=== 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: 8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf ################# 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/01 job id: 50906023 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/02 job id: 50906024 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/03 job id: 50906025 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/04 job id: 50906026 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/05 job id: 50906027 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/06 job id: 50906028 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/07 job id: 50906029 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/08 job id: 50906030 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/09 job id: 50906031 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/10 job id: 50906032 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/11 job id: 50906033 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/12 job id: 50906034 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/13 job id: 50906035 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/14 job id: 50906036 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/15 job id: 50906037 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/16 job id: 50906038 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/17 job id: 50906039 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/18 job id: 50906040 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/19 job id: 50906041 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/20 job id: 50906042 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/21 job id: 50906043 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/22 job id: 50906044 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/23 job id: 50906045 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/24 job id: 50906046 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/25 job id: 50906047 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/26 job id: 50906048 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/27 job id: 50906049 --- 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/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 172669. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.033 134.927 134.930 farming_run 1 2.0 134.411 134.414 134.894 134.899 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.505260E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.005 0.026 115.698 115.701 qs_energies 1 2.0 0.000 0.000 115.496 115.500 mp2_main 1 3.0 0.000 0.000 113.251 113.254 mp2_gpw_main 1 4.0 0.027 0.033 112.244 112.247 mp2_ri_gpw_compute_in 1 5.0 0.172 0.175 93.084 93.482 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.710 55.109 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.164 41.120 46.193 get_2c_integrals 1 6.0 0.008 0.009 37.304 38.200 integrate_v_rspace 273 8.0 0.441 0.458 24.692 29.484 pw_transfer 6555 10.6 0.378 0.389 26.932 27.745 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.567 26.302 grid_integrate_task_list 273 9.0 20.571 25.838 20.571 25.838 fft_wrap_pw1pw2_100 2178 12.4 0.030 0.032 23.102 23.810 compute_2c_integrals 1 7.0 0.002 0.002 19.389 19.391 rpa_ri_compute_en 1 5.0 0.019 0.025 19.057 19.331 compute_2c_integrals_loop_lm 1 8.0 0.001 0.003 18.640 19.110 mp2_eri_2c_integrate_gpw 1 9.0 2.380 2.426 18.638 19.109 cp_fm_cholesky_decompose 12 8.2 17.858 18.748 17.858 18.748 cholesky_decomp 1 7.0 0.000 0.000 16.763 17.663 fft3d_s 5443 13.4 16.137 16.450 16.160 16.471 ao_to_mo_and_store_B_mult_1 272 7.0 10.761 15.311 10.761 15.311 calculate_wavefunction 272 8.0 5.395 5.546 12.339 12.948 rpa_num_int 1 6.0 0.000 0.001 10.604 10.614 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.583 10.592 calc_mat_Q 8 8.0 0.000 0.000 9.462 9.546 contract_S_to_Q 8 9.0 0.000 0.000 8.886 8.969 calc_potential_gpw 544 9.5 0.004 0.005 8.245 8.629 parallel_gemm_fm 14 9.1 0.000 0.000 8.470 8.565 parallel_gemm_fm_cosma 14 10.1 8.470 8.565 8.470 8.565 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.096 8.425 potential_pw2rs 545 10.0 0.106 0.108 7.524 8.282 create_integ_mat 1 6.0 0.015 0.028 7.960 7.970 collocate_single_gaussian 272 10.0 0.039 0.041 7.357 7.600 array2fm 1 7.0 0.000 0.000 6.693 7.356 pw_scatter_s 2720 13.7 4.360 4.505 4.360 4.505 pw_gather_s 2722 13.2 3.478 3.864 3.478 3.864 array2fm_buffer_send 1 8.0 2.961 3.164 2.961 3.164 pw_poisson_solve 545 10.5 1.124 1.177 2.195 2.330 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.245402, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2806.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_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 0.026 0.037 423.130 423.133 farming_run 1 2.0 422.461 422.465 423.085 423.088 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.240003E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.051 217.893 217.894 qs_energies 1 2.0 0.000 0.000 217.625 217.638 scf_env_do_scf 1 3.0 0.000 0.000 115.554 115.555 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.401 114.408 rebuild_ks_matrix 4 6.0 0.000 0.000 114.399 114.407 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.064 114.399 114.407 hfx_ks_matrix 4 8.0 0.001 0.001 114.002 114.007 integrate_four_center 4 9.0 0.153 0.469 114.001 114.006 integrate_four_center_main 4 10.0 0.119 0.547 101.908 105.875 integrate_four_center_bin 265 11.0 101.788 105.860 101.788 105.860 mp2_main 1 3.0 0.000 0.000 101.778 101.790 mp2_gpw_main 1 4.0 0.046 0.073 100.664 100.676 init_scf_loop 1 4.0 0.000 0.000 97.605 97.605 mp2_ri_gpw_compute_in 1 5.0 0.065 0.068 73.904 74.946 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.796 54.837 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.160 41.637 46.557 integrate_v_rspace 95 8.0 0.398 0.567 28.155 32.897 pw_transfer 2240 10.6 0.146 0.162 29.503 29.908 ao_to_mo_and_store_B_mult_1 91 7.0 10.486 29.433 10.486 29.433 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.516 28.905 grid_integrate_task_list 95 9.0 23.526 28.484 23.526 28.484 mp2_ri_gpw_compute_en 1 5.0 0.061 0.075 26.611 28.175 fft_wrap_pw1pw2_100 730 12.4 0.013 0.013 26.236 26.652 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.849 1.920 24.913 24.923 get_2c_integrals 1 6.0 0.000 0.000 20.023 20.043 compute_2c_integrals 1 7.0 0.002 0.003 18.996 19.004 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.598 18.872 mp2_eri_2c_integrate_gpw 1 9.0 1.742 1.964 18.597 18.871 fft3d_s 1823 13.4 18.465 18.816 18.479 18.829 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.947 17.947 calculate_wavefunction 91 8.0 2.011 2.060 9.618 9.844 potential_pw2rs 186 10.0 0.033 0.034 8.466 9.073 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.582 8.728 9.070 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.136 8.527 local_gemm 172 8.0 8.171 8.502 8.171 8.502 mp2_ri_gpw_compute_en_comm 22 7.0 0.492 0.512 7.941 8.359 calc_potential_gpw 182 9.5 0.002 0.002 7.889 8.099 collocate_single_gaussian 91 10.0 0.017 0.023 7.741 7.977 mp_sync 37 10.5 4.001 6.924 4.001 6.924 integrate_four_center_load 4 10.0 0.000 0.000 6.797 6.801 hfx_load_balance 1 11.0 0.000 0.000 6.796 6.801 mp2_ri_gpw_compute_en_ener 172 7.0 6.349 6.417 6.349 6.417 mp_sendrecv_dm3 2068 8.0 5.991 6.389 5.991 6.389 pw_gather_s 912 13.2 4.488 5.130 4.488 5.130 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.654250, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1499.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_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 454.250496E+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 0.010 0.030 61.206 61.207 qs_mol_dyn_low 1 2.0 0.003 0.003 60.937 60.945 qs_forces 11 3.9 0.003 0.003 60.869 60.870 qs_energies 11 4.9 0.002 0.003 59.210 59.230 scf_env_do_scf 11 5.9 0.001 0.001 52.415 52.416 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 49.753 49.754 qs_scf_new_mos 108 7.5 0.000 0.001 37.876 38.190 qs_scf_loop_do_ot 108 8.5 0.000 0.001 37.876 38.190 dbcsr_multiply_generic 2286 12.5 0.095 0.098 36.624 37.127 ot_scf_mini 108 9.5 0.002 0.002 36.058 36.231 velocity_verlet 10 3.0 0.001 0.002 30.166 30.167 multiply_cannon 2286 13.5 0.192 0.205 27.419 29.174 multiply_cannon_loop 2286 14.5 1.823 1.924 26.448 28.149 ot_mini 108 10.5 0.001 0.001 21.566 21.819 qs_ot_get_derivative 108 11.5 0.001 0.001 18.323 18.544 mp_waitall_1 245248 16.5 9.772 16.087 9.772 16.087 multiply_cannon_metrocomm3 54864 15.5 0.073 0.077 6.484 13.736 multiply_cannon_multrec 54864 15.5 3.602 5.664 7.824 11.317 qs_ot_get_p 119 10.4 0.001 0.001 9.462 9.732 rebuild_ks_matrix 119 8.3 0.000 0.000 8.939 9.090 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.939 9.090 mp_sum_l 7287 12.8 6.343 8.080 6.343 8.080 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.889 8.029 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.230 6.683 multiply_cannon_sync_h2d 54864 15.5 5.066 6.397 5.066 6.397 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.042 6.165 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.989 6.043 dbcsr_mm_accdrv_process 76910 16.1 1.837 2.886 4.135 5.886 qs_rho_update_rho_low 119 7.7 0.000 0.001 5.257 5.399 calculate_rho_elec 119 8.7 0.011 0.016 5.256 5.398 init_scf_run 11 5.9 0.000 0.001 5.370 5.371 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.370 5.370 sum_up_and_integrate 119 10.3 0.001 0.001 5.277 5.282 integrate_v_rspace 119 11.3 0.002 0.002 5.267 5.273 cp_dbcsr_syevd 50 12.0 0.002 0.003 5.070 5.070 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.705 4.706 cp_fm_redistribute_end 50 14.0 2.391 4.653 2.406 4.660 cp_fm_diag_elpa_base 50 14.0 2.243 4.509 2.250 4.521 calculate_dm_sparse 119 9.5 0.000 0.000 3.174 3.363 multiply_cannon_metrocomm1 54864 15.5 0.057 0.062 2.075 3.337 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.041 3.274 apply_single 119 13.6 0.000 0.000 3.041 3.274 density_rs2pw 119 9.7 0.004 0.004 3.081 3.161 jit_kernel_multiply 13 15.8 2.233 2.925 2.233 2.925 ot_diis_step 108 11.5 0.006 0.006 2.882 2.883 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.705 2.816 acc_transpose_blocks 54864 15.5 0.225 0.251 2.269 2.813 pw_transfer 1439 11.6 0.051 0.057 2.641 2.728 calculate_first_density_matrix 1 7.0 0.000 0.000 2.705 2.709 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 2.566 2.656 init_scf_loop 11 6.9 0.000 0.000 2.631 2.632 make_m2s 4572 13.5 0.053 0.055 2.554 2.626 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.591 2.594 potential_pw2rs 119 12.3 0.004 0.004 2.543 2.575 wfi_extrapolate 11 7.9 0.001 0.001 2.561 2.561 make_images 4572 14.5 0.133 0.139 2.472 2.544 fft3d_ps 1201 14.6 0.376 0.490 2.282 2.365 mp_sum_d 4137 12.0 1.616 2.223 1.616 2.223 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.138 2.218 grid_integrate_task_list 119 12.3 1.999 2.106 1.999 2.106 mp_alltoall_d11v 2130 13.8 1.940 2.089 1.940 2.089 fft_wrap_pw1pw2_140 487 13.2 0.106 0.112 1.965 2.055 transfer_rs2pw 487 10.6 0.005 0.006 1.797 1.883 transfer_pw2rs 487 13.2 0.006 0.006 1.742 1.773 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.555 1.587 mp_waitany 12084 13.8 1.332 1.484 1.332 1.484 make_images_sizes 4572 15.5 0.004 0.005 1.154 1.463 mp_alltoall_i44 4572 16.5 1.150 1.458 1.150 1.458 acc_transpose_blocks_sync 164592 16.5 1.209 1.448 1.209 1.448 grid_collocate_task_list 119 9.7 1.353 1.432 1.353 1.432 mp_alltoall_z22v 1201 16.6 1.348 1.415 1.348 1.415 prepare_preconditioner 11 7.9 0.000 0.000 1.347 1.376 make_preconditioner 11 8.9 0.000 0.000 1.347 1.376 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.263 1.298 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.906 1.293 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=61.207000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.818182, yerr=1.695887 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_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 491.425792E+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 1161964. 