.002 20.020 20.380 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.094 19.666 apply_single 128 13.6 0.001 0.001 19.094 19.665 ot_diis_step 117 11.6 0.018 0.019 19.606 19.608 cp_fm_upper_to_full 105 14.8 11.575 17.006 11.575 17.006 cp_fm_cholesky_invert 11 10.9 16.058 16.067 16.058 16.067 qs_ot_p2m_diag 83 11.4 0.342 0.390 15.822 15.876 multiply_cannon_metrocomm3 30084 15.6 0.049 0.052 6.461 14.931 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.414 14.415 sum_up_and_integrate 128 10.3 0.002 0.003 14.037 14.065 integrate_v_rspace 128 11.3 0.003 0.004 13.977 14.007 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.808 13.846 calculate_rho_elec 128 8.7 0.172 0.186 13.808 13.846 dbcsr_complete_redistribute 395 12.7 1.508 1.644 9.225 13.045 dbcsr_mm_accdrv_process 62264 16.2 8.504 9.207 12.484 12.931 make_images_data 5014 15.6 0.067 0.072 10.887 12.659 hybrid_alltoall_any 5200 16.5 0.531 2.203 9.816 11.955 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.848 11.690 multiply_cannon_sync_h2d 30084 15.6 10.445 11.336 10.445 11.336 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.155 11.171 cp_fm_diag_elpa_base 83 14.4 10.182 10.515 11.148 11.163 init_scf_run 11 5.9 0.000 0.001 10.875 10.876 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.874 10.876 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.446 10.187 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.799 10.054 mp_alltoall_i22 716 14.1 5.607 9.497 5.607 9.497 pw_transfer 1547 11.6 0.086 0.102 9.198 9.261 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.972 9.040 fft_wrap_pw1pw2_140 523 13.2 1.254 1.299 7.984 8.066 grid_integrate_task_list 128 12.3 7.531 7.890 7.531 7.890 wfi_extrapolate 11 7.9 0.001 0.001 7.727 7.727 cp_fm_cholesky_decompose 22 10.9 7.521 7.615 7.521 7.615 density_rs2pw 128 9.7 0.006 0.006 6.860 7.162 multiply_cannon_metrocomm4 25070 15.6 0.079 0.088 2.865 7.147 mp_irecv_dv 76098 16.2 2.713 6.873 2.713 6.873 calculate_dm_sparse 128 9.5 0.001 0.001 6.298 6.415 fft3d_ps 1291 14.7 2.989 3.045 6.014 6.059 grid_collocate_task_list 128 9.7 5.281 5.679 5.281 5.679 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.521 5.556 mp_alltoall_d11v 2415 14.1 5.006 5.308 5.006 5.308 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.472 4.589 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.420 4.500 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.394 4.396 potential_pw2rs 128 12.3 0.023 0.023 4.323 4.333 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.254 4.322 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.498000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1087.090909, yerr=20.882592 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.562509E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532237E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58346. MP_Allreduce 11057 1167. MP_Sync 87 MP_Alltoall 1712 18838186. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.047 172.194 172.195 qs_mol_dyn_low 1 2.0 0.003 0.003 171.683 171.705 qs_forces 11 3.9 0.004 0.004 171.519 171.523 qs_energies 11 4.9 0.001 0.002 164.071 164.079 scf_env_do_scf 11 5.9 0.001 0.001 146.266 146.280 velocity_verlet 10 3.0 0.031 0.037 112.347 112.354 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 110.468 110.470 qs_scf_new_mos 117 7.6 0.001 0.001 73.074 73.207 qs_scf_loop_do_ot 117 8.6 0.001 0.001 73.074 73.207 dbcsr_multiply_generic 2507 12.6 0.185 0.192 72.750 73.126 ot_scf_mini 117 9.6 0.003 0.004 68.654 68.783 multiply_cannon 2507 13.6 0.559 0.582 53.825 56.985 multiply_cannon_loop 2507 14.6 0.811 0.853 50.822 51.632 ot_mini 117 10.6 0.001 0.001 37.849 37.986 init_scf_loop 11 6.9 0.000 0.000 35.647 35.649 mp_waitall_1 125778 16.7 25.382 31.942 25.382 31.942 prepare_preconditioner 11 7.9 0.000 0.000 31.756 31.789 make_preconditioner 11 8.9 0.000 0.000 31.756 31.789 make_full_inverse_cholesky 11 9.9 0.017 0.027 29.669 29.929 rebuild_ks_matrix 128 8.3 0.001 0.001 29.442 29.575 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 