=== 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: f5d7817782ccd5e60ecd1b01248fcc43fdeed73e ################# 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 # # 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: 04.10.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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_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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/01 job id: 50387113 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/02 job id: 50387114 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/03 job id: 50387115 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/04 job id: 50387116 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/05 job id: 50387117 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/06 job id: 50387118 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/07 job id: 50387119 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/08 job id: 50387120 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/09 job id: 50387121 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/10 job id: 50387122 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/11 job id: 50387123 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/12 job id: 50387124 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/13 job id: 50387126 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/14 job id: 50387127 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/15 job id: 50387128 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/16 job id: 50387129 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/17 job id: 50387130 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/18 job id: 50387131 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/19 job id: 50387132 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/20 job id: 50387133 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/21 job id: 50387134 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/22 job id: 50387135 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/23 job id: 50387136 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/24 job id: 50387137 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/25 job id: 50387138 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/26 job id: 50387139 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/27 job id: 50387140 --- 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/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 177869. 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 0.032 144.360 144.361 farming_run 1 2.0 139.646 139.657 144.332 144.336 create_force_env 2 3.0 4.606 4.606 4.606 4.606 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.490604E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.027 115.104 115.104 qs_energies 1 2.0 0.000 0.000 114.846 114.849 mp2_main 1 3.0 0.000 0.000 112.738 112.741 mp2_gpw_main 1 4.0 0.038 0.044 111.702 111.705 mp2_ri_gpw_compute_in 1 5.0 0.172 0.177 92.802 93.115 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.652 54.967 mp2_eri_3c_integrate_gpw 272 7.0 0.150 0.162 41.057 46.095 get_2c_integrals 1 6.0 0.008 0.009 37.206 37.977 integrate_v_rspace 273 8.0 0.439 0.455 24.625 29.510 pw_transfer 6555 10.6 0.380 0.392 26.885 27.370 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 25.496 25.899 grid_integrate_task_list 273 9.0 20.523 25.889 20.523 25.889 fft_wrap_pw1pw2_100 2178 12.4 0.029 0.031 23.040 23.412 compute_2c_integrals 1 7.0 0.002 0.002 19.501 19.501 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.649 19.219 mp2_eri_2c_integrate_gpw 1 9.0 2.369 2.414 18.648 19.216 rpa_ri_compute_en 1 5.0 0.019 0.022 18.787 18.898 cp_fm_cholesky_decompose 12 8.2 17.644 18.438 17.644 18.438 cholesky_decomp 1 7.0 0.000 0.000 16.542 17.340 fft3d_s 5443 13.4 16.139 16.385 16.160 16.407 ao_to_mo_and_store_B_mult_1 272 7.0 10.770 15.321 