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 0.018 0.040 42.354 42.356 qs_mol_dyn_low 1 2.0 0.003 0.008 42.063 42.070 qs_forces 11 3.9 0.003 0.011 41.969 41.972 qs_energies 11 4.9 0.009 0.051 40.189 40.198 scf_env_do_scf 11 5.9 0.002 0.013 34.549 34.550 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 31.626 31.626 dbcsr_multiply_generic 2286 12.5 0.101 0.105 22.834 23.184 qs_scf_new_mos 108 7.5 0.001 0.001 22.134 22.366 qs_scf_loop_do_ot 108 8.5 0.001 0.001 22.133 22.366 ot_scf_mini 108 9.5 0.003 0.004 21.174 21.341 velocity_verlet 10 3.0 0.001 0.002 19.917 19.919 multiply_cannon 2286 13.5 0.210 0.218 17.014 18.361 multiply_cannon_loop 2286 14.5 1.180 1.243 15.703 17.252 ot_mini 108 10.5 0.001 0.003 12.788 13.017 mp_waitall_1 200699 16.5 5.946 10.984 5.946 10.984 qs_ot_get_derivative 108 11.5 0.001 0.002 10.360 10.527 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.006 9.380 multiply_cannon_multrec 27432 15.5 1.827 4.262 6.139 8.951 rebuild_ks_matrix 119 8.3 0.000 0.000 7.300 7.452 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.019 7.300 7.452 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.424 6.559 dbcsr_mm_accdrv_process 47894 16.0 3.464 5.584 4.229 6.280 qs_ot_get_p 119 10.4 0.001 0.001 5.358 5.575 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.815 4.734 mp_sum_l 7287 12.8 2.309 4.332 2.309 4.332 init_scf_run 11 5.9 0.000 0.001 4.202 4.203 scf_env_initial_rho_setup 11 6.9 0.002 0.007 4.202 4.203 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.154 4.188 calculate_rho_elec 119 8.7 0.021 0.024 4.154 4.188 sum_up_and_integrate 119 10.3 0.001 0.003 4.129 4.135 integrate_v_rspace 119 11.3 0.002 0.003 4.114 4.121 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 3.112 4.100 apply_single 119 13.6 0.000 0.000 3.112 4.100 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.468 3.485 make_m2s 4572 13.5 0.052 0.054 3.068 3.381 make_images 4572 14.5 0.206 0.244 2.978 3.295 cp_dbcsr_syevd 50 12.0 0.003 0.004 3.031 3.032 init_scf_loop 11 6.9 0.001 0.004 2.881 2.881 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.551 2.636 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.578 2.580 cp_fm_redistribute_end 50 14.0 1.302 2.533 1.311 2.538 density_rs2pw 119 9.7 0.004 0.004 2.414 2.495 calculate_first_density_matrix 1 7.0 0.001 0.004 2.467 2.471 cp_fm_diag_elpa_base 50 14.0 1.190 2.411 1.220 2.455 ot_diis_step 108 11.5 0.011 0.013 2.376 2.377 multiply_cannon_sync_h2d 27432 15.5 1.694 2.238 1.694 2.238 pw_transfer 1439 11.6 0.064 0.068 2.191 2.227 calculate_dm_sparse 119 9.5 0.000 0.001 2.114 2.196 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.100 2.136 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.016 2.021 grid_integrate_task_list 119 12.3 1.834 1.966 1.834 1.966 acc_transpose_blocks 27432 15.5 0.113 0.118 1.612 1.952 jit_kernel_multiply 10 16.3 0.707 1.945 0.707 1.945 prepare_preconditioner 11 7.9 0.000 0.000 1.888 1.916 make_preconditioner 11 8.9 0.001 0.002 1.888 1.916 potential_pw2rs 119 12.3 0.006 0.006 1.847 1.861 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.780 1.841 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.728 1.771 fft3d_ps 1201 14.6 0.527 0.586 1.733 1.769 make_images_data 4572 15.5 0.048 0.055 1.356 1.759 mp_alltoall_d11v 2130 13.8 1.451 1.685 1.451 1.685 wfi_extrapolate 11 7.9 0.001 0.001 1.651 1.651 hybrid_alltoall_any 4725 16.4 0.054 0.117 1.215 1.597 fft_wrap_pw1pw2_140 487 13.2 0.126 0.131 1.553 1.589 transfer_rs2pw 487 10.6 0.005 0.005 1.481 1.577 grid_collocate_task_list 119 9.7 1.292 1.364 1.292 1.364 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.271 1.316 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.282 1.293 transfer_pw2rs 487 13.2 0.004 0.005 1.186 1.198 make_images_sizes 4572 15.5 0.005 0.005 0.847 1.174 mp_alltoall_i44 4572 16.5 0.843 1.169 0.843 1.169 mp_sum_d 4137 12.0 0.669 1.167 0.669 1.167 mp_allgather_i34 2286 14.5 0.734 1.130 0.734 1.130 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.062 1.066 mp_alltoall_z22v 1201 16.6 0.923 1.009 0.923 1.009 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.943 0.958 acc_transpose_blocks_sync 82296 16.5 0.812 0.940 0.812 0.940 acc_transpose_blocks_kernels 27432 16.5 0.189 0.279 0.660 0.869 cp_fm_cholesky_invert 11 10.9 0.858 0.862 0.858 0.862 dbcsr_complete_redistribute 329 12.2 0.083 0.125 0.781 0.857 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=42.356000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=467.636364, yerr=1.720081 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_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 522.653696E+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 0.013 0.031 35.964 35.966 qs_mol_dyn_low 1 2.0 0.003 0.003 35.668 35.677 qs_forces 11 3.9 0.002 0.003 35.600 35.601 qs_energies 11 4.9 0.002 0.002 33.946 33.948 scf_env_do_scf 11 5.9 0.001 0.001 28.698 28.698 scf_env_do_scf_inner_loop 108 6.5 0.002 0.011 25.894 25.894 dbcsr_multiply_generic 2286 12.5 0.095 0.099 17.970 18.062 qs_scf_new_mos 108 7.5 0.001 0.001 17.033 17.052 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.033 17.051 velocity_verlet 10 3.0 0.001 0.002 16.983 16.985 ot_scf_mini 108 9.5 0.002 0.002 16.237 16.253 multiply_cannon 2286 13.5 0.194 0.199 13.862 14.567 multiply_cannon_loop 2286 14.5 0.858 0.907 12.967 13.689 ot_mini 108 10.5 0.001 0.001 9.710 9.728 qs_ot_get_derivative 108 11.5 0.001 0.001 8.169 8.183 multiply_cannon_multrec 18288 15.5 1.889 2.909 6.967 7.334 rebuild_ks_matrix 119 8.3 0.000 0.000 6.682 6.701 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.022 6.682 6.701 dbcsr_mm_accdrv_process 38222 16.0 4.925 5.855 4.983 5.919 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.892 5.910 mp_waitall_1 158411 16.6 3.257 4.476 3.257 4.476 qs_ot_get_p 119 10.4 0.001 0.001 4.168 4.197 sum_up_and_integrate 119 10.3 0.001 0.003 4.145 4.149 integrate_v_rspace 119 11.3 0.002 0.003 4.132 4.138 init_scf_run 11 5.9 0.000 0.001 3.990 3.990 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.989 3.990 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.967 3.974 calculate_rho_elec 119 8.7 0.031 0.032 3.966 3.973 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.039 3.660 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.791 2.798 init_scf_loop 11 6.9 0.000 0.000 2.776 2.777 make_m2s 4572 13.5 0.044 0.045 2.465 2.620 density_rs2pw 119 9.7 0.004 0.004 2.489 2.558 make_images 4572 14.5 0.191 0.204 2.380 2.534 calculate_first_density_matrix 1 7.0 0.000 0.000 2.522 2.523 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.458 2.458 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.472 2.405 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.061 2.383 apply_single 119 13.6 0.000 0.000 2.061 2.382 pw_transfer 1439 11.6 0.065 0.068 2.274 2.302 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.182 2.209 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.120 2.129 cp_fm_diag_elpa_base 50 14.0 2.089 2.108 2.117 2.126 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.054 2.055 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.924 1.935 prepare_preconditioner 11 7.9 0.000 0.000 1.912 1.916 make_preconditioner 11 8.9 0.000 0.000 1.912 1.916 grid_integrate_task_list 119 12.3 1.800 1.916 1.800 1.916 potential_pw2rs 119 12.3 0.007 0.008 1.889 1.907 calculate_dm_sparse 119 9.5 0.000 0.001 1.892 1.904 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.754 1.843 acc_transpose_blocks 18288 15.5 0.079 0.080 1.682 1.779 fft3d_ps 1201 14.6 0.552 0.575 1.740 1.769 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.750 1.759 fft_wrap_pw1pw2_140 487 13.2 0.178 0.183 1.626 1.655 mp_sum_l 7287 12.8 1.268 1.587 1.268 1.587 ot_diis_step 108 11.5 0.011 0.012 1.524 1.524 transfer_rs2pw 487 10.6 0.005 0.005 1.414 1.490 wfi_extrapolate 11 7.9 0.001 0.001 1.404 1.404 grid_collocate_task_list 119 9.7 1.235 1.305 1.235 1.305 transfer_pw2rs 487 13.2 0.004 0.004 1.208 1.224 make_images_data 4572 15.5 0.048 0.051 0.994 1.222 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.208 1.213 multiply_cannon_sync_h2d 18288 15.5 1.032 1.197 1.032 1.197 mp_alltoall_d11v 2130 13.8 0.918 1.087 0.918 1.087 hybrid_alltoall_any 4725 16.4 0.059 0.118 0.855 1.079 mp_alltoall_z22v 1201 16.6 0.993 1.064 0.993 1.064 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.019 1.039 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 0.967 0.968 acc_transpose_blocks_kernels 18288 16.5 0.218 0.226 0.846 0.912 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.868 0.870 cp_fm_cholesky_invert 11 10.9 0.835 0.841 0.835 0.841 make_images_sizes 4572 15.5 0.005 0.005 0.617 0.832 mp_alltoall_i44 4572 16.5 0.612 0.828 0.612 0.828 dbcsr_complete_redistribute 329 12.2 0.130 0.189 0.707 0.817 acc_transpose_blocks_sync 54864 16.5 0.737 0.814 0.737 0.814 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.433 0.756 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.676 0.735 mp_allgather_i34 2286 14.5 0.344 0.729 0.344 0.729 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=35.966000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=498.000000, yerr=1.477098 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_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 557.768704E+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 0.007 0.028 37.609 37.609 qs_mol_dyn_low 1 2.0 0.003 0.003 37.438 37.446 qs_forces 11 3.9 0.002 0.003 37.374 37.374 qs_energies 11 4.9 0.002 0.002 35.639 35.647 scf_env_do_scf 11 5.9 0.000 0.001 30.359 30.361 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 26.648 26.649 dbcsr_multiply_generic 2286 12.5 0.100 0.103 20.078 20.172 velocity_verlet 10 3.0 0.001 0.002 19.268 19.270 qs_scf_new_mos 108 7.5 0.001 0.001 18.322 18.380 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.321 18.380 ot_scf_mini 108 9.5 0.002 0.003 17.275 17.333 multiply_cannon 2286 13.5 0.218 0.225 16.242 16.868 multiply_cannon_loop 2286 14.5 1.529 1.613 15.249 15.729 ot_mini 108 10.5 0.001 0.001 10.556 10.620 multiply_cannon_multrec 27432 15.5 2.450 3.142 9.029 9.303 qs_ot_get_derivative 108 11.5 0.001 0.001 8.652 8.711 dbcsr_mm_accdrv_process 47916 15.9 6.071 7.538 6.475 7.837 rebuild_ks_matrix 119 8.3 0.000 0.000 6.465 6.513 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.465 6.513 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.729 5.771 init_scf_run 11 5.9 0.000 0.001 3.926 3.926 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.925 3.926 qs_ot_get_p 119 10.4 0.001 0.001 3.828 3.908 init_scf_loop 11 6.9 0.000 0.000 3.690 3.691 sum_up_and_integrate 119 10.3 0.001 0.001 3.600 3.608 integrate_v_rspace 119 11.3 0.002 0.003 3.589 3.596 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.268 3.594 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.537 3.582 calculate_rho_elec 119 8.7 0.040 0.046 3.537 3.582 prepare_preconditioner 11 7.9 0.000 0.000 2.815 2.823 make_preconditioner 11 8.9 0.000 0.000 2.815 2.823 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.414 2.746 make_m2s 4572 13.5 0.053 0.055 2.416 2.539 acc_transpose_blocks 27432 15.5 0.118 0.120 2.409 2.515 mp_waitall_1 137007 16.6 1.956 2.446 1.956 2.446 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.403 2.432 make_images 4572 14.5 0.273 0.334 2.308 2.430 calculate_first_density_matrix 1 7.0 0.000 0.000 2.410 2.412 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.348 2.358 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.059 2.258 apply_single 119 13.6 0.000 0.000 2.059 2.258 calculate_dm_sparse 119 9.5 0.000 0.000 2.159 2.219 pw_transfer 1439 11.6 0.064 0.069 2.092 2.132 density_rs2pw 119 9.7 0.004 0.004 1.972 2.089 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.000 2.043 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.938 1.938 grid_integrate_task_list 119 12.3 1.834 1.923 1.834 1.923 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.866 1.867 ot_diis_step 108 11.5 0.012 0.012 1.862 1.862 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.671 1.678 cp_fm_diag_elpa_base 50 14.0 1.635 1.651 1.668 1.676 fft_wrap_pw1pw2_140 487 13.2 0.227 0.235 1.629 1.673 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.654 1.669 jit_kernel_multiply 8 16.4 0.338 1.530 0.338 1.530 acc_transpose_blocks_sync 82296 16.5 1.431 1.528 1.431 1.528 fft3d_ps 1201 14.6 0.593 0.657 1.488 1.513 wfi_extrapolate 11 7.9 0.001 0.001 1.457 1.457 potential_pw2rs 119 12.3 0.008 0.009 1.399 1.407 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.845 1.389 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.325 1.340 grid_collocate_task_list 119 9.7 1.249 1.333 1.249 1.333 mp_sum_l 7287 12.8 0.992 1.306 0.992 1.306 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.245 1.257 cp_fm_upper_to_full 72 14.2 0.839 1.192 0.839 1.192 dbcsr_complete_redistribute 329 12.2 0.124 0.150 0.861 1.146 transfer_rs2pw 487 10.6 0.004 0.005 1.000 1.117 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.084 1.085 make_images_data 4572 15.5 0.048 0.053 0.901 1.028 mp_alltoall_d11v 2130 13.8 0.829 0.979 0.829 0.979 hybrid_alltoall_any 4725 16.4 0.067 0.156 0.784 0.951 cp_fm_cholesky_invert 11 10.9 0.908 0.912 0.908 0.912 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.612 0.891 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.801 0.881 acc_transpose_blocks_kernels 27432 16.5 0.272 0.281 0.831 0.838 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.827 0.832 mp_alltoall_z22v 1201 16.6 0.776 0.804 0.776 0.804 transfer_pw2rs 487 13.2 0.004 0.004 0.799 0.803 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.609000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.818182, yerr=3.157046 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/07/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 117.977176E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1384136 0.0% 0.0% 100.0% average stack size 0.0 0.0 106.2 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 621.846528E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.032 32.986 32.987 qs_mol_dyn_low 1 2.0 0.003 0.003 32.783 32.793 qs_forces 11 3.9 0.002 0.002 32.717 32.719 qs_energies 11 4.9 0.002 0.002 30.953 30.956 scf_env_do_scf 11 5.9 0.000 0.001 25.791 25.792 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 22.788 22.789 velocity_verlet 10 3.0 0.008 0.009 16.786 16.789 dbcsr_multiply_generic 2286 12.5 0.094 0.097 14.461 14.544 qs_scf_new_mos 108 7.5 0.001 0.001 14.172 14.205 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.172 14.204 ot_scf_mini 108 9.5 0.002 0.002 13.395 13.419 multiply_cannon 2286 13.5 0.225 0.231 10.979 11.524 multiply_cannon_loop 2286 14.5 0.643 0.657 9.899 10.093 ot_mini 108 10.5 0.001 0.001 7.410 7.439 rebuild_ks_matrix 119 8.3 0.000 0.000 6.317 6.346 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.316 6.346 multiply_cannon_multrec 9144 15.5 1.694 1.928 5.993 6.243 qs_ot_get_derivative 108 11.5 0.001 0.001 5.993 6.017 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.651 5.678 dbcsr_mm_accdrv_process 12550 15.8 3.608 4.111 4.185 4.272 qs_ot_get_p 119 