29.442 29.575 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.842 26.967 multiply_cannon_multrec 10028 15.6 10.309 14.517 18.014 22.556 multiply_cannon_metrocomm3 10028 15.6 0.024 0.026 13.142 20.772 qs_ot_get_derivative 117 11.6 0.002 0.002 20.641 20.766 cp_fm_cholesky_invert 11 10.9 18.204 18.209 18.204 18.209 qs_ot_get_p 128 10.4 0.002 0.002 17.997 18.125 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 17.289 17.444 apply_single 128 13.6 0.001 0.001 17.289 17.444 ot_diis_step 117 11.6 0.020 0.020 17.137 17.138 make_m2s 5014 13.6 0.064 0.069 15.005 15.826 make_images 5014 14.6 2.159 2.588 14.698 15.512 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.352 14.403 calculate_rho_elec 128 8.7 0.255 0.266 14.351 14.403 sum_up_and_integrate 128 10.3 0.002 0.002 14.302 14.345 integrate_v_rspace 128 11.3 0.004 0.004 14.242 14.288 qs_ot_p2m_diag 83 11.4 0.495 0.501 14.172 14.189 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.996 12.996 multiply_cannon_sync_h2d 10028 15.6 10.702 11.241 10.702 11.241 init_scf_run 11 5.9 0.000 0.001 10.946 10.946 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.946 10.946 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.072 10.086 cp_fm_diag_elpa_base 83 14.4 9.832 9.896 10.068 10.081 pw_transfer 1547 11.6 0.086 0.096 9.981 10.014 hybrid_alltoall_any 5200 16.5 0.841 3.775 8.506 9.937 make_images_data 5014 15.6 0.056 0.066 8.706 9.935 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 9.758 9.799 fft_wrap_pw1pw2_140 523 13.2 1.677 1.719 8.620 8.667 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.109 8.210 grid_integrate_task_list 128 12.3 7.789 8.151 7.789 8.151 dbcsr_mm_accdrv_process 20762 16.1 3.262 4.453 7.354 8.065 cp_fm_cholesky_decompose 22 10.9 7.870 7.994 7.870 7.994 multiply_cannon_metrocomm1 10028 15.6 0.030 0.031 4.260 7.528 wfi_extrapolate 11 7.9 0.001 0.001 7.507 7.507 density_rs2pw 128 9.7 0.005 0.005 7.025 7.301 fft3d_ps 1291 14.7 3.141 3.238 6.204 6.235 calculate_dm_sparse 128 9.5 0.001 0.001 6.148 6.226 grid_collocate_task_list 128 9.7 5.573 5.835 5.573 5.835 dbcsr_complete_redistribute 395 12.7 2.124 2.182 5.298 5.683 multiply_cannon_metrocomm4 7521 15.6 0.026 0.029 1.860 5.643 mp_irecv_dv 28860 15.9 1.820 5.553 1.820 5.553 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.435 5.479 qs_energies_init_hamiltonians 11 5.9 0.012 0.022 5.312 5.313 mp_alltoall_d11v 2415 14.1 4.706 5.128 4.706 5.128 mp_allgather_i34 2507 14.6 1.179 4.712 1.179 4.712 potential_pw2rs 128 12.3 0.027 0.028 4.430 4.438 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.249 4.311 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.581 3.882 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.432 3.744 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.636 3.721 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.647 3.685 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.476 3.632 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.505 3.522 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=172.195000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1472.272727, yerr=33.387779 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_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 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 11.786061E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 3.173712E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+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 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4045 58530. MP_Allreduce 11189 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 0.034 0.095 297.711 297.712 qs_mol_dyn_low 1 2.0 0.003 0.003 297.080 297.108 qs_forces 11 3.9 0.004 0.005 296.971 296.975 qs_energies 11 4.9 0.002 0.002 288.013 288.021 scf_env_do_scf 11 5.9 0.001 0.001 264.935 264.942 velocity_verlet 10 3.0 0.020 0.022 214.453 214.461 scf_env_do_scf_inner_loop 118 6.6 0.035 0.039 138.385 138.388 init_scf_loop 11 6.9 0.000 0.000 126.237 126.243 