10.770 15.321 calculate_wavefunction 272 8.0 5.408 5.568 12.326 13.003 rpa_num_int 1 6.0 0.000 0.001 10.647 10.657 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.624 10.633 calc_mat_Q 8 8.0 0.000 0.000 9.484 9.572 contract_S_to_Q 8 9.0 0.000 0.000 8.906 8.997 calc_potential_gpw 544 9.5 0.004 0.005 8.251 8.782 parallel_gemm_fm 14 9.1 0.000 0.000 8.484 8.576 parallel_gemm_fm_cosma 14 10.1 8.484 8.576 8.484 8.576 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.109 8.476 potential_pw2rs 545 10.0 0.106 0.107 7.526 8.172 create_integ_mat 1 6.0 0.015 0.029 7.733 7.743 collocate_single_gaussian 272 10.0 0.039 0.041 7.364 7.684 array2fm 1 7.0 0.000 0.000 6.824 7.198 pw_scatter_s 2720 13.7 4.344 4.561 4.344 4.561 pw_gather_s 2722 13.2 3.465 3.850 3.465 3.850 array2fm_buffer_send 1 8.0 3.052 3.173 3.052 3.173 pw_poisson_solve 545 10.5 1.127 1.192 2.219 2.477 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.702143, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2807.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 205321. 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 0.037 427.434 427.437 farming_run 1 2.0 426.493 426.502 427.386 427.389 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.249456E+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 0.011 0.034 217.358 217.359 qs_energies 1 2.0 0.013 0.053 217.103 217.139 scf_env_do_scf 1 3.0 0.000 0.000 114.799 114.799 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.920 113.928 rebuild_ks_matrix 4 6.0 0.000 0.000 113.918 113.927 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.062 113.918 113.927 hfx_ks_matrix 4 8.0 0.001 0.001 113.577 113.581 integrate_four_center 4 9.0 0.153 0.466 113.576 113.581 integrate_four_center_main 4 10.0 0.138 0.597 102.286 105.339 integrate_four_center_bin 262 11.0 102.149 105.257 102.149 105.257 mp2_main 1 3.0 0.008 0.060 101.999 102.012 mp2_gpw_main 1 4.0 0.065 0.149 101.055 101.064 init_scf_loop 1 4.0 0.000 0.000 96.925 96.925 mp2_ri_gpw_compute_in 1 5.0 0.087 0.160 73.953 75.027 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.011 53.784 54.854 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.160 41.174 46.409 integrate_v_rspace 95 8.0 0.395 0.569 27.648 32.713 pw_transfer 2240 10.6 0.146 0.158 29.412 29.795 mp2_ri_gpw_compute_en 1 5.0 0.064 0.133 26.882 29.019 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.453 28.829 grid_integrate_task_list 95 9.0 23.066 28.372 23.066 28.372 ao_to_mo_and_store_B_mult_1 91 7.0 10.942 27.581 10.942 27.581 fft_wrap_pw1pw2_100 730 12.4 0.012 0.014 26.220 26.600 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.832 1.879 25.082 25.091 get_2c_integrals 1 6.0 0.025 0.139 20.000 20.108 compute_2c_integrals 1 7.0 0.003 0.004 18.953 18.993 fft3d_s 1823 13.4 18.453 18.821 18.466 18.834 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.462 18.712 mp2_eri_2c_integrate_gpw 1 9.0 1.748 1.839 18.461 18.711 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.872 17.873 calculate_wavefunction 91 8.0 2.010 2.047 9.616 9.815 mp2_ri_gpw_compute_en_expansio 172 7.0 0.559 0.586 8.788 9.220 potential_pw2rs 186 10.0 0.033 0.034 8.377 8.807 local_gemm 172 8.0 8.230 8.654 8.230 8.654 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.519 8.071 8.527 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.000 0.001 8.055 8.275 calc_potential_gpw 182 9.5 0.002 0.002 7.850 8.036 collocate_single_gaussian 91 10.0 0.023 0.081 7.688 8.009 integrate_four_center_load 4 10.0 0.000 0.000 6.750 6.753 hfx_load_balance 1 11.0 0.000 0.000 6.750 6.753 mp_sendrecv_dm3 2068 8.0 6.107 6.579 6.107 6.579 mp2_ri_gpw_compute_en_ener 172 7.0 6.346 6.407 6.346 6.407 mp_sync 37 10.5 3.336 5.682 3.336 5.682 pw_gather_s 912 13.2 4.508 4.878 4.508 4.878 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.052062, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1508.