10.4 0.001 0.001 3.897 3.925 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.896 3.905 calculate_rho_elec 119 8.7 0.060 0.061 3.895 3.904 sum_up_and_integrate 119 10.3 0.001 0.001 3.823 3.829 integrate_v_rspace 119 11.3 0.003 0.003 3.814 3.819 init_scf_run 11 5.9 0.000 0.001 3.642 3.642 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.641 3.642 init_scf_loop 11 6.9 0.000 0.000 2.973 2.974 mp_waitall_1 115863 16.7 2.199 2.844 2.199 2.844 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.555 2.557 pw_transfer 1439 11.6 0.065 0.068 2.479 2.490 make_m2s 4572 13.5 0.033 0.034 2.329 2.483 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.386 2.400 make_images 4572 14.5 0.269 0.300 2.240 2.395 calculate_first_density_matrix 1 7.0 0.000 0.000 2.301 2.302 density_rs2pw 119 9.7 0.003 0.003 2.228 2.296 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.268 2.269 prepare_preconditioner 11 7.9 0.000 0.000 2.161 2.165 make_preconditioner 11 8.9 0.000 0.000 2.160 2.165 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.032 2.062 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.037 2.055 grid_integrate_task_list 119 12.3 1.873 1.938 1.873 1.938 fft_wrap_pw1pw2_140 487 13.2 0.325 0.332 1.871 1.885 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.870 1.878 cp_fm_diag_elpa_base 50 14.0 1.838 1.857 1.867 1.876 calculate_dm_sparse 119 9.5 0.000 0.000 1.824 1.845 fft3d_ps 1201 14.6 0.656 0.669 1.740 1.756 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.649 1.662 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.621 1.623 potential_pw2rs 119 12.3 0.010 0.011 1.576 1.582 acc_transpose_blocks 9144 15.5 0.042 0.043 1.428 1.457 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.407 1.414 ot_diis_step 108 11.5 0.013 0.013 1.392 1.393 grid_collocate_task_list 119 9.7 1.304 1.381 1.304 1.381 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.297 1.321 apply_single 119 13.6 0.000 0.000 1.296 1.321 make_images_data 4572 15.5 0.042 0.047 1.038 1.286 wfi_extrapolate 11 7.9 0.001 0.001 1.276 1.277 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.251 1.252 hybrid_alltoall_any 4725 16.4 0.066 0.177 0.999 1.227 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.163 1.168 mp_alltoall_d11v 2130 13.8 1.057 1.147 1.057 1.147 cp_fm_cholesky_invert 11 10.9 1.141 1.146 1.141 1.146 transfer_rs2pw 487 10.6 0.004 0.004 1.050 1.116 jit_kernel_multiply 6 15.2 0.537 1.059 0.537 1.059 mp_alltoall_z22v 1201 16.6 0.958 1.001 0.958 1.001 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.958 0.964 mp_allgather_i34 2286 14.5 0.396 0.952 0.396 0.952 mp_sum_l 7287 12.8 0.711 0.952 0.711 0.952 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.875 0.925 transfer_pw2rs 487 13.2 0.003 0.003 0.847 0.851 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.449 0.845 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.795 0.797 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.696 0.748 acc_transpose_blocks_sync 27432 16.5 0.723 0.744 0.723 0.744 dbcsr_complete_redistribute 329 12.2 0.163 0.171 0.703 0.739 make_images_sizes 4572 15.5 0.005 0.005 0.509 0.717 mp_alltoall_i44 4572 16.5 0.504 0.712 0.504 0.712 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.547 0.705 arnoldi_extremal 119 11.4 0.002 0.002 0.647 0.671 arnoldi_normal_ev 119 12.4 0.002 0.003 0.645 0.669 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=32.987000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=590.545455, yerr=4.075729 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_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 782.069760E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.033 43.174 43.175 qs_mol_dyn_low 1 2.0 0.003 0.003 42.963 42.971 qs_forces 11 3.9 0.002 0.003 42.863 42.864 qs_energies 11 4.9 0.002 0.002 40.848 40.851 scf_env_do_scf 11 5.9 0.001 0.001 34.947 34.947 scf_env_do_scf_inner_loop 108 6.5 0.029 0.217 27.158 27.159 velocity_verlet 10 3.0 0.040 0.043 24.139 24.145 dbcsr_multiply_generic 2286 12.5 0.101 0.103 18.909 19.085 qs_scf_new_mos 108 7.5 0.001 0.001 17.240 17.339 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.239 17.338 ot_scf_mini 108 9.5 0.002 0.002 16.086 16.188 multiply_cannon 2286 13.5 0.296 0.305 15.017 15.909 multiply_cannon_loop 2286 14.5 0.861 0.884 13.775 14.698 ot_mini 108 10.5 0.001 0.001 9.769 9.886 multiply_cannon_multrec 9144 15.5 3.430 4.771 8.814 8.903 qs_ot_get_derivative 108 11.5 0.001 0.001 7.729 7.831 init_scf_loop 11 6.9 0.000 0.000 7.762 7.763 rebuild_ks_matrix 119 8.3 0.000 0.000 7.263 7.406 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.263 7.406 prepare_preconditioner 11 7.9 0.000 0.000 6.798 6.811 make_preconditioner 11 8.9 0.000 0.000 6.798 6.811 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.385 6.693 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.541 6.670 dbcsr_mm_accdrv_process 12550 15.8 4.486 5.841 5.244 6.576 cp_fm_upper_to_full 72 14.2 3.185 4.548 3.185 4.548 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.416 4.447 calculate_rho_elec 119 8.7 0.117 0.120 4.416 4.447 sum_up_and_integrate 119 10.3 0.001 0.001 3.853 3.860 integrate_v_rspace 119 11.3 0.003 0.003 3.842 3.849 init_scf_run 11 5.9 0.000 0.001 3.817 3.817 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.817 3.817 qs_ot_get_p 119 10.4 0.001 0.001 3.393 3.530 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.830 3.246 mp_waitall_1 94719 16.7 2.230 3.210 2.230 3.210 pw_transfer 1439 11.6 0.068 0.069 3.043 3.050 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.945 2.951 dbcsr_complete_redistribute 329 12.2 0.283 0.290 1.946 2.740 fft_wrap_pw1pw2_140 487 13.2 0.649 0.653 2.510 2.519 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.198 2.514 apply_single 119 13.6 0.000 0.000 2.197 2.513 make_m2s 4572 13.5 0.037 0.037 2.337 2.480 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.643 2.435 make_images 4572 14.5 0.356 0.388 2.216 2.361 calculate_first_density_matrix 1 7.0 0.000 0.000 2.342 2.344 mp_alltoall_i22 627 13.8 1.454 2.268 1.454 2.268 density_rs2pw 119 9.7 0.003 0.004 2.246 2.263 calculate_dm_sparse 119 9.5 0.000 0.000 2.230 2.246 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.318 2.205 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.408 2.193 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.073 2.126 grid_integrate_task_list 119 12.3 2.078 2.103 2.078 2.103 qs_ot_p2m_diag 50 11.0 0.044 0.044 2.021 2.023 ot_diis_step 108 11.5 0.014 0.014 2.015 2.015 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.927 1.928 fft3d_ps 1201 14.6 0.870 0.889 1.866 1.870 acc_transpose_blocks 9144 15.5 0.044 0.045 1.814 1.841 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.791 1.792 mp_sum_l 7287 12.8 1.013 1.755 1.013 1.755 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.663 1.663 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.614 1.638 grid_collocate_task_list 119 9.7 1.520 1.535 1.520 1.535 cp_fm_cholesky_invert 11 10.9 1.430 1.435 1.430 1.435 mp_alltoall_d11v 2130 13.8 1.353 1.417 1.353 1.417 wfi_extrapolate 11 7.9 0.001 0.001 1.411 1.411 potential_pw2rs 119 12.3 0.014 0.014 1.376 1.377 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.369 1.369 cp_fm_diag_elpa_base 50 14.0 1.223 1.275 1.368 1.368 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.311 1.329 hybrid_alltoall_any 4725 16.4 0.092 0.154 1.055 1.272 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.242 1.249 make_images_data 4572 15.5 0.047 0.050 0.980 1.172 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.119 1.136 acc_transpose_blocks_sync 27432 16.5 1.102 1.133 1.102 1.133 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.993 1.027 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.931 0.945 qs_create_task_list 11 7.9 0.001 0.001 0.932 0.942 generate_qs_task_list 11 8.9 0.370 0.390 0.931 0.941 mp_alltoall_z22v 1201 16.6 0.860 0.883 0.860 0.883 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.175000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=733.272727, yerr=13.232817 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_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 502.747136E+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 0.007 0.027 83.423 83.424 qs_mol_dyn_low 1 2.0 0.003 0.003 83.162 83.173 qs_forces 11 3.9 0.003 0.004 83.090 83.091 qs_energies 11 4.9 0.002 0.002 80.182 80.191 scf_env_do_scf 11 5.9 0.000 0.001 71.138 71.141 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.439 65.440 dbcsr_multiply_generic 2055 12.4 0.104 0.108 51.282 51.485 qs_scf_new_mos 99 7.5 0.000 0.001 48.037 48.148 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.037 48.147 ot_scf_mini 99 9.5 0.002 0.002 45.660 45.736 velocity_verlet 10 3.0 0.001 0.002 43.878 43.878 multiply_cannon 2055 13.4 0.182 0.188 42.783 43.665 multiply_cannon_loop 2055 14.4 1.787 1.827 41.766 42.665 ot_mini 99 10.5 0.001 0.001 27.015 27.096 qs_ot_get_derivative 99 11.5 0.001 0.001 20.152 20.247 multiply_cannon_multrec 49320 15.4 11.377 12.039 17.366 18.117 rebuild_ks_matrix 110 8.3 0.000 0.001 14.760 14.904 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.759 14.903 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.939 13.071 mp_waitall_1 220248 16.4 11.136 11.848 11.136 11.848 qs_ot_get_p 110 10.4 0.001 0.002 10.152 10.274 multiply_cannon_sync_h2d 49320 15.4 9.636 10.226 9.636 10.226 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.894 8.422 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.213 7.709 apply_single 110 13.6 0.000 0.001 7.213 7.709 multiply_cannon_metrocomm3 49320 15.4 0.083 0.087 6.635 7.604 sum_up_and_integrate 110 10.3 0.001 0.003 7.206 7.221 integrate_v_rspace 110 11.3 0.002 0.003 7.181 7.203 init_scf_run 11 5.9 0.000 0.001 6.976 6.976 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.976 6.976 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.904 6.950 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.533 6.663 calculate_rho_elec 110 8.6 0.020 0.024 6.532 6.662 ot_diis_step 99 11.5 0.005 0.006 6.624 6.624 dbcsr_mm_accdrv_process 87628 16.1 3.033 3.112 5.857 6.143 cp_dbcsr_syevd 48 12.0 0.002 0.003 6.006 6.006 init_scf_loop 11 6.9 0.000 0.000 5.669 5.670 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.453 5.505 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.388 5.411 cp_fm_diag_elpa_base 48 14.0 5.372 5.396 5.386 5.409 mp_sum_l 6594 12.7 3.967 4.685 3.967 4.685 make_m2s 4110 13.4 0.060 0.065 4.076 4.187 wfi_extrapolate 11 7.9 0.001 0.001 4.147 4.147 make_images 4110 14.4 0.177 0.190 3.981 4.095 calculate_dm_sparse 110 9.5 0.001 0.001 3.746 3.838 density_rs2pw 110 9.6 0.004 0.005 3.573 3.743 multiply_cannon_metrocomm1 49320 15.4 0.066 0.069 2.483 3.557 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.546 3.551 prepare_preconditioner 11 7.9 0.000 0.000 3.460 3.476 make_preconditioner 11 8.9 0.000 0.000 3.460 3.476 grid_integrate_task_list 110 12.3 3.252 3.427 3.252 3.427 pw_transfer 1331 11.6 0.055 0.063 3.263 3.364 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.175 3.279 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.237 3.279 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.200 3.250 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.200 3.247 potential_pw2rs 110 12.3 0.005 0.006 2.775 2.802 fft_wrap_pw1pw2_140 451 13.1 0.310 0.332 2.668 2.786 calculate_first_density_matrix 1 7.0 0.000 0.000 2.716 2.720 acc_transpose_blocks 49320 15.4 0.211 0.217 2.558 2.603 mp_alltoall_d11v 2046 13.8 2.148 2.587 2.148 2.587 jit_kernel_multiply 13 15.9 2.542 2.555 2.542 2.555 fft3d_ps 1111 14.6 0.797 0.887 2.470 2.549 grid_collocate_task_list 110 9.6 2.159 2.285 2.159 2.285 transfer_rs2pw 451 10.6 0.005 0.006 2.024 2.195 mp_waitany 14300 13.8 1.849 2.154 1.849 2.154 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.954 1.968 make_images_data 4110 15.4 0.044 0.049 1.822 1.953 transfer_pw2rs 451 13.1 0.006 0.007 1.889 1.910 cp_fm_cholesky_invert 11 10.9 1.890 1.894 1.890 1.894 mp_sum_d 3891 11.9 1.399 1.874 1.399 1.874 hybrid_alltoall_any 4261 16.3 0.084 0.483 1.596 1.816 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.654 1.678 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.424000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.818182, yerr=2.948595 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_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 593.018880E+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 1305749. 