prepare_preconditioner 11 7.9 0.000 0.000 120.813 120.848 make_preconditioner 11 8.9 0.000 0.000 120.813 120.848 make_full_inverse_cholesky 11 9.9 0.038 0.039 96.241 117.955 qs_scf_new_mos 118 7.6 0.001 0.001 92.879 93.009 qs_scf_loop_do_ot 118 8.6 0.001 0.001 92.878 93.008 ot_scf_mini 118 9.6 0.004 0.004 87.863 87.944 dbcsr_multiply_generic 2529 12.6 0.219 0.230 84.330 85.033 cp_fm_upper_to_full 106 14.8 54.071 77.746 54.071 77.746 multiply_cannon 2529 13.6 0.687 0.730 60.445 61.632 multiply_cannon_loop 2529 14.6 1.063 1.078 56.135 58.050 ot_mini 118 10.6 0.001 0.001 45.751 45.851 dbcsr_complete_redistribute 397 12.7 4.003 4.077 30.632 43.809 copy_fm_to_dbcsr 210 11.7 0.001 0.002 27.097 40.293 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.529 37.534 mp_alltoall_i22 720 14.1 22.344 35.723 22.344 35.723 rebuild_ks_matrix 129 8.3 0.001 0.001 35.268 35.395 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.019 35.267 35.395 mp_waitall_1 104580 16.8 29.294 34.567 29.294 34.567 cp_fm_cholesky_invert 11 10.9 33.325 33.331 33.325 33.331 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.756 32.858 qs_ot_get_p 129 10.4 0.443 0.452 27.169 27.265 qs_ot_get_derivative 118 11.6 0.002 0.002 25.263 25.345 qs_ot_p2m_diag 84 11.4 0.890 0.896 22.414 22.448 multiply_cannon_metrocomm3 10116 15.6 0.025 0.026 19.728 21.699 make_m2s 5058 13.6 0.075 0.077 19.021 20.700 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.580 20.581 ot_diis_step 118 11.6 0.022 0.023 20.441 20.442 make_images 5058 14.6 3.085 3.293 18.534 20.213 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.598 19.912 apply_single 129 13.6 0.001 0.001 19.598 19.911 multiply_cannon_multrec 10116 15.6 10.387 12.275 18.299 18.346 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.478 17.487 calculate_rho_elec 129 8.7 0.482 0.483 17.478 17.486 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.336 17.337 cp_fm_diag_elpa_base 84 14.4 12.759 14.467 17.332 17.332 sum_up_and_integrate 129 10.3 0.002 0.002 16.559 16.651 integrate_v_rspace 129 11.3 0.005 0.005 16.496 16.587 multiply_cannon_sync_h2d 10116 15.6 14.406 14.429 14.406 14.429 hybrid_alltoall_any 5245 16.5 1.325 3.094 11.118 13.559 pw_transfer 1559 11.6 0.096 0.097 13.362 13.369 make_images_data 5058 15.6 0.065 0.071 10.839 13.203 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 13.121 13.127 init_scf_run 11 5.9 0.000 0.001 12.561 12.563 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.561 12.563 fft_wrap_pw1pw2_140 527 13.2 3.257 3.366 11.682 11.699 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 10.056 10.113 dbcsr_mm_accdrv_process 20934 16.1 4.448 6.273 7.657 9.512 wfi_extrapolate 11 7.9 0.001 0.001 9.157 9.157 cp_fm_cholesky_decompose 22 10.9 8.824 8.843 8.824 8.843 grid_integrate_task_list 129 12.3 8.614 8.819 8.614 8.819 density_rs2pw 129 9.7 0.005 0.005 8.440 8.490 qs_energies_init_hamiltonians 11 5.9 0.019 0.033 8.268 8.269 fft3d_ps 1301 14.7 4.005 4.029 7.529 7.582 mp_alltoall_d11v 2429 14.1 6.990 7.176 6.990 7.176 calculate_dm_sparse 129 9.5 0.001 0.001 6.883 6.953 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.630 6.681 grid_collocate_task_list 129 9.7 6.462 6.547 6.462 6.547 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.302 6.390 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=297.712000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2882.909091, yerr=158.867271 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_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 1.261208E+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 8123016. 