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 453.734400E+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 485046. 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 0.011 0.033 53.395 53.397 qs_mol_dyn_low 1 2.0 0.003 0.005 52.778 53.164 qs_forces 11 3.9 0.009 0.061 52.685 52.694 qs_energies 11 4.9 0.002 0.002 51.165 51.178 scf_env_do_scf 11 5.9 0.000 0.001 44.946 44.946 scf_env_do_scf_inner_loop 108 6.5 0.004 0.015 42.888 42.888 qs_scf_new_mos 108 7.5 0.000 0.001 33.092 33.367 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.092 33.367 dbcsr_multiply_generic 2286 12.5 0.093 0.099 32.918 33.286 ot_scf_mini 108 9.5 0.002 0.003 31.480 31.665 multiply_cannon 2286 13.5 0.192 0.202 25.842 27.243 multiply_cannon_loop 2286 14.5 1.816 1.936 25.193 26.624 velocity_verlet 10 3.0 0.001 0.002 25.797 25.797 ot_mini 108 10.5 0.001 0.003 19.041 19.284 qs_ot_get_derivative 108 11.5 0.001 0.002 16.001 16.202 mp_waitall_1 245248 16.5 8.035 14.000 8.035 14.000 multiply_cannon_metrocomm3 54864 15.5 0.074 0.079 5.793 12.256 multiply_cannon_multrec 54864 15.5 3.654 5.818 7.600 10.996 qs_ot_get_p 119 10.4 0.001 0.002 7.848 8.127 rebuild_ks_matrix 119 8.3 0.000 0.000 7.755 7.910 qs_ks_build_kohn_sham_matrix 119 9.3 0.016 0.030 7.755 7.909 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.826 6.964 mp_sum_l 7287 12.8 5.220 6.711 5.220 6.711 multiply_cannon_sync_h2d 54864 15.5 5.207 6.029 5.207 6.029 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.537 5.997 dbcsr_mm_accdrv_process 76910 16.1 1.827 2.911 3.860 5.545 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.194 5.283 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.113 5.166 init_scf_run 11 5.9 0.000 0.001 4.934 4.934 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.934 4.934 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.387 4.387 sum_up_and_integrate 119 10.3 0.001 0.001 4.354 4.360 integrate_v_rspace 119 11.3 0.002 0.002 4.343 4.351 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.203 4.203 cp_fm_redistribute_end 50 14.0 2.145 4.180 2.150 4.182 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.015 4.127 calculate_rho_elec 119 8.7 0.011 0.017 4.015 4.126 cp_fm_diag_elpa_base 50 14.0 2.027 4.077 2.030 4.085 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.955 3.156 apply_single 119 13.6 0.000 0.000 2.954 3.155 calculate_dm_sparse 119 9.5 0.000 0.000 2.906 3.025 acc_transpose_blocks 54864 15.5 0.227 0.252 2.209 2.802 ot_diis_step 108 11.5 0.008 0.023 2.784 2.794 calculate_first_density_matrix 1 7.0 0.001 0.002 2.742 2.751 multiply_cannon_metrocomm1 54864 15.5 0.056 0.062 1.642 2.711 jit_kernel_multiply 13 15.8 1.968 2.565 1.968 2.565 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.377 2.444 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.198 2.200 density_rs2pw 119 9.7 0.004 0.004 2.057 2.153 wfi_extrapolate 11 7.9 0.001 0.001 2.118 2.118 grid_integrate_task_list 119 12.3 2.020 2.114 2.020 2.114 init_scf_loop 11 6.9 0.000 0.000 2.042 2.042 mp_sum_d 4137 12.0 1.379 2.013 1.379 2.013 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.925 1.987 potential_pw2rs 119 12.3 0.004 0.004 1.742 1.751 pw_transfer 1439 11.6 0.051 0.058 1.583 1.648 make_m2s 4572 13.5 0.053 0.056 1.549 1.594 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.507 1.575 make_images 4572 14.5 0.133 0.140 1.467 1.511 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.460 1.487 acc_transpose_blocks_sync 164592 16.5 1.188 1.443 1.188 1.443 mp_waitany 12084 13.8 1.235 1.431 1.235 1.431 grid_collocate_task_list 119 9.7 1.352 1.430 1.352 1.430 transfer_rs2pw 487 10.6 0.005 0.006 1.322 1.404 mp_alltoall_d11v 2130 13.8 1.242 1.403 1.242 1.403 transfer_pw2rs 487 13.2 0.006 0.007 1.281 1.289 fft3d_ps 1201 14.6 0.369 0.478 1.219 1.282 fft_wrap_pw1pw2_140 487 13.2 0.109 0.114 1.171 1.240 dbcsr_dot_sd 1205 11.9 0.048 0.059 0.766 1.184 parallel_gemm_fm 81 9.0 0.000 0.000 1.113 1.136 parallel_gemm_fm_cosma 81 10.0 1.113 1.135 1.113 1.135 acc_transpose_blocks_kernels 54864 16.5 0.247 0.385 0.751 1.071 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.397000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=432.363636, yerr=0.979121 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 489.250816E+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 769475. 