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 0.014 0.034 71.844 71.847 qs_mol_dyn_low 1 2.0 0.003 0.005 71.575 71.585 qs_forces 11 3.9 0.003 0.003 71.502 71.503 qs_energies 11 4.9 0.003 0.008 68.152 68.156 scf_env_do_scf 11 5.9 0.001 0.001 59.529 59.533 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 51.379 51.380 dbcsr_multiply_generic 2055 12.4 0.115 0.119 38.674 38.900 velocity_verlet 10 3.0 0.001 0.002 37.156 37.157 qs_scf_new_mos 99 7.5 0.001 0.001 35.035 35.185 qs_scf_loop_do_ot 99 8.5 0.001 0.001 35.035 35.184 ot_scf_mini 99 9.5 0.003 0.004 33.338 33.485 multiply_cannon 2055 13.4 0.224 0.246 31.531 32.431 multiply_cannon_loop 2055 14.4 1.159 1.188 30.154 30.968 ot_mini 99 10.5 0.001 0.001 19.035 19.183 multiply_cannon_multrec 24660 15.4 6.958 8.684 14.017 15.583 rebuild_ks_matrix 110 8.3 0.000 0.000 13.826 13.997 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 13.826 13.997 qs_ot_get_derivative 99 11.5 0.001 0.001 13.170 13.314 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.188 12.349 mp_waitall_1 176588 16.5 7.891 10.438 7.891 10.438 init_scf_loop 11 6.9 0.001 0.004 8.111 8.112 multiply_cannon_metrocomm3 24660 15.4 0.074 0.077 5.267 7.903 qs_ot_get_p 110 10.4 0.001 0.001 7.530 7.684 multiply_cannon_sync_h2d 24660 15.4 6.383 7.423 6.383 7.423 dbcsr_mm_accdrv_process 52282 16.1 5.555 6.506 6.889 7.233 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.533 7.202 apply_single 110 13.6 0.000 0.001 6.533 7.202 sum_up_and_integrate 110 10.3 0.002 0.003 6.592 6.601 integrate_v_rspace 110 11.3 0.002 0.003 6.565 6.576 init_scf_run 11 5.9 0.000 0.001 6.168 6.169 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.168 6.169 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.135 6.144 calculate_rho_elec 110 8.6 0.039 0.047 6.134 6.144 prepare_preconditioner 11 7.9 0.000 0.000 6.065 6.087 make_preconditioner 11 8.9 0.001 0.004 6.065 6.087 ot_diis_step 99 11.5 0.010 0.010 5.810 5.810 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.653 5.810 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.851 5.591 qs_ot_p2m_diag 48 11.0 0.029 0.044 5.448 5.476 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.944 4.945 make_m2s 4110 13.4 0.056 0.059 4.503 4.933 make_images 4110 14.4 0.406 0.463 4.391 4.816 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.275 4.284 cp_fm_diag_elpa_base 48 14.0 4.222 4.240 4.272 4.282 pw_transfer 1331 11.6 0.066 0.074 3.772 3.925 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.665 3.819 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.578 3.662 wfi_extrapolate 11 7.9 0.001 0.001 3.571 3.571 density_rs2pw 110 9.6 0.004 0.004 3.350 3.568 grid_integrate_task_list 110 12.3 3.154 3.367 3.154 3.367 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.261 3.263 fft_wrap_pw1pw2_140 451 13.1 0.357 0.379 3.043 3.194 calculate_dm_sparse 110 9.5 0.001 0.001 3.001 3.031 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.906 2.959 cp_fm_cholesky_invert 11 10.9 2.919 2.926 2.919 2.926 fft3d_ps 1111 14.6 1.119 1.353 2.759 2.924 make_images_data 4110 15.4 0.049 0.053 2.428 2.873 hybrid_alltoall_any 4261 16.3 0.106 0.455 2.154 2.871 mp_sum_l 6594 12.7 2.014 2.759 2.014 2.759 calculate_first_density_matrix 1 7.0 0.000 0.002 2.500 2.501 potential_pw2rs 110 12.3 0.008 0.009 2.438 2.452 grid_collocate_task_list 110 9.6 2.172 2.358 2.172 2.358 mp_alltoall_d11v 2046 13.8 1.914 2.227 1.914 2.227 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.047 2.070 acc_transpose_blocks 24660 15.4 0.115 0.120 1.980 2.007 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.922 1.923 transfer_rs2pw 451 10.6 0.006 0.007 1.576 1.832 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.799 1.810 mp_allgather_i34 2055 14.4 0.689 1.774 0.689 1.774 multiply_cannon_metrocomm4 22605 15.4 0.073 0.078 0.762 1.772 jit_kernel_multiply 9 16.3 0.964 1.769 0.964 1.769 mp_irecv_dv 57340 16.2 0.638 1.553 0.638 1.553 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.543 1.552 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.389 1.494 dbcsr_complete_redistribute 325 12.2 0.237 0.313 1.219 1.487 cp_fm_cholesky_decompose 22 10.9 1.461 1.467 1.461 1.467 mp_alltoall_z22v 1111 16.6 1.368 1.450 1.368 1.450 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.847000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.727273, yerr=7.556946 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 668.135424E+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 0.013 0.045 61.934 61.941 qs_mol_dyn_low 1 2.0 0.003 0.004 61.650 61.663 qs_forces 11 3.9 0.003 0.004 61.579 61.580 qs_energies 11 4.9 0.002 0.008 58.376 58.380 scf_env_do_scf 11 5.9 0.001 0.002 50.289 50.289 scf_env_do_scf_inner_loop 99 6.5 0.003 0.010 41.438 41.438 velocity_verlet 10 3.0 0.001 0.002 33.548 33.550 dbcsr_multiply_generic 2055 12.4 0.110 0.112 29.466 29.788 qs_scf_new_mos 99 7.5 0.001 0.001 26.405 26.515 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.405 26.515 ot_scf_mini 99 9.5 0.002 0.003 25.133 25.263 multiply_cannon 2055 13.4 0.211 0.219 22.818 23.965 multiply_cannon_loop 2055 14.4 0.812 0.860 21.593 22.886 ot_mini 99 10.5 0.001 0.001 14.251 14.384 rebuild_ks_matrix 110 8.3 0.000 0.000 12.406 12.571 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.405 12.570 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.933 11.085 multiply_cannon_multrec 16440 15.4 3.524 4.672 9.727 10.917 mp_waitall_1 139946 16.5 7.529 10.732 7.529 10.732 qs_ot_get_derivative 99 11.5 0.001 0.001 9.691 9.821 init_scf_loop 11 6.9 0.001 0.003 8.811 8.812 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.748 7.774 prepare_preconditioner 11 7.9 0.000 0.000 7.015 7.032 make_preconditioner 11 8.9 0.000 0.001 7.015 7.032 make_full_inverse_cholesky 11 9.9 0.000 0.001 6.341 6.699 sum_up_and_integrate 110 10.3 0.001 0.002 6.411 6.426 integrate_v_rspace 110 11.3 0.003 0.003 6.385 6.400 dbcsr_mm_accdrv_process 34862 16.1 5.296 5.762 6.050 6.257 qs_ot_get_p 110 10.4 0.001 0.002 5.940 6.095 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.952 5.964 calculate_rho_elec 110 8.6 0.059 0.059 5.952 5.964 init_scf_run 11 5.9 0.000 0.001 5.652 5.652 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.652 5.652 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.017 5.545 apply_single 110 13.6 0.000 0.000 5.017 5.545 make_m2s 4110 13.4 0.049 0.051 4.395 4.754 make_images 4110 14.4 0.401 0.534 4.278 4.635 ot_diis_step 99 11.5 0.011 0.011 4.521 4.522 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.197 4.204 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.316 3.940 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.812 3.812 multiply_cannon_sync_h2d 16440 15.4 3.178 3.797 3.178 3.797 pw_transfer 1331 11.6 0.065 0.072 3.736 3.750 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.629 3.645 grid_integrate_task_list 110 12.3 3.192 3.430 3.192 3.430 density_rs2pw 110 9.6 0.004 0.004 3.164 3.357 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.224 3.233 cp_fm_diag_elpa_base 48 14.0 3.153 3.187 3.222 3.230 fft_wrap_pw1pw2_140 451 13.1 0.462 0.475 3.094 3.110 wfi_extrapolate 11 7.9 0.001 0.001 3.096 3.096 make_images_data 4110 15.4 0.047 0.051 2.473 2.951 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.868 2.869 cp_fm_cholesky_invert 11 10.9 2.822 2.828 2.822 2.828 hybrid_alltoall_any 4261 16.3 0.109 0.388 2.178 2.781 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.609 2.688 calculate_dm_sparse 110 9.5 0.001 0.001 2.554 2.582 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.517 2.579 multiply_cannon_metrocomm4 14385 15.4 0.047 0.051 0.918 2.576 fft3d_ps 1111 14.6 1.121 1.134 2.544 2.565 calculate_first_density_matrix 1 7.0 0.000 0.003 2.454 2.455 mp_irecv_dv 48980 15.7 0.843 2.440 0.843 2.440 mp_sum_l 6594 12.7 1.631 2.354 1.631 2.354 grid_collocate_task_list 110 9.6 2.220 2.330 2.220 2.330 potential_pw2rs 110 12.3 0.010 0.011 2.238 2.248 mp_alltoall_d11v 2046 13.8 1.860 2.129 1.860 2.129 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.973 1.974 dbcsr_complete_redistribute 325 12.2 0.325 0.346 1.491 1.962 cp_fm_upper_to_full 70 14.2 1.426 1.796 1.426 1.796 cp_fm_cholesky_decompose 22 10.9 1.679 1.706 1.679 1.706 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.670 1.688 mp_allgather_i34 2055 14.4 0.542 1.667 0.542 1.667 acc_transpose_blocks 16440 15.4 0.078 0.080 1.585 1.666 transfer_rs2pw 451 10.6 0.005 0.006 1.382 1.615 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.477 1.487 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.998 1.471 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.367 1.468 mp_waitany 17072 13.8 1.156 1.441 1.156 1.441 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.340 1.348 transfer_pw2rs 451 13.1 0.005 0.005 1.244 1.254 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.941000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=633.000000, yerr=9.695360 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_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 738.709504E+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 0.010 0.030 68.560 68.561 qs_mol_dyn_low 1 2.0 0.003 0.003 68.325 68.335 qs_forces 11 3.9 0.003 0.004 68.246 68.250 qs_energies 11 4.9 0.002 0.002 64.801 64.807 scf_env_do_scf 11 5.9 0.000 0.001 56.349 56.352 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 44.014 44.014 velocity_verlet 10 3.0 0.001 0.002 38.561 38.563 dbcsr_multiply_generic 2055 12.4 0.116 0.121 31.220 31.401 qs_scf_new_mos 99 7.5 0.001 0.001 28.935 29.044 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.935 29.043 ot_scf_mini 99 9.5 0.002 0.003 27.224 27.333 multiply_cannon 2055 13.4 0.236 0.249 23.658 24.737 multiply_cannon_loop 2055 14.4 1.419 1.475 22.118 22.736 ot_mini 99 10.5 0.001 0.001 15.225 15.367 multiply_cannon_multrec 24660 15.4 4.089 6.710 13.007 14.255 rebuild_ks_matrix 110 8.3 0.000 0.000 12.306 12.425 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.305 12.424 init_scf_loop 11 6.9 0.000 0.000 12.292 12.292 qs_ot_get_derivative 99 11.5 0.001 0.001 11.017 11.133 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.879 10.981 prepare_preconditioner 11 7.9 0.000 0.000 10.494 10.512 make_preconditioner 11 8.9 0.000 0.000 10.494 10.512 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.686 10.172 dbcsr_mm_accdrv_process 52304 16.0 7.749 9.089 8.761 9.679 qs_ot_get_p 110 10.4 0.001 0.001 6.775 6.935 sum_up_and_integrate 110 10.3 0.001 0.002 6.468 6.485 integrate_v_rspace 110 11.3 0.003 0.003 6.442 6.457 make_m2s 4110 13.4 0.059 0.062 5.795 6.172 mp_waitall_1 121746 16.5 4.448 6.164 4.448 6.164 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.080 6.091 calculate_rho_elec 110 8.6 0.078 0.081 6.079 6.091 make_images 4110 14.4 0.575 0.693 5.654 6.028 init_scf_run 11 5.9 0.000 0.001 5.678 5.678 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.677 5.678 cp_fm_upper_to_full 70 14.2 3.450 4.987 3.450 4.987 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.746 4.761 ot_diis_step 99 11.5 0.011 0.011 4.166 4.166 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.160 4.161 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.057 4.137 apply_single 110 13.6 0.000 0.000 4.057 4.137 pw_transfer 1331 11.6 0.065 0.073 3.980 4.015 dbcsr_complete_redistribute 325 12.2 0.412 0.456 2.782 3.933 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.873 3.912 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.644 3.697 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.512 3.578 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.531 3.542 cp_fm_diag_elpa_base 48 14.0 3.367 3.433 3.528 3.539 grid_integrate_task_list 110 12.3 3.280 3.413 3.280 3.413 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.201 3.336 density_rs2pw 110 9.6 0.004 0.004 3.166 3.330 fft_wrap_pw1pw2_140 451 13.1 0.532 0.543 3.262 3.302 hybrid_alltoall_any 4261 16.3 0.123 0.459 2.413 3.292 make_images_data 4110 15.4 0.049 0.053 2.871 3.283 wfi_extrapolate 11 7.9 0.001 0.001 3.148 3.148 calculate_dm_sparse 110 9.5 0.001 0.001 3.022 3.051 cp_fm_cholesky_invert 11 10.9 3.002 3.011 3.002 3.011 multiply_cannon_metrocomm3 24660 15.4 0.037 0.039 1.324 2.981 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.799 2.913 mp_alltoall_i22 605 13.7 1.698 2.894 1.698 2.894 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.864 2.865 fft3d_ps 1111 14.6 1.154 1.195 2.699 2.714 acc_transpose_blocks 24660 15.4 0.111 0.114 2.433 2.530 multiply_cannon_sync_h2d 24660 15.4 2.359 2.521 2.359 2.521 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.463 2.504 calculate_first_density_matrix 1 7.0 0.000 0.000 2.426 2.429 grid_collocate_task_list 110 9.6 2.263 2.404 2.263 2.404 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.253 2.256 potential_pw2rs 110 12.3 0.012 0.012 2.232 2.240 mp_alltoall_d11v 2046 13.8 1.948 2.147 1.948 2.147 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.840 1.877 cp_fm_cholesky_decompose 22 10.9 1.835 1.869 1.835 1.869 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.615 1.717 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.674 1.688 mp_allgather_i34 2055 14.4 0.626 1.643 0.626 1.643 mp_sum_l 6594 12.7 1.077 1.604 1.077 1.604 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.560 1.573 acc_transpose_blocks_sync 73980 16.4 1.414 1.506 1.414 1.506 multiply_cannon_metrocomm4 20550 15.4 0.059 0.061 0.856 1.504 transfer_rs2pw 451 10.6 0.005 0.005 1.267 1.457 jit_kernel_multiply 8 15.9 0.663 1.433 0.663 1.433 mp_irecv_dv 62702 16.1 0.754 1.423 0.754 1.423 mp_alltoall_z22v 1111 16.6 1.330 1.405 1.330 1.405 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=68.561000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=698.090909, yerr=11.492721 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_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 864.641024E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.030 57.724 57.725 qs_mol_dyn_low 1 2.0 0.003 0.003 57.474 57.485 qs_forces 11 3.9 0.003 0.003 57.403 57.404 qs_energies 11 4.9 0.002 0.002 53.708 53.712 scf_env_do_scf 11 5.9 0.000 0.001 45.350 45.350 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 37.181 37.182 velocity_verlet 10 3.0 0.002 0.002 32.149 32.152 dbcsr_multiply_generic 2055 12.4 0.107 0.110 23.875 24.031 qs_scf_new_mos 99 7.5 0.001 0.001 22.138 22.173 qs_scf_loop_do_ot 99 8.5 0.001 0.001 22.138 22.172 ot_scf_mini 99 9.5 0.002 0.002 20.847 20.869 multiply_cannon 2055 13.4 0.235 0.244 17.910 19.426 multiply_cannon_loop 2055 14.4 0.603 0.622 16.587 16.965 rebuild_ks_matrix 110 8.3 0.000 0.000 