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 0.038 0.083 85.731 85.746 qs_energies 1 2.0 0.000 0.000 85.039 85.050 ls_scf 1 3.0 0.002 0.017 83.834 83.845 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.560 72.731 multiply_cannon 111 7.7 0.017 0.020 55.870 57.109 multiply_cannon_loop 111 8.7 0.231 0.243 52.507 53.792 ls_scf_main 1 4.0 0.003 0.023 52.371 52.398 density_matrix_trs4 2 5.0 0.002 0.003 46.778 46.868 ls_scf_init_scf 1 4.0 0.000 0.001 28.392 28.409 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.080 27.133 mp_waitall_1 11031 10.9 22.369 26.226 22.369 26.226 matrix_sqrt_Newton_Schulz 2 6.5 0.004 0.023 24.986 25.001 multiply_cannon_multrec 2664 9.7 8.174 8.975 15.585 17.373 multiply_cannon_sync_h2d 2664 9.7 13.598 15.721 13.598 15.721 make_m2s 222 7.7 0.010 0.012 13.057 13.499 make_images 222 8.7 0.100 0.110 13.035 13.477 multiply_cannon_metrocomm1 2664 9.7 0.011 0.012 9.637 12.398 make_images_data 222 9.7 0.005 0.006 7.625 8.129 dbcsr_mm_accdrv_process 4760 10.4 0.591 0.699 7.027 8.019 hybrid_alltoall_any 227 10.6 0.218 1.848 6.574 7.886 multiply_cannon_metrocomm3 2664 9.7 0.010 0.012 5.440 7.808 dbcsr_mm_accdrv_process_sort 4760 11.4 6.234 7.124 6.234 7.124 calculate_norms 4752 9.8 5.523 6.179 5.523 6.179 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.049 5.238 mp_sum_l 887 5.1 3.192 4.603 3.192 4.603 make_images_sizes 222 9.7 0.000 0.000 0.675 3.743 mp_alltoall_i44 222 10.7 0.675 3.743 0.675 3.743 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.317 3.327 arnoldi_extremal 4 6.8 0.000 0.000 3.251 3.281 arnoldi_normal_ev 4 7.8 0.001 0.003 3.251 3.281 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.067 3.128 build_subspace 16 8.4 0.009 0.012 3.123 3.125 mp_irecv_dv 6231 10.9 2.049 3.101 2.049 3.101 ls_scf_post 1 4.0 0.001 0.010 3.070 3.082 ls_scf_store_result 1 5.0 0.000 0.000 2.880 2.917 dbcsr_special_finalize 555 9.7 0.006 0.007 2.395 2.729 dbcsr_merge_single_wm 555 10.7 0.454 0.573 2.386 2.721 make_images_pack 222 9.7 2.213 2.631 2.215 2.633 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.322 2.564 dbcsr_sort_data 658 11.4 2.191 2.504 2.191 2.504 dbcsr_matrix_vector_mult_local 304 10.0 2.065 2.455 2.067 2.457 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.316 2.404 buffer_matrices_ensure_size 222 8.7 1.746 2.061 1.746 2.061 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.968 1.985 rebuild_ks_matrix 3 7.3 0.000 0.000 1.959 1.976 qs_ks_build_kohn_sham_matrix 3 8.3 0.035 0.072 1.959 1.976 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.746000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_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 2.196050E+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 0.057 0.070 92.187 92.188 qs_energies 1 2.0 0.005 0.040 91.547 91.660 ls_scf 1 3.0 0.000 0.000 90.180 90.262 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.640 75.947 multiply_cannon 111 7.7 0.027 0.042 53.138 58.033 ls_scf_main 1 4.0 0.001 0.029 55.063 55.070 multiply_cannon_loop 111 8.7 0.136 0.146 49.965 53.963 density_matrix_trs4 2 5.0 0.002 0.006 49.331 49.615 mp_waitall_1 9105 10.9 21.095 31.042 21.095 31.042 ls_scf_init_scf 1 4.0 0.000 0.001 30.007 30.012 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.807 28.906 multiply_cannon_multrec 1332 9.7 13.402 17.541 22.697 27.891 matrix_sqrt_Newton_Schulz 2 6.5 0.008 0.018 26.375 26.384 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.431 21.495 make_m2s 222 7.7 0.007 0.008 15.668 16.634 make_images 222 8.7 1.374 1.699 15.638 16.605 dbcsr_mm_accdrv_process 4041 10.4 0.364 0.541 8.890 10.497 make_images_data 222 9.7 0.004 0.005 9.158 10.384 hybrid_alltoall_any 227 10.6 0.541 2.565 8.381 9.958 dbcsr_mm_accdrv_process_sort 4041 11.4 8.411 9.956 8.411 9.956 mp_sum_l 887 5.1 5.365 8.626 5.365 8.626 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.196 7.965 mp_irecv_dv 3311 11.0 3.176 7.917 3.176 7.917 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.083 7.004 calculate_norms 2376 9.8 6.075 6.651 6.075 6.651 multiply_cannon_sync_h2d 1332 9.7 4.923 6.412 4.923 6.412 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.036 5.255 ls_scf_post 1 4.0 0.009 0.068 5.110 5.192 arnoldi_extremal 4 6.8 0.000 0.002 4.681 4.701 arnoldi_normal_ev 4 7.8 0.002 0.007 4.681 4.700 build_subspace 16 8.4 0.014 0.021 4.406 4.415 ls_scf_store_result 1 5.0 0.000 0.000 4.173 4.299 mp_allgather_i34 111 8.7 1.156 3.583 1.156 