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 0.059 0.113 40.002 40.002 qs_mol_dyn_low 1 2.0 0.003 0.004 39.389 39.397 qs_forces 11 3.9 0.002 0.003 39.329 39.331 qs_energies 11 4.9 0.006 0.019 37.633 37.638 scf_env_do_scf 11 5.9 0.002 0.013 32.356 32.356 scf_env_do_scf_inner_loop 108 6.5 0.005 0.028 29.849 29.850 dbcsr_multiply_generic 2286 12.5 0.101 0.104 22.134 22.491 qs_scf_new_mos 108 7.5 0.001 0.001 21.044 21.274 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.043 21.273 ot_scf_mini 108 9.5 0.003 0.004 20.114 20.282 velocity_verlet 10 3.0 0.002 0.008 18.562 18.563 multiply_cannon 2286 13.5 0.210 0.216 16.846 18.261 multiply_cannon_loop 2286 14.5 1.190 1.245 15.656 17.186 ot_mini 108 10.5 0.001 0.001 12.256 12.491 mp_waitall_1 200699 16.5 5.523 10.763 5.523 10.763 qs_ot_get_derivative 108 11.5 0.001 0.001 9.895 10.063 multiply_cannon_metrocomm3 27432 15.5 0.073 0.075 3.964 9.454 multiply_cannon_multrec 27432 15.5 1.826 4.063 6.152 9.006 rebuild_ks_matrix 119 8.3 0.000 0.000 6.862 7.003 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.022 6.861 7.003 dbcsr_mm_accdrv_process 47894 16.0 3.494 5.809 4.244 6.340 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.022 6.150 qs_ot_get_p 119 10.4 0.001 0.001 4.930 5.153 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.670 4.632 mp_sum_l 7287 12.8 2.216 4.242 2.216 4.242 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.109 4.104 apply_single 119 13.6 0.000 0.000 3.109 4.104 init_scf_run 11 5.9 0.000 0.001 3.975 3.976 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.975 3.976 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.762 3.787 calculate_rho_elec 119 8.7 0.021 0.024 3.761 3.787 sum_up_and_integrate 119 10.3 0.001 0.001 3.764 3.769 integrate_v_rspace 119 11.3 0.002 0.010 3.750 3.755 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.216 3.233 make_m2s 4572 13.5 0.052 0.054 2.674 3.008 make_images 4572 14.5 0.207 0.246 2.584 2.921 cp_dbcsr_syevd 50 12.0 0.012 0.073 2.815 2.815 init_scf_loop 11 6.9 0.001 0.004 2.472 2.473 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.342 2.432 calculate_first_density_matrix 1 7.0 0.000 0.003 2.423 2.426 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.373 2.374 cp_fm_redistribute_end 50 14.0 1.198 2.337 1.203 2.340 ot_diis_step 108 11.5 0.011 0.011 2.313 2.314 cp_fm_diag_elpa_base 50 14.0 1.102 2.237 1.132 2.273 multiply_cannon_sync_h2d 27432 15.5 1.695 2.168 1.695 2.168 calculate_dm_sparse 119 9.5 0.000 0.001 2.080 2.159 density_rs2pw 119 9.7 0.004 0.004 2.035 2.144 acc_transpose_blocks 27432 15.5 0.116 0.122 1.619 1.955 pw_transfer 1439 11.6 0.065 0.068 1.917 1.953 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.944 1.947 grid_integrate_task_list 119 12.3 1.841 1.919 1.841 1.919 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.826 1.864 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.693 1.733 make_images_data 4572 15.5 0.049 0.055 1.253 1.591 prepare_preconditioner 11 7.9 0.000 0.000 1.561 1.588 make_preconditioner 11 8.9 0.000 0.002 1.561 1.588 jit_kernel_multiply 8 16.3 0.693 1.569 0.693 1.569 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.459 1.516 potential_pw2rs 119 12.3 0.006 0.006 1.502 1.510 wfi_extrapolate 11 7.9 0.001 0.001 1.492 1.492 fft3d_ps 1201 14.6 0.522 0.576 1.457 1.488 hybrid_alltoall_any 4725 16.4 0.054 0.115 1.109 1.474 fft_wrap_pw1pw2_140 487 13.2 0.128 0.132 1.413 1.451 mp_alltoall_d11v 2130 13.8 1.260 1.379 1.260 1.379 grid_collocate_task_list 119 9.7 1.291 1.345 1.291 