11.978 11.997 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.978 11.996 ot_mini 99 10.5 0.001 0.001 11.206 11.220 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.649 10.664 multiply_cannon_multrec 8220 15.4 3.229 4.362 7.630 8.553 init_scf_loop 11 6.9 0.000 0.000 8.119 8.121 mp_waitall_1 103326 16.6 6.174 7.811 6.174 7.811 qs_ot_get_derivative 99 11.5 0.001 0.001 7.486 7.507 sum_up_and_integrate 110 10.3 0.001 0.002 6.426 6.440 prepare_preconditioner 11 7.9 0.000 0.000 6.416 6.421 make_preconditioner 11 8.9 0.000 0.000 6.416 6.421 integrate_v_rspace 110 11.3 0.004 0.016 6.400 6.413 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.219 6.235 calculate_rho_elec 110 8.6 0.113 0.113 6.219 6.235 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.995 6.072 qs_ot_get_p 110 10.4 0.001 0.001 5.615 5.635 init_scf_run 11 5.9 0.000 0.001 5.186 5.186 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.186 5.186 dbcsr_mm_accdrv_process 17442 15.9 3.046 4.103 4.261 5.172 make_m2s 4110 13.4 0.037 0.039 4.468 4.756 make_images 4110 14.4 0.645 0.709 4.338 4.627 pw_transfer 1331 11.6 0.065 0.069 4.306 4.316 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.918 4.255 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.199 4.211 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.086 4.091 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.729 3.800 apply_single 110 13.6 0.000 0.000 3.729 3.799 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.721 3.722 ot_diis_step 99 11.5 0.012 0.012 3.692 3.693 fft_wrap_pw1pw2_140 451 13.1 0.721 0.733 3.590 3.609 grid_integrate_task_list 110 12.3 3.375 3.500 3.375 3.500 density_rs2pw 110 9.6 0.004 0.004 3.213 3.332 cp_fm_cholesky_invert 11 10.9 3.148 3.153 3.148 3.153 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.130 3.138 cp_fm_diag_elpa_base 48 14.0 3.069 3.100 3.127 3.135 wfi_extrapolate 11 7.9 0.001 0.001 2.799 2.799 hybrid_alltoall_any 4261 16.3 0.200 0.848 2.336 2.793 make_images_data 4110 15.4 0.041 0.047 2.383 2.761 fft3d_ps 1111 14.6 1.296 1.303 2.742 2.755 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.692 2.693 calculate_dm_sparse 110 9.5 0.001 0.001 2.524 2.562 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.550 2.551 multiply_cannon_sync_h2d 8220 15.4 2.378 2.537 2.378 2.537 grid_collocate_task_list 110 9.6 2.370 2.504 2.370 2.504 calculate_first_density_matrix 1 7.0 0.000 0.000 2.286 2.287 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.164 2.181 potential_pw2rs 110 12.3 0.015 0.015 2.173 2.177 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.116 2.143 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.067 2.078 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.787 1.996 mp_alltoall_d11v 2046 13.8 1.786 1.932 1.786 1.932 cp_fm_cholesky_decompose 22 10.9 1.801 1.814 1.801 1.814 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.661 1.665 mp_allgather_i34 2055 14.4 0.496 1.645 0.496 1.645 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.517 1.633 dbcsr_complete_redistribute 325 12.2 0.548 0.564 1.508 1.613 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.984 1.521 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.441 1.453 acc_transpose_blocks 8220 15.4 0.040 0.041 1.392 1.422 jit_kernel_multiply 8 15.6 0.892 1.381 0.892 1.381 transfer_rs2pw 451 10.6 0.005 0.005 1.195 1.344 qs_create_task_list 11 7.9 0.001 0.001 1.221 1.321 generate_qs_task_list 11 8.9 0.377 0.446 1.220 1.320 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.226 1.247 mp_alltoall_z22v 1111 16.6 1.212 1.234 1.212 1.234 mp_waitany 9240 13.8 1.083 1.232 1.083 1.232 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.467 1.164 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.725000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=814.090909, yerr=10.299643 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_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 1.387086E+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 0.017 0.036 90.202 90.203 qs_mol_dyn_low 1 2.0 0.003 0.003 89.884 89.895 qs_forces 11 3.9 0.003 0.003 89.812 89.812 qs_energies 11 4.9 0.002 0.002 85.629 85.631 scf_env_do_scf 11 5.9 0.000 0.001 75.334 75.334 velocity_verlet 10 3.0 0.003 0.003 57.154 57.160 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 46.228 46.230 dbcsr_multiply_generic 2055 12.4 0.122 0.126 29.870 29.957 init_scf_loop 11 6.9 0.000 0.000 29.029 29.034 qs_scf_new_mos 99 7.5 0.001 0.001 27.857 27.891 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.856 27.891 prepare_preconditioner 11 7.9 0.000 0.000 26.950 26.958 make_preconditioner 11 8.9 0.000 0.000 26.950 26.958 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.026 26.417 ot_scf_mini 99 9.5 0.002 0.002 26.028 26.059 multiply_cannon 2055 13.4 0.333 0.354 22.500 23.280 multiply_cannon_loop 2055 14.4 0.823 0.836 20.704 20.993 cp_fm_upper_to_full 70 14.2 12.837 18.488 12.837 18.488 ot_mini 99 10.5 0.001 0.001 14.354 14.383 rebuild_ks_matrix 110 8.3 0.000 0.001 14.210 14.246 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.016 14.210 14.245 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.891 12.925 dbcsr_complete_redistribute 325 12.2 1.010 1.050 7.526 10.807 multiply_cannon_multrec 8220 15.4 4.051 4.231 9.844 9.960 qs_ot_get_derivative 99 11.5 0.001 0.001 9.760 9.792 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.498 9.789 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.908 9.165 mp_alltoall_i22 605 13.7 5.532 8.844 5.532 8.844 mp_waitall_1 84994 16.7 7.650 8.503 7.650 8.503 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.743 7.777 calculate_rho_elec 110 8.6 0.222 0.223 7.742 7.776 sum_up_and_integrate 110 10.3 0.002 0.002 7.101 7.117 integrate_v_rspace 110 11.3 0.003 0.004 7.073 7.090 qs_ot_get_p 110 10.4 0.001 0.001 6.360 6.395 pw_transfer 1331 11.6 0.074 0.075 6.078 6.082 make_m2s 4110 13.4 0.042 0.043 5.577 6.058 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.961 5.965 init_scf_run 11 5.9 0.000 0.001 5.909 5.909 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.909 5.909 make_images 4110 14.4 0.882 0.939 5.388 5.869 dbcsr_mm_accdrv_process 11614 15.7 3.816 4.104 5.646 5.866 cp_fm_cholesky_invert 11 10.9 5.557 5.562 5.557 5.562 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.726 5.191 apply_single 110 13.6 0.000 0.000 4.726 5.191 fft_wrap_pw1pw2_140 451 13.1 1.365 1.368 5.175 5.181 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.825 5.134 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.663 4.672 ot_diis_step 99 11.5 0.015 0.016 4.576 4.576 density_rs2pw 110 9.6 0.004 0.004 4.163 4.184 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.130 4.131 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.762 3.762 grid_integrate_task_list 110 12.3 3.685 3.754 3.685 3.754 fft3d_ps 1111 14.6 1.887 1.894 3.669 3.676 hybrid_alltoall_any 4261 16.3 0.268 0.588 2.898 3.650 make_images_data 4110 15.4 0.046 0.049 2.919 3.582 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.076 3.534 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.488 3.489 cp_fm_diag_elpa_base 48 14.0 2.937 3.147 3.485 3.486 wfi_extrapolate 11 7.9 0.001 0.001 3.391 3.391 calculate_dm_sparse 110 9.5 0.001 0.001 3.198 3.225 multiply_cannon_sync_h2d 8220 15.4 3.125 3.141 3.125 3.141 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.019 3.020 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.750 2.766 grid_collocate_task_list 110 9.6 2.674 2.689 2.674 2.689 potential_pw2rs 110 12.3 0.021 0.021 2.598 2.601 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.507 2.517 calculate_first_density_matrix 1 7.0 0.000 0.000 2.397 2.398 qs_env_update_s_mstruct 11 6.9 0.001 0.002 2.242 2.330 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.100 2.191 mp_alltoall_d11v 2046 13.8 2.135 2.185 2.135 2.185 cp_fm_cholesky_decompose 22 10.9 2.156 2.168 2.156 2.168 qs_create_task_list 11 7.9 0.001 0.001 1.886 1.931 generate_qs_task_list 11 8.9 0.736 0.790 1.886 1.930 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.922 1.924 jit_kernel_multiply 10 15.2 1.630 1.873 1.630 1.873 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=90.203000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1246.181818, yerr=56.499104 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 633.675776E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 57688. MP_Allreduce 11097 796. MP_Sync 86 MP_Alltoall 2210 1936085. MP_SendRecv 24130 18752. MP_ISendRecv 24130 18752. MP_Wait 42150 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.061 250.137 250.152 qs_mol_dyn_low 1 2.0 0.003 0.005 249.450 249.509 qs_forces 11 3.9 0.005 0.005 249.339 249.341 qs_energies 11 4.9 0.002 0.003 243.163 243.176 scf_env_do_scf 11 5.9 0.001 0.002 225.011 225.016 scf_env_do_scf_inner_loop 116 6.6 0.003 0.010 197.672 197.674 qs_scf_new_mos 116 7.6 0.001 0.001 150.231 150.540 qs_scf_loop_do_ot 116 8.6 0.001 0.001 150.230 150.539 velocity_verlet 10 3.0 0.002 0.005 146.012 146.014 ot_scf_mini 116 9.6 0.003 0.003 143.404 143.736 dbcsr_multiply_generic 2485 12.5 0.178 0.184 129.342 130.016 multiply_cannon 2485 13.5 0.236 0.248 102.053 104.164 multiply_cannon_loop 2485 14.5 2.401 2.452 99.090 101.336 ot_mini 116 10.6 0.001 0.001 69.781 70.058 qs_ot_get_p 127 10.4 0.001 0.001 51.169 51.421 qs_ot_get_derivative 116 11.6 0.001 0.001 44.888 45.214 multiply_cannon_multrec 59640 15.5 31.307 33.110 41.124 42.918 qs_ot_p2m_diag 82 11.4 0.078 0.090 42.839 42.884 cp_dbcsr_syevd 82 12.4 0.005 0.005 39.841 39.846 rebuild_ks_matrix 127 8.3 0.001 0.001 38.113 38.586 qs_ks_build_kohn_sham_matrix 127 9.3 0.029 0.060 38.113 38.585 mp_waitall_1 264810 16.5 33.619 36.533 33.619 36.533 cp_fm_diag_elpa 82 13.4 0.000 0.000 35.283 35.311 cp_fm_diag_elpa_base 82 14.4 35.097 35.189 35.272 35.301 qs_ks_update_qs_env 127 7.6 0.001 0.001 34.276 34.704 multiply_cannon_sync_h2d 59640 15.5 25.791 27.689 25.791 27.689 init_scf_loop 11 6.9 0.000 0.001 27.220 27.222 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.358 25.000 apply_single 127 13.6 0.001 0.001 24.358 25.000 ot_diis_step 116 11.6 0.007 0.008 24.607 24.609 prepare_preconditioner 11 7.9 0.000 0.000 22.061 22.117 make_preconditioner 11 8.9 0.000 0.000 22.061 22.117 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 21.512 21.745 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.150 21.386 multiply_cannon_metrocomm3 59640 15.5 0.117 0.122 16.056 18.840 sum_up_and_integrate 127 10.3 0.002 0.004 18.239 18.255 integrate_v_rspace 127 11.3 0.003 0.004 18.181 18.199 qs_rho_update_rho_low 127 7.7 0.001 0.001 17.929 18.050 calculate_rho_elec 127 8.7 0.045 0.063 17.928 18.049 make_m2s 4970 13.5 0.102 0.109 16.405 16.787 make_images 4970 14.5 0.398 0.418 16.224 16.617 init_scf_run 11 5.9 0.000 0.001 13.671 13.673 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.671 13.672 cp_fm_cholesky_invert 11 10.9 12.565 12.576 12.565 12.576 density_rs2pw 127 9.7 0.006 0.007 11.470 12.098 pw_transfer 1535 11.6 0.074 0.088 10.341 10.967 mp_sum_l 7884 12.9 9.672 10.940 9.672 10.940 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 10.140 10.764 wfi_extrapolate 11 7.9 0.001 0.001 10.296 10.296 dbcsr_mm_accdrv_process 123452 16.2 4.734 4.856 9.387 9.920 multiply_cannon_metrocomm1 59640 15.5 0.096 0.101 7.021 9.163 fft3d_ps 1281 14.7 2.176 2.792 8.211 8.978 calculate_dm_sparse 127 9.5 0.001 0.001 8.691 8.811 make_images_data 4970 15.5 0.067 0.073 7.888 8.791 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.624 8.788 potential_pw2rs 127 12.3 0.008 0.010 8.377 8.738 qs_ot_get_orbitals 116 10.6 0.001 0.001 8.330 8.441 fft_wrap_pw1pw2_140 519 13.2 0.821 0.870 7.376 7.981 hybrid_alltoall_any 5155 16.4 0.290 2.266 6.769 7.840 grid_integrate_task_list 127 12.3 6.842 7.708 6.842 7.708 mp_alltoall_d11v 2401 14.1 5.685 7.465 5.685 7.465 cp_fm_cholesky_decompose 22 10.9 7.066 7.085 7.066 7.085 transfer_rs2pw 519 10.6 0.007 0.008 6.126 7.055 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.005 6.876 6.889 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.846 5.993 transfer_pw2rs 519 13.2 0.008 0.010 5.342 5.708 mp_alltoall_z22v 1281 16.7 4.731 5.281 4.731 5.281 grid_collocate_task_list 127 9.7 4.802 5.232 4.802 5.232 make_images_sizes 4970 15.5 0.006 0.007 3.197 5.221 mp_alltoall_i44 4970 16.5 3.190 5.215 3.190 5.215 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=250.152000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=599.272727, yerr=6.648184 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 842.326016E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 5446828. 