3.583 dbcsr_matrix_vector_mult 304 9.0 0.010 0.023 3.175 3.423 dbcsr_matrix_vector_mult_local 304 10.0 2.767 3.252 2.769 3.254 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.152 2.715 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.592 2.683 dbcsr_data_new 4174 10.1 2.105 2.393 2.105 2.393 make_images_pack 222 9.7 1.830 2.134 1.833 2.137 dbcsr_sort_data 436 11.2 1.823 2.117 1.823 2.117 make_images_sizes 222 9.7 0.000 0.000 0.739 2.099 mp_alltoall_i44 222 10.7 0.738 2.099 0.738 2.099 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.849 1.855 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.188000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1807.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_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 2.888360E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.076 96.060 96.067 qs_energies 1 2.0 0.000 0.001 95.092 95.106 ls_scf 1 3.0 0.004 0.033 93.655 93.668 dbcsr_multiply_generic 111 6.7 0.016 0.016 77.346 77.784 multiply_cannon 111 7.7 0.032 0.075 53.755 58.378 ls_scf_main 1 4.0 0.000 0.004 58.127 58.132 multiply_cannon_loop 111 8.7 0.117 0.130 50.057 54.559 density_matrix_trs4 2 5.0 0.002 0.009 50.746 50.948 mp_waitall_1 7281 11.0 24.449 34.888 24.449 34.888 ls_scf_init_scf 1 4.0 0.000 0.002 30.550 30.553 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.324 29.430 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.904 26.915 multiply_cannon_multrec 888 9.7 12.765 15.406 21.354 24.607 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 10.990 23.591 make_m2s 222 7.7 0.006 0.008 16.634 18.101 make_images 222 8.7 1.586 1.899 16.596 18.062 make_images_data 222 9.7 0.004 0.005 9.775 11.648 hybrid_alltoall_any 227 10.6 0.641 2.944 9.091 11.283 dbcsr_mm_accdrv_process 3754 10.4 0.297 0.491 8.112 9.361 mp_sum_l 887 5.1 5.310 9.203 5.310 9.203 dbcsr_mm_accdrv_process_sort 3754 11.4 7.676 8.870 7.676 8.870 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 4.329 7.611 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.075 7.296 multiply_cannon_sync_h2d 888 9.7 6.085 7.193 6.085 7.193 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.464 7.186 mp_irecv_dv 2335 11.1 2.449 7.133 2.449 7.133 apply_matrix_preconditioner 6 5.3 0.000 0.000 6.270 6.496 arnoldi_extremal 4 6.8 0.000 0.000 5.213 5.231 arnoldi_normal_ev 4 7.8 0.001 0.006 5.213 5.231 ls_scf_post 1 4.0 0.000 0.003 4.973 4.981 mp_allgather_i34 111 8.7 1.806 4.961 1.806 4.961 build_subspace 16 8.4 0.014 0.020 4.903 4.911 ls_scf_store_result 1 5.0 0.000 0.000 4.707 4.787 calculate_norms 1584 9.8 4.345 4.719 4.345 4.719 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.517 3.842 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.603 3.710 dbcsr_matrix_vector_mult_local 304 10.0 3.048 3.643 3.050 3.646 dbcsr_data_new 4116 9.9 2.099 2.457 2.099 2.457 make_images_sizes 222 9.7 0.000 0.000 1.026 2.178 mp_alltoall_i44 222 10.7 1.026 2.177 1.026 2.177 qs_rho_update_rho_low 3 6.3 0.000 0.000 2.138 2.175 calculate_rho_elec 3 7.3 0.011 0.011 2.138 2.175 dbcsr_sort_data 325 11.1 1.864 2.137 1.864 2.137 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=96.067000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2242.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_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 3.336421E+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 0.112 0.167 100.630 100.632 qs_energies 1 2.0 0.002 0.017 99.523 99.537 ls_scf 1 3.0 0.000 0.001 97.833 97.840 dbcsr_multiply_generic 111 6.7 0.017 0.017 81.436 81.745 ls_scf_main 1 4.0 0.000 0.002 61.641 61.643 multiply_cannon 111 7.7 0.044 0.087 54.072 58.771 density_matrix_trs4 2 5.0 0.002 0.004 55.512 55.655 multiply_cannon_loop 111 8.7 0.153 0.165 48.863 52.117 ls_scf_init_scf 1 4.0 0.000 0.003 32.875 32.878 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.650 31.718 mp_waitall_1 6369 11.0 25.245 31.013 25.245 31.013 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.151 29.166 multiply_cannon_multrec 