1.345 transfer_rs2pw 487 10.6 0.005 0.005 1.192 1.294 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.266 1.276 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.209 1.258 mp_sum_d 4137 12.0 0.573 1.057 0.573 1.057 mp_allgather_i34 2286 14.5 0.613 1.009 0.613 1.009 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.998 0.999 transfer_pw2rs 487 13.2 0.004 0.005 0.947 0.953 acc_transpose_blocks_sync 82296 16.5 0.811 0.938 0.811 0.938 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.899 0.913 acc_transpose_blocks_kernels 27432 16.5 0.188 0.277 0.667 0.885 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.002000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.727273, yerr=0.616575 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 520.523776E+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 0.112 0.131 43.178 43.188 qs_mol_dyn_low 1 2.0 0.003 0.003 39.575 39.592 qs_forces 11 3.9 0.085 0.116 39.516 39.517 qs_energies 11 4.9 0.002 0.002 37.879 37.913 scf_env_do_scf 11 5.9 0.000 0.001 25.789 25.789 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 23.366 23.366 dbcsr_multiply_generic 2286 12.5 0.096 0.101 21.154 21.242 multiply_cannon 2286 13.5 0.194 0.201 17.827 18.564 multiply_cannon_loop 2286 14.5 0.866 0.893 17.050 17.809 qs_scf_new_mos 108 7.5 0.001 0.001 15.646 15.659 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.645 15.659 velocity_verlet 10 3.0 0.003 0.004 15.219 15.220 ot_scf_mini 108 9.5 0.002 0.002 14.885 14.898 init_scf_run 11 5.9 0.000 0.001 9.912 9.912 scf_env_initial_rho_setup 11 6.9 0.304 0.499 9.911 9.911 ot_mini 108 10.5 0.001 0.001 9.225 9.236 calculate_first_density_matrix 1 7.0 0.000 0.001 8.391 8.585 qs_ot_get_derivative 108 11.5 0.001 0.001 7.727 7.741 multiply_cannon_multrec 18288 15.5 1.925 2.943 7.446 7.692 dbcsr_mm_accdrv_process 38222 16.0 5.329 6.291 5.428 6.357 rebuild_ks_matrix 119 8.3 0.000 0.000 5.991 6.004 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 5.990 6.004 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 5.884 5.886 acc_transpose_blocks 18288 15.5 0.080 0.083 5.397 5.466 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.423 5.428 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.280 5.292 acc_transpose_blocks_kernels 18288 16.5 0.218 0.226 4.555 4.576 jit_kernel_transpose 5 15.6 4.336 4.350 4.336 4.350 mp_waitall_1 158411 16.6 2.546 3.574 2.546 3.574 sum_up_and_integrate 119 10.3 0.001 0.001 3.559 3.564 integrate_v_rspace 119 11.3 0.002 0.003 3.546 3.553 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.931 3.527 qs_ot_get_p 119 10.4 0.001 0.002 3.410 3.431 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.289 3.296 calculate_rho_elec 119 8.7 0.031 0.031 3.288 3.295 qs_init_subsys 1 2.0 0.915 1.248 2.935 2.938 multiply_cannon_metrocomm3 18288 15.5 0.049 0.051 1.496 2.437 init_scf_loop 11 6.9 0.000 0.000 2.405 2.406 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.010 2.352 apply_single 119 13.6 0.000 0.000 2.010 2.351 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.278 2.285 calculate_dm_sparse 119 9.5 0.000 0.001 2.037 2.045 make_m2s 4572 13.5 0.044 0.046 1.870 2.018 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.980 1.980 density_rs2pw 119 9.7 0.004 0.004 1.860 1.966 grid_integrate_task_list 119 12.3 1.811 1.937 1.811 1.937 make_images 4572 14.5 0.193 0.206 1.784 1.932 pw_transfer 1439 11.6 0.065 0.068 1.862 1.870 qs_energies_init_hamiltonians 11 5.9 0.191 0.193 1.774 1.808 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.769 1.778 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.733 1.742 cp_fm_diag_elpa_base 50 14.0 1.710 1.721 1.732 1.740 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.730 1.738 prepare_preconditioner 11 7.9 0.000 0.000 1.647 1.650 make_preconditioner 11 8.9 0.000 0.001 1.647 1.650 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.506 1.588 calculate_atomic_block_dm 