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 0.014 0.036 199.447 199.449 qs_mol_dyn_low 1 2.0 0.003 0.003 199.053 199.070 qs_forces 11 3.9 0.004 0.005 198.935 198.937 qs_energies 11 4.9 0.002 0.003 192.065 192.075 scf_env_do_scf 11 5.9 0.001 0.001 175.317 175.327 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 140.995 140.999 velocity_verlet 10 3.0 0.002 0.002 124.324 124.327 qs_scf_new_mos 117 7.6 0.001 0.001 101.019 101.515 qs_scf_loop_do_ot 117 8.6 0.001 0.001 101.018 101.514 dbcsr_multiply_generic 2507 12.6 0.188 0.193 98.435 99.724 ot_scf_mini 117 9.6 0.004 0.004 96.163 96.754 multiply_cannon 2507 13.6 0.505 0.559 77.432 82.347 multiply_cannon_loop 2507 14.6 1.560 1.629 73.870 76.715 ot_mini 117 10.6 0.001 0.002 50.975 51.533 mp_waitall_1 214728 16.6 24.463 37.976 24.463 37.976 multiply_cannon_multrec 30084 15.6 21.108 25.076 31.771 36.069 init_scf_loop 11 6.9 0.000 0.000 34.220 34.221 rebuild_ks_matrix 128 8.3 0.001 0.001 33.151 33.787 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.024 33.150 33.786 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.821 30.404 prepare_preconditioner 11 7.9 0.000 0.000 29.813 29.877 make_preconditioner 11 8.9 0.000 0.000 29.813 29.877 qs_ot_get_derivative 117 11.6 0.001 0.002 28.988 29.565 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.411 28.983 multiply_cannon_metrocomm3 30084 15.6 0.100 0.108 15.395 28.910 qs_ot_get_p 128 10.4 0.001 0.002 27.086 27.774 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.121 23.212 apply_single 128 13.6 0.001 0.001 22.121 23.212 qs_ot_p2m_diag 83 11.4 0.189 0.217 21.779 21.816 ot_diis_step 117 11.6 0.014 0.015 21.812 21.814 cp_dbcsr_syevd 83 12.4 0.005 0.006 20.506 20.507 multiply_cannon_sync_h2d 30084 15.6 18.069 20.405 18.069 20.405 cp_fm_cholesky_invert 11 10.9 17.446 17.460 17.446 17.460 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.999 17.031 cp_fm_diag_elpa_base 83 14.4 16.716 16.838 16.991 17.019 make_m2s 5014 13.6 0.086 0.091 14.828 16.462 make_images 5014 14.6 1.166 1.371 14.614 16.252 sum_up_and_integrate 128 10.3 0.002 0.006 14.696 14.731 integrate_v_rspace 128 11.3 0.003 0.003 14.636 14.676 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.127 14.160 calculate_rho_elec 128 8.7 0.087 0.104 14.126 14.159 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.565 12.017 init_scf_run 11 5.9 0.000 0.001 11.852 11.854 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.852 11.853 multiply_cannon_metrocomm4 27577 15.6 0.099 0.113 3.823 10.859 make_images_data 5014 15.6 0.067 0.077 8.673 10.707 dbcsr_mm_accdrv_process 62242 16.2 5.437 6.384 10.117 10.686 mp_irecv_dv 69486 16.3 3.625 10.466 3.625 10.466 hybrid_alltoall_any 5200 16.5 0.352 1.553 7.335 9.891 pw_transfer 1547 11.6 0.085 0.093 9.177 9.238 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.953 9.012 wfi_extrapolate 11 7.9 0.001 0.001 8.606 8.606 density_rs2pw 128 9.7 0.006 0.007 7.700 8.309 fft_wrap_pw1pw2_140 523 13.2 0.935 0.963 7.721 7.792 grid_integrate_task_list 128 12.3 7.158 7.525 7.158 7.525 cp_fm_cholesky_decompose 22 10.9 7.344 7.418 7.344 7.418 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.423 7.195 mp_sum_l 7950 12.9 4.387 6.785 4.387 6.785 calculate_dm_sparse 128 9.5 0.001 0.001 6.529 6.670 fft3d_ps 1291 14.7 2.829 3.007 6.511 6.576 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.274 6.284 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.420 5.489 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.297 5.478 grid_collocate_task_list 128 9.7 5.006 5.429 5.006 5.429 mp_allgather_i34 2507 14.6 2.090 5.265 2.090 5.265 mp_alltoall_d11v 2415 14.1 4.661 5.264 4.661 5.264 potential_pw2rs 128 12.3 0.015 0.017 5.084 5.110 dbcsr_complete_redistribute 395 12.7 0.760 0.864 3.495 4.359 mp_sum_d 4472 12.1 2.683 4.025 2.683 4.025 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=199.449000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=802.272727, yerr=4.594283 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 987.308032E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172905552 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58192. MP_Allreduce 11085 1000. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.047 185.428 185.430 qs_mol_dyn_low 1 2.0 0.003 0.003 184.906 184.920 qs_forces 11 3.9 0.006 0.017 184.797 184.799 qs_energies 11 4.9 0.005 0.028 178.048 178.057 scf_env_do_scf 11 5.9 0.001 0.001 162.064 162.068 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 125.351 125.352 velocity_verlet 10 3.0 0.002 0.002 117.390 117.393 qs_scf_new_mos 116 7.6 0.001 0.001 87.133 87.452 qs_scf_loop_do_ot 116 8.6 0.001 0.001 87.132 87.451 dbcsr_multiply_generic 2485 12.5 0.184 0.187 82.372 83.385 ot_scf_mini 116 9.6 0.003 0.005 82.872 83.264 multiply_cannon 2485 13.5 0.497 0.515 61.941 66.793 multiply_cannon_loop 2485 14.5 1.118 1.183 58.784 61.130 ot_mini 116 10.6 0.001 0.001 43.057 43.433 init_scf_loop 11 6.9 0.001 0.008 36.601 36.604 mp_waitall_1 169034 16.6 24.928 33.728 24.928 33.728 prepare_preconditioner 11 7.9 0.000 0.000 32.472 32.533 make_preconditioner 11 8.9 0.000 0.002 32.472 32.533 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.004 31.484 rebuild_ks_matrix 127 8.3 0.001 0.001 30.886 31.344 qs_ks_build_kohn_sham_matrix 127 9.3 0.033 0.060 30.885 31.344 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.798 28.224 qs_ot_get_p 127 10.4 0.001 0.002 25.125 25.569 multiply_cannon_multrec 19880 15.5 13.042 15.892 22.356 25.143 multiply_cannon_metrocomm3 19880 15.5 0.062 0.066 15.044 24.135 qs_ot_get_derivative 116 11.6 0.001 0.002 23.611 24.007 qs_ot_p2m_diag 82 11.4 0.262 0.270 20.471 20.481 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.511 20.463 apply_single 127 13.6 0.001 0.001 19.511 20.462 cp_dbcsr_syevd 82 12.4 0.005 0.005 19.368 19.370 ot_diis_step 116 11.6 0.018 0.018 19.328 19.329 make_m2s 4970 13.5 0.077 0.081 15.185 16.441 make_images 4970 14.5 1.150 1.242 14.952 16.203 cp_fm_diag_elpa 82 13.4 0.000 0.000 15.783 15.806 cp_fm_diag_elpa_base 82 14.4 15.332 15.527 15.777 15.800 cp_fm_cholesky_invert 11 10.9 15.295 15.304 15.295 15.304 multiply_cannon_sync_h2d 19880 15.5 13.519 14.791 13.519 14.791 sum_up_and_integrate 127 10.3 0.003 0.014 14.465 14.488 integrate_v_rspace 127 11.3 0.003 0.004 14.405 14.431 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.255 14.293 calculate_rho_elec 127 8.7 0.128 0.142 14.255 14.292 init_scf_run 11 5.9 0.000 0.001 10.816 10.816 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.816 10.816 make_images_data 4970 15.5 0.064 0.074 8.974 10.679 hybrid_alltoall_any 5155 16.4 0.445 2.017 7.756 10.006 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.321 9.576 pw_transfer 1535 11.6 0.083 0.101 9.398 9.514 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.177 9.297 multiply_cannon_metrocomm4 17395 15.5 0.064 0.073 3.429 9.120 dbcsr_mm_accdrv_process 41158 16.2 5.609 5.933 8.774 8.955 mp_irecv_dv 49801 16.2 3.303 8.867 3.303 8.867 density_rs2pw 127 9.7 0.006 0.006 7.559 8.098 fft_wrap_pw1pw2_140 519 13.2 1.056 1.092 7.932 8.055 cp_fm_cholesky_decompose 22 10.9 7.804 7.839 7.804 7.839 wfi_extrapolate 11 7.9 0.001 0.001 7.727 7.727 grid_integrate_task_list 127 12.3 7.222 7.632 7.222 7.632 cp_fm_upper_to_full 104 14.8 5.855 7.387 5.855 7.387 dbcsr_complete_redistribute 393 12.7 1.162 1.199 4.987 6.793 fft3d_ps 1281 14.7 2.803 3.023 6.513 6.603 calculate_dm_sparse 127 9.5 0.001 0.001 5.878 5.972 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.577 5.582 grid_collocate_task_list 127 9.7 5.119 5.543 5.119 5.543 mp_alltoall_d11v 2401 14.1 4.867 5.533 4.867 5.533 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.657 5.455 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.740 5.408 mp_sum_l 7884 12.9 3.453 4.935 3.453 4.935 mp_allgather_i34 2485 14.5 1.682 4.868 1.682 4.868 potential_pw2rs 127 12.3 0.019 0.021 4.842 4.860 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.590 4.719 transfer_fm_to_dbcsr 11 9.9 0.018 0.022 2.448 4.208 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.975 4.001 mp_alltoall_i22 712 14.1 2.017 3.971 2.017 3.971 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.804 3.805 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=185.430000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=928.454545, yerr=21.980833 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.151607E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58205. MP_Allreduce 11082 1082. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.086 205.989 205.992 qs_mol_dyn_low 1 2.0 0.003 0.003 205.395 205.410 qs_forces 11 3.9 0.004 0.005 205.267 205.277 qs_energies 11 4.9 0.003 0.012 198.010 198.022 scf_env_do_scf 11 5.9 0.001 0.002 180.640 180.652 velocity_verlet 10 3.0 0.001 0.002 134.273 134.276 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 131.267 131.269 qs_scf_new_mos 117 7.6 0.001 0.001 93.049 93.350 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.049 93.349 ot_scf_mini 117 9.6 0.003 0.004 88.405 88.734 dbcsr_multiply_generic 2507 12.6 0.193 0.199 82.204 83.002 multiply_cannon 2507 13.6 0.541 0.567 55.913 59.190 multiply_cannon_loop 2507 14.6 1.843 1.912 51.985 53.925 init_scf_loop 11 6.9 0.001 0.006 49.225 49.226 prepare_preconditioner 11 7.9 0.000 0.000 44.956 44.990 make_preconditioner 11 8.9 0.000 0.002 44.955 44.990 ot_mini 117 10.6 0.001 0.001 44.311 44.622 make_full_inverse_cholesky 11 9.9 0.013 0.020 38.286 43.523 multiply_cannon_multrec 30084 15.6 13.794 18.770 26.700 31.354 rebuild_ks_matrix 128 8.3 0.001 0.001 30.499 30.804 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.028 30.499 30.804 qs_ot_get_p 128 10.4 0.001 0.002 30.000 30.313 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.578 27.833 mp_waitall_1 147882 16.7 17.880 27.491 17.880 27.491 qs_ot_p2m_diag 83 11.4 0.343 0.390 25.150 25.203 qs_ot_get_derivative 117 11.6 0.001 0.002 24.543 24.875 cp_dbcsr_syevd 83 12.4 0.005 0.007 23.649 23.652 make_m2s 5014 13.6 0.093 0.097 21.477 22.426 make_images 5014 14.6 1.950 2.275 21.164 22.113 cp_fm_diag_elpa 83 13.4 0.000 0.000 19.930 19.948 cp_fm_diag_elpa_base 83 14.4 18.874 19.238 19.920 19.936 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.069 19.628 apply_single 128 13.6 0.001 0.001 19.068 19.627 ot_diis_step 117 11.6 0.018 0.019 19.623 19.625 cp_fm_upper_to_full 105 14.8 11.863 17.520 11.863 17.520 cp_fm_cholesky_invert 11 10.9 17.449 17.459 17.449 17.459 sum_up_and_integrate 128 10.3 0.002 0.006 14.854 14.883 integrate_v_rspace 128 11.3 0.003 0.004 14.794 14.826 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.763 14.807 calculate_rho_elec 128 8.7 0.170 0.186 14.762 14.806 multiply_cannon_metrocomm3 30084 15.6 0.049 0.052 6.042 14.576 dbcsr_complete_redistribute 395 12.7 1.488 1.600 9.806 13.702 make_images_data 5014 15.6 0.066 0.072 11.511 13.111 dbcsr_mm_accdrv_process 62264 16.2 8.389 9.151 12.482 12.979 copy_fm_to_dbcsr 209 11.7 0.001 0.002 8.203 12.090 hybrid_alltoall_any 5200 16.5 0.531 2.202 9.998 12.073 multiply_cannon_sync_h2d 30084 15.6 10.443 11.492 10.443 11.492 init_scf_run 11 5.9 0.000 0.001 11.264 11.265 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.263 11.265 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.403 10.634 transfer_fm_to_dbcsr 11 9.9 0.001 0.009 6.648 10.470 pw_transfer 1547 11.6 0.086 0.102 10.049 10.130 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.823 9.911 mp_alltoall_i22 716 14.1 5.860 9.820 5.860 9.820 fft_wrap_pw1pw2_140 523 13.2 1.251 1.297 8.488 8.589 cp_fm_cholesky_decompose 22 10.9 8.359 8.453 8.359 8.453 wfi_extrapolate 11 7.9 0.001 0.001 8.115 8.115 density_rs2pw 128 9.7 0.006 0.006 7.708 8.060 grid_integrate_task_list 128 12.3 7.517 7.971 7.517 7.971 multiply_cannon_metrocomm4 25070 15.6 0.079 0.088 2.728 7.210 mp_irecv_dv 76098 16.2 2.580 6.936 2.580 6.936 fft3d_ps 1291 14.7 3.004 3.093 6.877 6.933 calculate_dm_sparse 128 9.5 0.001 0.001 6.355 6.425 mp_alltoall_d11v 2415 14.1 5.613 6.263 5.613 6.263 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.630 5.715 grid_collocate_task_list 128 9.7 5.287 5.620 5.287 5.620 potential_pw2rs 128 12.3 0.023 0.024 4.970 4.998 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.706 4.788 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.417 4.504 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.485 4.485 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.370 4.428 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=205.992000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1087.272727, yerr=16.023227 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.560408E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108638608 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58273. MP_Allreduce 11070 1167. MP_Sync 86 MP_Alltoall 1700 18828160. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.034 174.452 174.452 qs_mol_dyn_low 1 2.0 0.003 0.003 174.046 174.059 qs_forces 11 3.9 0.032 0.045 173.944 173.948 qs_energies 11 4.9 0.002 0.003 166.441 166.462 scf_env_do_scf 11 5.9 0.001 0.001 148.929 148.937 velocity_verlet 10 3.0 0.004 0.005 113.673 113.677 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 112.436 112.438 qs_scf_new_mos 116 7.6 0.001 0.001 75.092 75.244 qs_scf_loop_do_ot 116 8.6 0.001 0.001 75.092 75.243 dbcsr_multiply_generic 2485 12.5 0.183 0.191 72.621 73.177 ot_scf_mini 116 9.6 0.003 0.004 70.696 70.853 multiply_cannon 2485 13.5 0.552 0.574 53.465 56.740 multiply_cannon_loop 2485 14.5 0.802 0.843 50.392 51.515 ot_mini 116 10.6 0.001 0.001 38.142 38.313 init_scf_loop 11 6.9 0.000 0.000 36.328 36.329 prepare_preconditioner 11 7.9 0.000 0.000 32.435 32.480 make_preconditioner 11 8.9 0.000 0.000 32.435 32.479 mp_waitall_1 124680 16.7 25.417 32.388 25.417 32.388 make_full_inverse_cholesky 11 9.9 0.018 0.027 30.405 30.679 rebuild_ks_matrix 127 8.3 0.001 0.001 29.382 29.540 qs_ks_build_kohn_sham_matrix 127 9.3 0.028 0.054 29.382 29.539 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.762 26.909 multiply_cannon_multrec 9940 15.5 10.238 15.118 17.798 22.442 qs_ot_get_derivative 116 11.6 0.001 0.002 20.996 21.160 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 12.955 20.635 qs_ot_get_p 127 10.4 0.001 0.001 19.761 19.922 cp_fm_cholesky_invert 11 10.9 18.611 18.618 18.611 18.618 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.199 17.392 apply_single 127 13.6 0.001 0.001 17.199 17.392 ot_diis_step 116 11.6 0.021 0.040 17.072 17.073 make_m2s 4970 13.5 0.063 0.069 15.164 16.084 qs_ot_p2m_diag 82 11.4 0.489 0.495 15.844 15.860 make_images 4970 14.5 2.141 2.564 14.859 15.781 cp_dbcsr_syevd 82 12.4 0.005 0.006 14.644 14.645 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.389 14.423 calculate_rho_elec 127 8.7 0.251 0.262 14.388 14.422 sum_up_and_integrate 127 10.3 0.002 0.002 14.278 14.329 integrate_v_rspace 127 11.3 0.004 0.004 14.218 14.268 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.628 11.640 cp_fm_diag_elpa_base 82 14.4 11.379 11.463 11.618 11.630 multiply_cannon_sync_h2d 9940 15.5 10.576 11.273 10.576 11.273 init_scf_run 11 5.9 0.000 0.001 10.537 10.538 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.537 10.537 pw_transfer 1535 11.6 0.084 0.093 10.028 10.072 make_images_data 4970 15.5 0.055 0.067 8.763 10.049 hybrid_alltoall_any 5155 16.4 0.833 3.737 8.523 9.944 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 9.808 9.853 fft_wrap_pw1pw2_140 519 13.2 1.664 1.696 8.614 8.663 cp_fm_cholesky_decompose 22 10.9 8.161 8.310 8.161 8.310 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 8.189 8.308 grid_integrate_task_list 127 12.3 7.723 8.043 7.723 8.043 multiply_cannon_metrocomm1 9940 15.5 0.030 0.030 4.402 7.989 dbcsr_mm_accdrv_process 20590 16.1 3.238 4.156 7.213 7.908 wfi_extrapolate 11 7.9 0.001 0.001 7.485 7.485 density_rs2pw 127 9.7 0.005 0.005 7.099 7.416 fft3d_ps 1281 14.7 3.111 3.222 6.287 6.310 calculate_dm_sparse 127 9.5 0.001 0.001 6.050 6.131 multiply_cannon_metrocomm4 7455 15.5 0.025 0.028 1.854 5.850 grid_collocate_task_list 127 9.7 5.527 5.805 5.527 5.805 mp_irecv_dv 28618 15.9 1.814 5.763 1.814 5.763 dbcsr_complete_redistribute 393 12.7 2.110 2.181 5.299 5.611 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.380 5.398 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.254 5.263 mp_alltoall_d11v 2401 14.1 4.710 5.190 4.710 5.190 mp_allgather_i34 2485 14.5 1.289 4.883 1.289 4.883 potential_pw2rs 127 12.3 0.026 0.026 4.480 4.487 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.088 4.163 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.602 3.878 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.757 3.820 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.383 3.706 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.652 3.691 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.572 3.664 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.521 3.532 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.452000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1466.545455, yerr=14.932354 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.183251E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3949 59902. MP_Allreduce 10950 1507. MP_Sync 86 MP_Alltoall 1700 36954374. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.069 302.116 302.124 qs_mol_dyn_low 1 2.0 0.003 0.003 301.386 301.401 qs_forces 11 3.9 0.004 0.005 301.261 301.263 qs_energies 11 4.9 0.002 0.002 292.230 292.239 scf_env_do_scf 11 5.9 0.001 0.001 269.232 269.243 velocity_verlet 10 3.0 0.003 0.003 216.889 216.899 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 142.186 142.188 init_scf_loop 11 6.9 0.000 0.000 126.760 126.762 prepare_preconditioner 11 7.9 0.000 0.000 121.723 121.745 make_preconditioner 11 8.9 0.000 0.000 121.723 121.745 make_full_inverse_cholesky 11 9.9 0.038 0.039 98.117 118.744 qs_scf_new_mos 116 7.6 0.001 0.001 96.205 96.268 qs_scf_loop_do_ot 116 8.6 0.001 0.001 96.204 96.267 ot_scf_mini 116 9.6 0.004 0.004 91.232 91.274 dbcsr_multiply_generic 2485 12.5 0.216 0.227 84.515 85.121 cp_fm_upper_to_full 104 14.8 53.137 75.618 53.137 75.618 multiply_cannon 2485 13.5 0.685 0.728 60.601 62.084 multiply_cannon_loop 2485 14.5 1.049 1.079 56.406 57.878 ot_mini 116 10.6 0.001 0.001 46.109 46.162 dbcsr_complete_redistribute 393 12.7 3.964 4.019 29.786 42.578 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.128 38.930 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.563 36.280 rebuild_ks_matrix 127 8.3 0.001 0.001 35.677 35.715 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.018 35.677 35.715 cp_fm_cholesky_invert 11 10.9 34.987 34.994 34.987 34.994 mp_waitall_1 102768 16.8 30.001 34.515 30.001 34.515 mp_alltoall_i22 712 14.1 21.345 34.033 21.345 34.033 qs_ks_update_qs_env 127 7.6 0.001 0.001 33.048 33.089 qs_ot_get_p 127 10.4 0.001 0.001 29.658 29.735 qs_ot_get_derivative 116 11.6 0.002 0.002 26.028 26.065 qs_ot_p2m_diag 82 11.4 0.868 0.875 25.225 25.257 cp_dbcsr_syevd 82 12.4 0.006 0.006 23.346 23.347 multiply_cannon_metrocomm3 9940 15.5 0.025 0.025 20.506 21.902 make_m2s 4970 13.5 0.074 0.075 18.933 20.525 make_images 4970 14.5 3.036 3.220 18.455 20.045 ot_diis_step 116 11.6 0.022 0.022 20.033 20.034 cp_fm_diag_elpa 82 13.4 0.000 0.000 19.930 19.931 cp_fm_diag_elpa_base 82 14.4 15.495 17.067 19.923 19.924 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.328 19.449 apply_single 127 13.6 0.001 0.001 19.328 19.449 multiply_cannon_multrec 9940 15.5 10.271 12.095 18.047 18.182 qs_rho_update_rho_low 127 7.7 0.001 0.001 17.849 17.864 calculate_rho_elec 127 8.7 0.474 0.475 17.848 17.864 sum_up_and_integrate 127 10.3 0.002 0.002 17.000 17.089 integrate_v_rspace 127 11.3 0.004 0.004 16.938 17.027 multiply_cannon_sync_h2d 9940 15.5 14.187 14.220 14.187 14.220 pw_transfer 1535 11.6 0.094 0.095 13.802 13.809 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 13.566 13.572 hybrid_alltoall_any 5155 16.4 1.300 3.022 10.735 13.103 make_images_data 4970 15.5 0.063 0.070 10.627 13.007 init_scf_run 11 5.9 0.000 0.001 12.566 12.567 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.565 12.567 fft_wrap_pw1pw2_140 519 13.2 3.190 3.246 11.975 11.983 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.350 10.373 cp_fm_cholesky_decompose 22 10.9 9.850 9.864 9.850 9.864 wfi_extrapolate 11 7.9 0.001 0.001 9.390 9.391 dbcsr_mm_accdrv_process 20590 16.0 4.248 6.111 7.529 9.380 density_rs2pw 127 9.7 0.005 0.005 8.789 8.839 grid_integrate_task_list 127 12.3 8.510 8.713 8.510 8.713 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.216 8.217 fft3d_ps 1281 14.7 3.946 3.969 8.077 8.122 mp_alltoall_d11v 2401 14.1 7.529 7.656 7.529 7.656 calculate_dm_sparse 127 9.5 0.001 0.001 6.880 6.940 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.510 6.598 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.372 6.492 grid_collocate_task_list 127 9.7 6.372 6.450 6.372 6.450 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=302.124000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2883.909091, yerr=148.511680 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_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 1.260835E+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 28240833. 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 0.015 0.037 85.636 85.637 qs_energies 1 2.0 0.000 0.000 85.148 85.163 ls_scf 1 3.0 0.000 0.000 84.244 84.258 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.825 72.995 multiply_cannon 111 7.7 0.017 0.020 55.929 56.951 multiply_cannon_loop 111 8.7 0.228 0.246 52.551 53.951 ls_scf_main 1 4.0 0.000 0.000 52.659 52.664 density_matrix_trs4 2 5.0 0.002 0.003 47.015 47.095 ls_scf_init_scf 1 4.0 0.000 0.001 28.532 28.533 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.390 27.438 mp_waitall_1 11031 10.9 22.592 25.453 22.592 25.453 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.268 25.285 multiply_cannon_multrec 2664 9.7 8.160 8.938 15.614 17.406 multiply_cannon_sync_h2d 2664 9.7 13.569 15.640 13.569 15.640 make_m2s 222 7.7 0.009 0.011 13.249 13.639 make_images 222 8.7 0.098 0.108 13.227 13.617 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.744 12.289 make_images_data 222 9.7 0.004 0.005 7.793 8.303 dbcsr_mm_accdrv_process 4760 10.4 0.584 0.711 7.070 8.106 hybrid_alltoall_any 227 10.6 0.215 1.820 6.723 7.892 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.377 7.849 dbcsr_mm_accdrv_process_sort 4760 11.4 6.282 7.197 6.282 7.197 calculate_norms 4752 9.8 5.513 6.279 5.513 6.279 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.031 5.186 mp_sum_l 887 5.1 3.121 4.791 3.121 4.791 make_images_sizes 222 9.7 0.000 0.000 0.712 3.987 mp_alltoall_i44 222 10.7 0.711 3.987 0.711 3.987 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.066 3.676 mp_irecv_dv 6231 10.9 2.050 3.647 2.050 3.647 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.329 3.514 arnoldi_extremal 4 6.8 0.000 0.000 3.461 3.490 arnoldi_normal_ev 4 7.8 0.001 0.003 3.461 3.490 build_subspace 16 8.4 0.009 0.012 3.364 3.365 ls_scf_post 1 4.0 0.000 0.000 3.054 3.068 ls_scf_store_result 1 5.0 0.000 0.000 2.878 2.914 dbcsr_special_finalize 555 9.7 0.005 0.006 2.394 2.747 dbcsr_merge_single_wm 555 10.7 0.461 0.605 2.385 2.738 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.373 2.640 make_images_pack 222 9.7 2.204 2.623 2.206 2.624 dbcsr_sort_data 658 11.4 2.182 2.501 2.182 2.501 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.389 2.500 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.478 2.069 2.480 buffer_matrices_ensure_size 222 8.7 1.751 2.079 1.751 2.079 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.798 1.802 rebuild_ks_matrix 3 7.3 0.000 0.000 1.789 1.793 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.789 1.793 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.637000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1132.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_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 2.192609E+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 0.021 0.040 90.265 90.267 qs_energies 1 2.0 0.000 0.000 89.801 89.806 ls_scf 1 3.0 0.000 0.000 88.458 88.462 dbcsr_multiply_generic 111 6.7 0.015 0.015 74.396 74.780 multiply_cannon 111 7.7 0.028 0.039 52.698 57.230 ls_scf_main 1 4.0 0.000 0.000 54.683 54.688 multiply_cannon_loop 111 8.7 0.135 0.146 49.862 52.993 density_matrix_trs4 2 5.0 0.002 0.003 48.976 49.184 ls_scf_init_scf 1 4.0 0.000 0.001 30.069 30.070 mp_waitall_1 9105 10.9 20.605 29.640 20.605 29.640 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.863 28.957 multiply_cannon_multrec 1332 9.7 13.481 16.662 22.864 27.181 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.444 26.454 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.157 19.896 make_m2s 222 7.7 0.006 0.008 15.280 15.876 make_images 222 8.7 1.366 1.692 15.250 15.847 dbcsr_mm_accdrv_process 4041 10.4 0.358 0.540 8.979 10.578 dbcsr_mm_accdrv_process_sort 4041 11.4 8.497 10.039 8.497 10.039 make_images_data 222 9.7 0.004 0.004 8.914 9.900 hybrid_alltoall_any 227 10.6 0.542 2.574 8.274 9.666 mp_sum_l 887 5.1 4.956 7.842 4.956 7.842 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.188 7.740 mp_irecv_dv 3311 11.0 3.169 7.685 3.169 7.685 calculate_norms 2376 9.8 6.063 6.738 6.063 6.738 multiply_cannon_sync_h2d 1332 9.7 4.911 6.695 4.911 6.695 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.751 6.125 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.967 5.152 arnoldi_extremal 4 6.8 0.000 0.000 4.695 4.709 arnoldi_normal_ev 4 7.8 0.001 0.004 4.695 4.708 build_subspace 16 8.4 0.014 0.021 4.438 4.441 ls_scf_post 1 4.0 0.000 0.000 3.706 3.709 ls_scf_store_result 1 5.0 0.000 0.000 3.404 3.529 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.154 3.385 dbcsr_matrix_vector_mult_local 304 10.0 2.742 3.222 2.744 3.223 mp_allgather_i34 111 8.7 0.832 3.099 0.832 3.099 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.639 2.724 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.177 2.709 dbcsr_data_new 4174 10.1 2.113 2.406 2.113 2.406 make_images_pack 222 9.7 1.826 2.128 1.828 2.130 dbcsr_sort_data 436 11.2 1.822 2.096 1.822 2.096 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.869 1.871 rebuild_ks_matrix 3 7.3 0.000 0.000 1.856 1.858 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.856 1.858 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.267000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1779.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_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 2.843824E+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 0.023 0.043 92.982 92.984 qs_energies 1 2.0 0.000 0.000 92.441 92.445 ls_scf 1 3.0 0.000 0.000 91.020 91.024 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.589 75.907 ls_scf_main 1 4.0 0.000 0.000 56.989 56.993 multiply_cannon 111 7.7 0.035 0.098 52.319 56.585 multiply_cannon_loop 111 8.7 0.116 0.129 49.560 53.452 density_matrix_trs4 2 5.0 0.002 0.003 51.050 51.212 mp_waitall_1 7281 11.0 23.831 33.207 23.831 33.207 ls_scf_init_scf 1 4.0 0.000 0.001 30.392 30.396 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.161 29.258 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.799 26.813 multiply_cannon_multrec 888 9.7 12.749 15.202 21.393 24.452 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.193 22.782 make_m2s 222 7.7 0.006 0.007 16.506 17.243 make_images 222 8.7 1.583 1.862 16.468 17.203 make_images_data 222 9.7 0.004 0.005 9.650 10.676 hybrid_alltoall_any 227 10.6 0.640 2.930 9.167 10.294 dbcsr_mm_accdrv_process 3754 10.4 0.303 0.488 8.175 9.439 dbcsr_mm_accdrv_process_sort 3754 11.4 7.741 8.951 7.741 8.951 mp_sum_l 887 5.1 5.063 8.614 5.063 8.614 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.633 7.352 multiply_cannon_sync_h2d 888 9.7 6.053 7.234 6.053 7.234 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.458 7.176 mp_irecv_dv 2335 11.1 2.441 7.103 2.441 7.103 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.949 6.989 arnoldi_extremal 4 6.8 0.000 0.000 5.147 5.166 arnoldi_normal_ev 4 7.8 0.001 0.005 5.146 5.166 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.761 4.969 build_subspace 16 8.4 0.014 0.020 4.838 4.843 calculate_norms 1584 9.8 4.358 4.766 4.358 4.766 mp_allgather_i34 111 8.7 0.876 3.781 0.876 3.781 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.448 3.780 ls_scf_post 1 4.0 0.000 0.000 3.639 3.643 dbcsr_matrix_vector_mult_local 304 10.0 3.020 3.597 3.022 3.599 ls_scf_store_result 1 5.0 0.000 0.000 3.385 3.467 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.864 2.943 dbcsr_data_new 4116 9.9 2.108 2.455 2.108 2.455 make_images_sizes 222 9.7 0.000 0.000 1.022 2.334 mp_alltoall_i44 222 10.7 1.022 2.334 1.022 2.334 dbcsr_sort_data 325 11.1 1.882 2.162 1.882 2.162 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.906 1.907 rebuild_ks_matrix 3 7.3 0.000 0.000 1.888 1.890 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.888 1.890 make_images_pack 222 9.7 1.619 1.877 1.622 1.880 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.984000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2212.