1332 9.7 14.151 17.406 22.075 24.513 make_m2s 222 7.7 0.008 0.009 21.498 22.816 make_images 222 8.7 3.143 3.606 21.447 22.767 multiply_cannon_metrocomm3 1332 9.7 0.004 0.004 10.081 19.048 make_images_data 222 9.7 0.004 0.005 12.120 13.912 hybrid_alltoall_any 227 10.6 0.799 3.784 11.322 13.169 dbcsr_mm_accdrv_process 3641 10.4 0.286 0.480 7.562 9.069 dbcsr_mm_accdrv_process_sort 3641 11.4 7.122 8.573 7.122 8.573 mp_sum_l 887 5.1 4.320 8.220 4.320 8.220 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 3.876 6.607 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.349 6.521 multiply_cannon_sync_h2d 1332 9.7 5.461 6.078 5.461 6.078 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.108 6.052 mp_irecv_dv 3229 10.9 2.089 5.971 2.089 5.971 arnoldi_extremal 4 6.8 0.000 0.000 5.350 5.363 arnoldi_normal_ev 4 7.8 0.001 0.005 5.350 5.363 build_subspace 16 8.4 0.014 0.021 5.011 5.019 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.563 4.746 mp_allgather_i34 111 8.7 2.268 4.509 2.268 4.509 calculate_norms 2376 9.8 4.164 4.504 4.164 4.504 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 3.682 3.973 dbcsr_matrix_vector_mult_local 304 10.0 3.224 3.740 3.226 3.742 dbcsr_sort_data 658 11.4 3.055 3.356 3.055 3.356 ls_scf_post 1 4.0 0.000 0.001 3.317 3.322 dbcsr_special_finalize 555 9.7 0.006 0.007 2.817 3.125 dbcsr_merge_single_wm 555 10.7 0.539 0.651 2.809 3.117 ls_scf_store_result 1 5.0 0.000 0.000 3.007 3.087 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.012 3.067 dbcsr_data_release 10477 10.7 1.585 2.392 1.585 2.392 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=100.632000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2713.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_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 4.735255E+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 0.035 0.053 94.072 94.074 qs_energies 1 2.0 0.000 0.000 93.295 93.300 ls_scf 1 3.0 0.000 0.000 91.272 91.277 dbcsr_multiply_generic 111 6.7 0.017 0.018 71.575 71.774 ls_scf_main 1 4.0 0.000 0.000 57.248 57.249 multiply_cannon 111 7.7 0.072 0.124 53.556 57.057 multiply_cannon_loop 111 8.7 0.088 0.094 51.012 52.168 density_matrix_trs4 2 5.0 0.002 0.003 49.185 49.234 ls_scf_init_scf 1 4.0 0.000 0.001 29.027 29.028 mp_waitall_1 5436 11.0 25.124 28.868 25.124 28.868 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.787 27.809 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.692 25.703 multiply_cannon_multrec 444 9.7 13.673 16.460 20.744 22.641 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.916 16.736 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.160 14.445 make_m2s 222 7.7 0.005 0.005 13.022 13.932 make_images 222 8.7 2.042 2.486 12.955 13.862 make_images_data 222 9.7 0.003 0.004 7.867 9.245 hybrid_alltoall_any 227 10.6 0.805 3.849 7.744 9.229 multiply_cannon_sync_h2d 444 9.7 6.698 8.944 6.698 8.944 dbcsr_mm_accdrv_process 3003 10.4 0.338 0.393 6.768 7.864 dbcsr_mm_accdrv_process_sort 3003 11.4 6.407 7.471 6.407 7.471 arnoldi_extremal 4 6.8 0.000 0.000 5.989 6.002 arnoldi_normal_ev 4 7.8 0.002 0.005 5.989 6.002 build_subspace 16 8.4 0.015 0.019 5.597 5.605 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.811 5.332 ls_scf_post 1 4.0 0.000 0.000 4.998 5.002 ls_scf_store_result 1 5.0 0.000 0.000 4.788 4.825 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.652 4.703 mp_sum_l 887 5.1 2.994 4.651 2.994 4.651 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.245 4.439 dbcsr_matrix_vector_mult_local 304 10.0 3.712 4.189 3.714 4.191 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.536 3.982 mp_irecv_dv 1241 11.2 1.524 3.955 1.524 3.955 calculate_norms 792 9.8 3.624 3.770 3.624 3.770 mp_allgather_i34 111 8.7 1.118 3.489 1.118 3.489 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.820 3.430 make_images_sizes 222 9.7 0.000 0.000 0.761 2.863 mp_alltoall_i44 222 10.7 0.761 2.862 0.761 2.862 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.632 2.633 rebuild_ks_matrix 3 7.3 