1 8.0 0.001 0.001 1.449 1.564 ot_diis_step 108 11.5 0.011 0.012 1.481 1.481 mp_sum_l 7287 12.8 1.079 1.479 1.079 1.479 qs_env_setup 1 3.0 0.000 0.000 1.417 1.475 qs_env_rebuild_pw_env 23 5.3 0.261 0.280 1.365 1.474 transfer_rs2pw 487 10.6 0.005 0.005 1.350 1.452 fft_wrap_pw1pw2_140 487 13.2 0.179 0.183 1.421 1.430 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.377 1.380 potential_pw2rs 119 12.3 0.007 0.008 1.362 1.365 grid_collocate_task_list 119 9.7 1.231 1.354 1.231 1.354 fft3d_ps 1201 14.6 0.546 0.569 1.324 1.333 calculate_atom 2 9.0 0.000 0.000 1.234 1.306 calculate_atom_restricted 2 10.0 1.025 1.090 1.234 1.306 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.154 1.204 multiply_cannon_sync_h2d 18288 15.5 1.050 1.171 1.050 1.171 wfi_extrapolate 11 7.9 0.001 0.001 1.166 1.166 pw_env_rebuild 1 5.0 0.000 0.000 0.937 0.995 transfer_rs2pw_140 130 11.5 0.297 0.428 0.866 0.965 make_images_data 4572 15.5 0.049 0.053 0.800 0.963 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.945 0.960 jit_kernel_multiply 4 16.0 0.041 0.953 0.041 0.953 mp_waitany 9880 13.7 0.694 0.910 0.694 0.910 hybrid_alltoall_any 4725 16.4 0.059 0.116 0.696 0.873 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=43.188000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.000000, yerr=1.477098 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 564.682752E+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 0.008 0.027 41.102 41.103 qs_mol_dyn_low 1 2.0 0.003 0.003 40.942 40.950 qs_forces 11 3.9 0.002 0.002 40.881 40.881 qs_energies 11 4.9 0.058 0.090 39.169 39.176 scf_env_do_scf 11 5.9 0.000 0.001 29.313 29.314 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.911 25.911 dbcsr_multiply_generic 2286 12.5 0.111 0.114 24.266 24.384 multiply_cannon 2286 13.5 0.218 0.226 20.476 20.961 multiply_cannon_loop 2286 14.5 1.526 1.614 19.514 19.934 velocity_verlet 10 3.0 0.001 0.002 18.499 18.501 qs_scf_new_mos 108 7.5 0.001 0.001 17.831 17.888 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.830 17.887 ot_scf_mini 108 9.5 0.002 0.003 16.799 16.858 ot_mini 108 10.5 0.001 0.001 10.303 10.369 multiply_cannon_multrec 27432 15.5 2.441 3.098 8.983 9.284 qs_ot_get_derivative 108 11.5 0.001 0.001 8.430 8.491 init_scf_run 11 5.9 0.000 0.001 8.354 8.355 scf_env_initial_rho_setup 11 6.9 0.007 0.010 8.354 8.354 dbcsr_mm_accdrv_process 47916 15.9 5.965 7.707 6.440 7.770 acc_transpose_blocks 27432 15.5 0.122 0.128 6.881 7.027 calculate_first_density_matrix 1 7.0 0.000 0.000 6.926 6.935 rebuild_ks_matrix 119 8.3 0.000 0.000 6.330 6.387 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.330 6.386 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 6.313 6.315 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.135 6.147 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.603 5.653 acc_transpose_blocks_kernels 27432 16.5 0.271 0.281 5.316 5.338 jit_kernel_transpose 5 15.6 5.045 5.064 5.045 5.064 qs_ot_get_p 119 10.4 0.001 0.002 3.654 3.732 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.058 3.515 sum_up_and_integrate 119 10.3 0.001 0.001 3.500 3.510 integrate_v_rspace 119 11.3 0.003 0.003 3.488 3.498 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.366 3.399 calculate_rho_elec 119 8.7 0.040 0.046 3.365 3.398 init_scf_loop 11 6.9 0.000 0.000 3.378 3.378 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.187 2.647 apply_single 119 13.6 0.000 0.000 2.187 2.647 prepare_preconditioner 11 7.9 0.000 0.000 2.547 2.554 make_preconditioner 11 8.9 0.000 0.001 2.546 2.554 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.154 2.481 make_m2s 4572 13.5 0.054 0.056 2.293 2.396 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.331 2.363 make_images 4572 14.5 0.275 0.336 2.185 2.287 mp_waitall_1 137007 16.6 1.678 2.251 1.678 2.251 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.226 2.237 calculate_dm_sparse 119 9.5 0.000 0.001 2.136 