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_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 3.325678E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.049 97.083 97.086 qs_energies 1 2.0 0.000 0.000 96.502 96.508 ls_scf 1 3.0 0.000 0.000 94.826 94.831 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.595 78.883 ls_scf_main 1 4.0 0.000 0.000 58.954 58.956 multiply_cannon 111 7.7 0.046 0.099 51.784 56.455 density_matrix_trs4 2 5.0 0.002 0.003 52.772 52.918 multiply_cannon_loop 111 8.7 0.151 0.163 46.666 49.404 ls_scf_init_scf 1 4.0 0.000 0.001 32.642 32.645 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.424 31.494 mp_waitall_1 6369 11.0 22.734 29.074 22.734 29.074 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.933 28.948 multiply_cannon_multrec 1332 9.7 14.164 17.133 22.147 24.872 make_m2s 222 7.7 0.006 0.007 21.159 22.599 make_images 222 8.7 3.141 3.604 21.109 22.551 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.172 16.746 make_images_data 222 9.7 0.004 0.004 11.865 13.552 hybrid_alltoall_any 227 10.6 0.797 3.758 11.246 13.167 dbcsr_mm_accdrv_process 3641 10.4 0.292 0.474 7.622 9.172 dbcsr_mm_accdrv_process_sort 3641 11.4 7.183 8.686 7.183 8.686 mp_sum_l 887 5.1 4.071 7.581 4.071 7.581 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.089 6.122 multiply_cannon_sync_h2d 1332 9.7 5.468 6.049 5.468 6.049 mp_irecv_dv 3229 10.9 2.064 6.045 2.064 6.045 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.112 5.937 arnoldi_extremal 4 6.8 0.000 0.000 5.203 5.219 arnoldi_normal_ev 4 7.8 0.001 0.005 5.203 5.219 build_subspace 16 8.4 0.014 0.021 4.865 4.873 mp_allgather_i34 111 8.7 2.222 4.757 2.222 4.757 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.588 4.730 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.527 4.478 calculate_norms 2376 9.8 4.166 4.453 4.166 4.453 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.568 3.878 dbcsr_matrix_vector_mult_local 304 10.0 3.184 3.678 3.186 3.680 dbcsr_sort_data 658 11.4 3.085 3.408 3.085 3.408 ls_scf_post 1 4.0 0.000 0.000 3.230 3.235 dbcsr_special_finalize 555 9.7 0.006 0.007 2.839 3.181 dbcsr_merge_single_wm 555 10.7 0.541 0.665 2.830 3.173 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.034 3.092 ls_scf_store_result 1 5.0 0.000 0.000 2.974 3.034 dbcsr_data_release 10477 10.7 1.585 2.438 1.585 2.438 dbcsr_finalize 304 7.8 0.049 0.061 1.806 2.001 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.086000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2687.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_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 4.773134E+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 92.421 92.423 qs_energies 1 2.0 0.000 0.000 91.675 91.681 ls_scf 1 3.0 0.000 0.000 89.341 89.347 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.866 71.029 ls_scf_main 1 4.0 0.000 0.000 56.506 56.507 multiply_cannon 111 7.7 0.082 0.135 52.693 55.833 multiply_cannon_loop 111 8.7 0.089 0.095 50.118 51.201 density_matrix_trs4 2 5.0 0.002 0.003 49.479 49.557 ls_scf_init_scf 1 4.0 0.009 0.013 29.441 29.441 mp_waitall_1 5436 11.0 24.489 28.781 24.489 28.781 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.181 28.219 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.073 26.082 multiply_cannon_multrec 444 9.7 13.733 16.697 20.892 22.368 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.037 15.493 make_m2s 222 7.7 0.005 0.005 13.537 14.454 make_images 222 8.7 2.039 2.476 13.470 14.387 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.027 13.501 make_images_data 222 9.7 0.003 0.004 8.238 9.745 hybrid_alltoall_any 227 10.6 0.799 3.816 8.191 9.530 dbcsr_mm_accdrv_process 3003 10.4 0.350 0.406 6.850 7.970 multiply_cannon_sync_h2d 444 9.7 6.709 7.834 6.709 7.834 dbcsr_mm_accdrv_process_sort 3003 11.4 6.489 7.572 6.489 7.572 arnoldi_extremal 4 6.8 0.000 0.000 5.830 5.840 arnoldi_normal_ev 4 7.8 0.004 0.013 5.830 5.840 build_subspace 16 8.4 0.015 0.020 5.428 5.439 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.453 4.679 mp_sum_l 887 5.1 2.631 4.668 2.631 4.668 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.177 4.360 dbcsr_matrix_vector_mult_local 304 10.0 3.697 4.161 3.699 4.163 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.863 3.844 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.487 3.741 calculate_norms 792 9.8 3.622 3.723 3.622 3.723 mp_irecv_dv 1241 11.2 1.475 3.715 1.475 3.715 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.547 3.608 mp_allgather_i34 111 8.7 1.149 3.577 1.149 3.577 ls_scf_post 1 4.0 0.000 0.000 3.394 3.400 make_images_sizes 222 9.7 0.000 0.000 0.909 3.366 mp_alltoall_i44 222 10.7 0.909 3.365 0.909 3.365 ls_scf_store_result 1 5.0 0.000 0.000 3.184 3.219 qs_energies_init_hamiltonians 1 3.0 0.028 0.029 2.317 2.317 dbcsr_data_new 4608 9.7 1.792 2.284 1.792 2.284 dbcsr_finalize 304 7.8 0.062 0.077 2.195 2.275 dbcsr_merge_all 275 8.9 0.479 0.520 2.052 2.104 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.055 2.056 rebuild_ks_matrix 3 7.3 0.000 0.000 2.022 2.023 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.022 2.023 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.423000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3792.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_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 8.788083E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.071 0.092 107.177 107.179 qs_energies 1 2.0 0.000 0.000 105.791 105.802 ls_scf 1 3.0 0.000 0.000 102.629 102.639 dbcsr_multiply_generic 111 6.7 0.023 0.026 75.573 75.705 ls_scf_main 1 4.0 0.000 0.000 64.428 64.428 density_matrix_trs4 2 5.0 0.002 0.003 55.334 55.386 multiply_cannon 111 7.7 0.174 0.235 48.803 50.816 multiply_cannon_loop 111 8.7 0.098 0.099 45.784 46.611 ls_scf_init_scf 1 4.0 0.077 0.097 34.380 34.381 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.716 32.757 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.951 29.960 mp_waitall_1 4527 11.1 22.243 25.732 22.243 25.732 make_m2s 222 7.7 0.005 0.005 22.850 23.861 make_images 222 8.7 3.574 3.871 22.742 23.752 multiply_cannon_multrec 444 9.7 17.914 18.643 22.600 23.353 hybrid_alltoall_any 227 10.6 1.657 3.621 12.989 15.898 make_images_data 222 9.7 0.003 0.004 13.215 15.278 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.427 10.988 multiply_cannon_sync_h2d 444 9.7 8.787 8.850 8.787 8.850 arnoldi_extremal 4 6.8 0.000 0.000 7.527 7.567 arnoldi_normal_ev 4 7.8 0.035 0.074 7.527 7.567 build_subspace 16 8.4 0.026 0.036 6.937 6.953 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.526 5.691 dbcsr_matrix_vector_mult_local 304 10.0 5.045 5.363 5.047 5.366 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.232 5.329 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.883 5.133 dbcsr_mm_accdrv_process 1814 10.4 0.291 0.356 4.510 4.646 dbcsr_mm_accdrv_process_sort 1814 11.4 4.173 4.322 4.173 4.322 ls_scf_post 1 4.0 0.000 0.000 3.822 3.834 make_images_sizes 222 9.7 0.000 0.000 1.480 3.605 mp_alltoall_i44 222 10.7 1.480 3.605 1.480 3.605 ls_scf_store_result 1 5.0 0.000 0.000 3.544 3.577 calculate_norms 792 9.8 3.235 3.284 3.235 3.284 mp_allgather_i34 111 8.7 1.031 3.267 1.031 3.267 dbcsr_finalize 304 7.8 0.082 0.090 3.090 3.175 qs_energies_init_hamiltonians 1 3.0 0.028 0.074 3.131 3.131 dbcsr_merge_all 275 8.9 0.888 0.922 2.865 2.950 dbcsr_complete_redistribute 5 7.6 1.428 1.471 2.748 2.872 matrix_ls_to_qs 2 6.0 0.000 0.000 2.402 2.541 dbcsr_sort_data 325 11.1 2.444 2.503 2.444 2.503 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.490 2.492 rebuild_ks_matrix 3 7.3 0.000 0.000 2.423 2.426 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.423 2.426 dbcsr_data_new 6591 9.6 1.844 2.369 1.844 2.369 dbcsr_new_transposed 4 7.5 0.243 0.256 2.316 2.328 mp_sum_l 887 5.1 1.515 2.265 1.515 2.265 dbcsr_frobenius_norm 74 6.6 2.058 2.141 2.213 2.246 dbcsr_data_release 12724 10.6 1.999 2.225 1.999 2.225 dbcsr_add_d 103 6.2 0.000 0.000 2.140 2.209 dbcsr_add_anytype 103 7.2 0.859 0.892 2.140 2.209 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.179000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6969.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf_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 589.934592E+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 4805336. 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 0.032 0.120 243.467 243.472 qs_mol_dyn_low 1 2.0 0.003 0.007 242.361 242.384 qs_forces 5 3.8 0.005 0.013 242.287 242.292 qs_energies 5 4.8 0.003 0.014 239.124 239.147 scf_env_do_scf 5 5.8 0.000 0.001 224.909 224.916 scf_env_do_scf_inner_loop 105 6.6 0.002 0.006 196.511 196.516 qs_scf_new_mos 105 7.6 0.000 0.001 153.246 153.392 qs_scf_loop_do_ot 105 8.6 0.001 0.001 153.246 153.392 ot_scf_mini 105 9.6 0.003 0.009 143.277 143.431 dbcsr_multiply_generic 1445 12.2 0.128 0.135 134.051 134.573 multiply_cannon 1445 13.2 0.273 0.284 114.151 116.227 multiply_cannon_loop 1445 14.2 2.817 2.973 112.282 113.545 velocity_verlet 4 3.0 0.001 0.002 109.420 109.422 ot_mini 105 10.6 0.001 0.003 60.929 61.043 qs_ot_get_p 112 10.4 0.001 0.002 52.226 52.555 multiply_cannon_multrec 69360 15.2 29.674 34.845 39.478 44.894 mp_waitall_1 488190 16.1 35.857 43.385 35.857 43.385 qs_ot_p2m_diag 40 11.0 0.020 0.030 40.793 40.862 qs_ot_get_derivative 55 11.6 0.001 0.001 39.152 39.284 cp_dbcsr_syevd 40 12.0 0.002 0.002 37.417 37.419 multiply_cannon_metrocomm3 69360 15.2 0.202 0.212 26.006 33.469 multiply_cannon_sync_h2d 69360 15.2 29.052 33.061 29.052 33.061 cp_fm_syevd 40 13.0 0.000 0.000 32.116 32.248 rebuild_ks_matrix 110 8.4 0.000 0.000 30.846 31.016 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.025 30.846 31.016 qs_ks_update_qs_env 112 7.6 0.001 0.001 28.315 28.469 init_scf_loop 7 6.6 0.000 0.007 28.346 28.348 cp_fm_redistribute_end 40 14.0 13.606 27.154 13.612 27.157 cp_fm_syevd_base 40 14.0 13.534 27.084 13.534 27.084 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.121 23.439 apply_single 62 13.6 0.000 0.000 23.121 23.439 prepare_preconditioner 7 7.6 0.000 0.000 23.279 23.317 make_preconditioner 7 8.6 0.001 0.010 23.279 23.317 ot_new_cg_direction 55 11.6 0.001 0.007 21.106 21.106 qs_rho_update_rho_low 110 7.6 0.001 0.001 18.467 18.860 calculate_rho_elec 110 8.6 0.030 0.032 18.466 18.860 make_full_inverse_cholesky 7 9.6 0.000 0.000 15.579 15.645 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.805 15.002 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.973 14.080 pw_transfer 1645 12.4 0.078 0.105 13.316 13.491 mp_sum_l 4764 12.2 12.452 13.400 12.452 13.400 fft_wrap_pw1pw2 1425 13.5 0.012 0.016 13.179 13.357 density_rs2pw 110 9.6 0.005 0.006 12.499 13.061 fft_wrap_pw1pw2_240 915 15.0 0.813 0.896 11.678 11.855 calculate_dm_sparse 110 9.5 0.000 0.001 11.514 11.691 init_scf_run 5 5.8 0.000 0.001 11.505 11.506 scf_env_initial_rho_setup 5 6.8 0.001 0.003 11.504 11.505 qs_vxc_create 110 10.4 0.002 0.005 11.426 11.474 cp_fm_cholesky_invert 7 10.6 10.870 10.878 10.870 10.878 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.433 10.514 dbcsr_mm_accdrv_process 154766 15.8 6.183 6.375 9.672 10.482 fft3d_pb 915 16.0 2.374 2.579 9.921 10.110 check_diag 80 13.5 8.683 8.934 9.741 9.875 transfer_rs2pw 445 10.6 0.007 0.008 8.804 9.424 sum_up_and_integrate 60 10.3 0.001 0.004 9.229 9.242 integrate_v_rspace 60 11.3 0.001 0.002 9.212 9.226 xc_rho_set_and_dset_create 110 12.4 0.075 0.101 7.910 8.171 xc_vxc_pw_create 60 11.3 0.039 0.052 7.692 7.741 acc_transpose_blocks 69360 15.2 0.357 0.373 7.136 7.605 make_m2s 2890 13.2 0.078 0.086 7.014 7.577 make_images 2890 14.2 0.239 0.259 6.908 7.467 make_full_single_inverse 7 9.6 0.001 0.011 7.391 7.430 multiply_cannon_metrocomm1 69360 15.2 0.099 0.106 4.702 7.410 mp_alltoall_z22v 2340 17.7 7.028 7.294 7.028 7.294 xc_pw_derive 510 13.4 0.005 0.007 7.079 7.149 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 7.015 7.032 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.591 6.647 calculate_first_density_matrix 1 7.0 0.000 0.003 6.530 6.544 potential_pw2rs 60 12.3 0.002 0.003 5.316 5.345 mp_waitany 7680 13.5 4.601 5.139 4.601 5.139 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=243.472000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.000000, yerr=2.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 8485e0d524418ebd6cdaf3188a914e6b9a2d9ebf Summary: empty Status: OK