0.000 0.000 2.599 2.600 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.002 2.599 2.600 mp_alltoall_d11v 48 9.2 2.410 2.521 2.410 2.521 dbcsr_finalize 304 7.8 0.062 0.077 2.200 2.279 dbcsr_data_new 4608 9.7 1.792 2.234 1.792 2.234 qs_rho_update_rho_low 3 6.3 0.000 0.000 2.054 2.126 calculate_rho_elec 3 7.3 0.020 0.022 2.054 2.126 dbcsr_merge_all 275 8.9 0.479 0.531 2.058 2.119 sum_up_and_integrate 3 9.3 0.001 0.001 2.113 2.114 integrate_v_rspace 3 10.3 0.000 0.000 2.105 2.106 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.007 2.007 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=94.074000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3697.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/26/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 30.358840E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 339931 0.0% 0.0% 100.0% average stack size 0.0 0.0 28341.7 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 8.856408E+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 0.089 0.243 109.139 109.141 qs_energies 1 2.0 0.000 0.000 107.136 107.140 ls_scf 1 3.0 0.000 0.000 104.133 104.137 dbcsr_multiply_generic 111 6.7 0.024 0.028 77.190 77.357 ls_scf_main 1 4.0 0.000 0.000 66.636 66.641 density_matrix_trs4 2 5.0 0.002 0.003 57.440 57.504 multiply_cannon 111 7.7 0.148 0.283 48.131 50.655 multiply_cannon_loop 111 8.7 0.098 0.099 45.074 45.396 ls_scf_init_scf 1 4.0 0.001 0.001 33.688 33.695 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.120 32.136 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.003 29.349 29.357 mp_waitall_1 4527 11.1 23.936 27.680 23.936 27.680 make_m2s 222 7.7 0.005 0.006 25.324 26.477 make_images 222 8.7 3.577 3.889 25.214 26.365 multiply_cannon_multrec 444 9.7 17.919 18.695 22.555 23.355 hybrid_alltoall_any 227 10.6 1.657 3.630 14.601 18.371 make_images_data 222 9.7 0.004 0.004 15.540 17.879 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.782 10.268 multiply_cannon_sync_h2d 444 9.7 8.784 8.852 8.784 8.852 arnoldi_extremal 4 6.8 0.000 0.000 7.547 7.564 arnoldi_normal_ev 4 7.8 0.005 0.011 7.547 7.564 build_subspace 16 8.4 0.026 0.036 6.985 6.994 dbcsr_matrix_vector_mult 304 9.0 0.017 0.035 5.579 5.741 dbcsr_matrix_vector_mult_local 304 10.0 5.099 5.412 5.102 5.415 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.293 5.382 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.877 5.122 make_images_sizes 222 9.7 0.000 0.000 1.616 4.649 mp_alltoall_i44 222 10.7 1.616 4.649 1.616 4.649 dbcsr_mm_accdrv_process 1814 10.4 0.289 0.355 4.461 4.590 dbcsr_mm_accdrv_process_sort 1814 11.4 4.125 4.258 4.125 4.258 ls_scf_post 1 4.0 0.000 0.000 3.808 3.813 mp_allgather_i34 111 8.7 1.157 3.618 1.157 3.618 ls_scf_store_result 1 5.0 0.000 0.000 3.525 3.565 calculate_norms 792 9.8 3.236 3.275 3.236 3.275 dbcsr_finalize 304 7.8 0.082 0.089 3.072 3.178 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.973 2.973 dbcsr_merge_all 275 8.9 0.885 0.918 2.857 2.953 dbcsr_complete_redistribute 5 7.6 1.417 1.459 2.759 2.879 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.576 2.584 rebuild_ks_matrix 3 7.3 0.000 0.000 2.510 2.518 qs_ks_build_kohn_sham_matrix 3 8.3 0.006 0.013 2.510 2.518 matrix_ls_to_qs 2 6.0 0.000 0.000 2.397 2.517 dbcsr_sort_data 325 11.1 2.438 2.503 2.438 2.503 dbcsr_data_new 6591 9.6 1.881 2.439 1.881 2.439 dbcsr_new_transposed 4 7.5 0.241 0.252 2.361 2.394 dbcsr_frobenius_norm 74 6.6 2.056 2.131 2.178 2.209 dbcsr_data_release 12724 10.6 1.974 2.208 1.974 2.208 dbcsr_add_d 103 6.2 0.000 0.000 2.126 2.206 dbcsr_add_anytype 103 7.2 0.858 0.890 2.125 2.206 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.141000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7036.