2.189 pw_transfer 1439 11.6 0.065 0.068 1.995 2.031 density_rs2pw 119 9.7 0.004 0.004 1.827 1.942 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.903 1.942 grid_integrate_task_list 119 12.3 1.824 1.921 1.824 1.921 ot_diis_step 108 11.5 0.012 0.013 1.833 1.833 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.822 1.822 fft_wrap_pw1pw2_140 487 13.2 0.227 0.235 1.569 1.609 mp_sum_l 7287 12.8 1.056 1.602 1.056 1.602 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.576 1.586 cp_fm_diag_elpa_base 50 14.0 1.542 1.560 1.574 1.584 acc_transpose_blocks_sync 82296 16.5 1.414 1.563 1.414 1.563 fft3d_ps 1201 14.6 0.592 0.653 1.391 1.417 wfi_extrapolate 11 7.9 0.001 0.001 1.367 1.367 grid_collocate_task_list 119 9.7 1.249 1.331 1.249 1.331 potential_pw2rs 119 12.3 0.009 0.009 1.310 1.315 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.286 1.304 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.237 1.247 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.687 1.226 cp_fm_upper_to_full 72 14.2 0.818 1.162 0.818 1.162 qs_energies_init_hamiltonians 11 5.9 0.003 0.005 1.146 1.147 jit_kernel_multiply 6 16.1 0.409 1.073 0.409 1.073 dbcsr_complete_redistribute 329 12.2 0.117 0.141 0.791 1.059 transfer_rs2pw 487 10.6 0.004 0.005 0.901 1.017 make_images_data 4572 15.5 0.050 0.054 0.850 0.965 hybrid_alltoall_any 4725 16.4 0.067 0.156 0.724 0.890 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.874 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.576 0.840 mp_alltoall_d11v 2130 13.8 0.739 0.832 0.739 0.832 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=41.103000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=534.636364, yerr=6.034337 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 618.475520E+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 0.049 0.052 32.881 32.882 qs_mol_dyn_low 1 2.0 0.003 0.003 32.497 32.506 qs_forces 11 3.9 0.002 0.002 32.432 32.433 qs_energies 11 4.9 0.002 0.002 30.711 30.713 scf_env_do_scf 11 5.9 0.000 0.001 22.018 22.018 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.545 19.546 dbcsr_multiply_generic 2286 12.5 0.096 0.100 16.984 17.055 multiply_cannon 2286 13.5 0.224 0.235 14.339 14.867 velocity_verlet 10 3.0 0.002 0.003 14.747 14.750 multiply_cannon_loop 2286 14.5 0.643 0.665 13.422 13.607 qs_scf_new_mos 108 7.5 0.001 0.001 11.843 11.864 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.842 11.863 ot_scf_mini 108 9.5 0.002 0.002 11.131 11.152 init_scf_run 11 5.9 0.000 0.001 7.261 7.261 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.261 7.261 ot_mini 108 10.5 0.001 0.001 6.342 6.368 multiply_cannon_multrec 9144 15.5 1.653 1.895 5.910 6.174 calculate_first_density_matrix 1 7.0 0.000 0.001 6.108 6.109 rebuild_ks_matrix 119 8.3 0.000 0.000 5.808 5.832 qs_ks_build_kohn_sham_matrix 119 9.3 0.015 0.018 5.807 5.831 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 5.556 5.557 acc_transpose_blocks 9144 15.5 0.044 0.045 5.393 5.431 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.345 5.355 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.163 5.184 qs_ot_get_derivative 108 11.5 0.001 0.001 5.054 5.073 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 4.617 4.622 jit_kernel_transpose 5 15.6 4.499 4.504 4.499 4.504 dbcsr_mm_accdrv_process 12550 15.8 3.560 4.171 4.147 4.218 sum_up_and_integrate 119 10.3 0.001 0.001 3.452 3.457 integrate_v_rspace 119 11.3 0.003 0.003 3.442 3.446 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.382 3.388 calculate_rho_elec 119 8.7 0.060 0.061 3.381 3.388 qs_ot_get_p 119 10.4 0.003 0.005 2.900 2.932 init_scf_loop 11 6.9 0.000 0.000 2.454 2.455 pw_transfer 1439 11.6 0.065 0.067 2.119 2.130 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.026 2.039 make_m2s 4572 13.5 0.033 0.035 1.809 1.955 grid_integrate_task_list 119 12.3 1.867 1.929 1.867 1.929 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.901 1.904 