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 593.702912E+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 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 0.036 0.096 240.655 240.675 qs_mol_dyn_low 1 2.0 0.003 0.011 239.477 239.575 qs_forces 5 3.8 0.007 0.034 238.876 238.878 qs_energies 5 4.8 0.006 0.085 235.701 235.748 scf_env_do_scf 5 5.8 0.005 0.128 220.251 220.255 scf_env_do_scf_inner_loop 105 6.6 0.006 0.067 191.662 191.671 qs_scf_new_mos 105 7.6 0.000 0.001 148.195 148.805 qs_scf_loop_do_ot 105 8.6 0.001 0.008 148.195 148.804 dbcsr_multiply_generic 1445 12.2 0.127 0.135 139.379 140.571 ot_scf_mini 105 9.6 0.004 0.020 137.936 138.705 multiply_cannon 1445 13.2 0.275 0.286 118.030 121.451 multiply_cannon_loop 1445 14.2 2.839 3.017 116.252 120.048 velocity_verlet 4 3.0 0.002 0.010 112.093 112.095 ot_mini 105 10.6 0.001 0.014 62.176 62.917 mp_waitall_1 488190 16.1 39.392 47.956 39.392 47.956 qs_ot_get_p 112 10.4 0.004 0.114 45.054 45.609 multiply_cannon_multrec 69360 15.2 29.709 34.242 39.842 45.005 qs_ot_get_derivative 55 11.6 0.001 0.001 39.790 40.560 multiply_cannon_metrocomm3 69360 15.2 0.200 0.211 27.999 38.404 multiply_cannon_sync_h2d 69360 15.2 29.116 33.117 29.116 33.117 qs_ot_p2m_diag 40 11.0 0.026 0.150 32.536 32.859 rebuild_ks_matrix 110 8.4 0.000 0.000 30.971 31.677 qs_ks_build_kohn_sham_matrix 110 9.4 0.016 0.085 30.971 31.677 qs_ks_update_qs_env 112 7.6 0.001 0.001 28.478 29.163 cp_dbcsr_syevd 40 12.0 0.003 0.054 28.664 28.675 init_scf_loop 7 6.6 0.002 0.041 28.552 28.559 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.462 23.949 apply_single 62 13.6 0.000 0.000 23.462 23.948 cp_fm_syevd 40 13.0 0.000 0.002 23.152 23.288 prepare_preconditioner 7 7.6 0.000 0.001 23.079 23.112 make_preconditioner 7 8.6 0.001 0.023 23.079 23.112 ot_new_cg_direction 55 11.6 0.001 0.001 21.705 21.705 qs_rho_update_rho_low 110 7.6 0.000 0.001 18.755 19.314 calculate_rho_elec 110 8.6 0.030 0.032 18.754 19.314 cp_fm_redistribute_end 40 14.0 9.127 18.205 9.132 18.207 cp_fm_syevd_base 40 14.0 9.064 18.144 9.064 18.144 mp_sum_l 4764 12.2 13.914 15.912 13.914 15.912 make_full_inverse_cholesky 7 9.6 0.000 0.005 15.483 15.550 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.136 15.432 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.402 15.151 density_rs2pw 110 9.6 0.005 0.006 12.712 13.158 init_scf_run 5 5.8 0.000 0.001 12.575 12.576 scf_env_initial_rho_setup 5 6.8 0.002 0.004 12.575 12.576 calculate_dm_sparse 110 9.5 0.000 0.001 12.086 12.491 pw_transfer 1645 12.4 0.080 0.104 11.854 12.224 multiply_cannon_metrocomm1 69360 15.2 0.093 0.100 6.067 12.211 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.715 12.088 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 11.244 11.326 dbcsr_mm_accdrv_process 154766 15.8 6.221 6.397 10.001 10.965 cp_fm_cholesky_invert 7 10.6 10.833 10.842 10.833 10.842 fft_wrap_pw1pw2_240 915 15.0 0.837 0.926 10.299 10.642 qs_vxc_create 110 10.4 0.003 0.034 10.484 10.532 transfer_rs2pw 445 10.6 0.007 0.008 9.475 9.932 check_diag 80 13.5 8.685 8.935 9.745 9.878 sum_up_and_integrate 60 10.3 0.002 0.042 9.428 9.465 integrate_v_rspace 60 11.3 0.002 0.008 9.410 9.447 fft3d_pb 915 16.0 2.398 2.701 8.472 8.839 acc_transpose_blocks 69360 15.2 0.359 0.383 7.293 7.795 calculate_first_density_matrix 1 7.0 0.001 0.037 7.561 7.586 xc_rho_set_and_dset_create 110 12.4 0.077 0.149 7.317 7.583 make_m2s 2890 13.2 0.079 0.088 6.954 7.578 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.012 7.484 7.499 make_images 2890 14.2 0.241 0.261 6.848 7.474 make_full_single_inverse 7 9.6 0.003 0.081 7.311 7.345 xc_vxc_pw_create 60 11.3 0.039 0.076 7.064 7.112 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.987 7.059 xc_pw_derive 510 13.4 0.005 0.007 6.256 6.335 mp_alltoall_z22v 2340 17.7 5.489 5.957 5.489 5.957 potential_pw2rs 60 12.3 0.002 0.003 5.581 5.646 mp_waitany 7680 13.5 4.511 5.127 4.511 5.127 acc_transpose_blocks_kernels 69360 16.2 0.840 0.885 4.537 4.922 multiply_cannon_metrocomm4 67915 15.2 0.184 0.202 2.007 4.829 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=240.675000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=563.200000, yerr=3.429286 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 9dc68573fd84d3eeaa2b8e8af0ad793de4465ca0 Summary: empty Status: OK