mp_waitall_1 115863 16.7 1.368 1.900 1.368 1.900 density_rs2pw 119 9.7 0.003 0.004 1.777 1.862 make_images 4572 14.5 0.269 0.300 1.720 1.862 calculate_dm_sparse 119 9.5 0.000 0.000 1.740 1.760 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.723 1.740 prepare_preconditioner 11 7.9 0.000 0.000 1.717 1.720 make_preconditioner 11 8.9 0.001 0.001 1.717 1.720 fft_wrap_pw1pw2_140 487 13.2 0.324 0.333 1.676 1.687 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.653 1.653 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.609 1.634 fft3d_ps 1201 14.6 0.649 0.664 1.379 1.389 grid_collocate_task_list 119 9.7 1.294 1.372 1.294 1.372 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.358 1.366 cp_fm_diag_elpa_base 50 14.0 1.331 1.349 1.357 1.365 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.353 1.363 ot_diis_step 108 11.5 0.013 0.013 1.274 1.274 potential_pw2rs 119 12.3 0.010 0.010 1.263 1.267 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.211 1.246 apply_single 119 13.6 0.000 0.000 1.210 1.246 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.219 1.221 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.123 1.129 wfi_extrapolate 11 7.9 0.001 0.001 1.103 1.103 jit_kernel_multiply 6 15.6 0.547 1.030 0.547 1.030 hybrid_alltoall_any 4725 16.4 0.066 0.180 0.777 0.987 make_images_data 4572 15.5 0.043 0.047 0.778 0.934 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.877 0.926 transfer_rs2pw 487 10.6 0.004 0.004 0.785 0.871 mp_alltoall_d11v 2130 13.8 0.788 0.864 0.788 0.864 cp_fm_cholesky_invert 11 10.9 0.858 0.861 0.858 0.861 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.831 0.838 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.766 0.769 acc_transpose_blocks_sync 27432 16.5 0.717 0.749 0.717 0.749 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.670 0.719 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.354 0.691 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=32.882000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=585.090909, yerr=6.868301 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 789.168128E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.032 47.157 47.158 qs_mol_dyn_low 1 2.0 0.003 0.003 46.941 46.950 qs_forces 11 3.9 0.028 0.041 46.878 46.879 qs_energies 11 4.9 0.038 0.038 44.864 44.871 scf_env_do_scf 11 5.9 0.001 0.001 34.492 34.492 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.758 26.760 velocity_verlet 10 3.0 0.002 0.002 23.945 23.951 dbcsr_multiply_generic 2286 12.5 0.103 0.105 23.179 23.421 multiply_cannon 2286 13.5 0.295 0.303 19.329 20.304 multiply_cannon_loop 2286 14.5 0.847 0.862 18.089 19.100 qs_scf_new_mos 108 7.5 0.001 0.001 17.091 17.190 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.090 17.189 ot_scf_mini 108 9.5 0.002 0.002 15.925 16.024 ot_mini 108 10.5 0.001 0.001 9.647 9.760 multiply_cannon_multrec 9144 15.5 3.357 4.607 8.751 8.859 init_scf_run 11 5.9 0.000 0.001 8.230 8.230 scf_env_initial_rho_setup 11 6.9 0.090 0.099 8.229 8.230 qs_ot_get_derivative 108 11.5 0.001 0.001 7.616 7.715 init_scf_loop 11 6.9 0.000 0.000 7.707 7.709 rebuild_ks_matrix 119 8.3 0.000 0.000 7.244 7.384 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.243 7.384 prepare_preconditioner 11 7.9 0.000 0.000 6.747 6.759 make_preconditioner 11 8.9 0.001 0.001 6.747 6.759 calculate_first_density_matrix 1 7.0 0.000 0.000 6.683 6.712 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.528 6.656 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.320 6.641 dbcsr_mm_accdrv_process 12550 15.8 4.529 5.919 5.257 6.550 acc_transpose_blocks 9144 15.5 0.045 0.047 6.247 6.272 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 6.255 6.255 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.005 6.043 acc_transpose_blocks_kernels 9144 16.5 0.120 0.122 5.110 5.112 jit_kernel_transpose 5 15.6 4.989 4.992 4.989 4.992 cp_fm_upper_to_full 72 14.2 3.164 4.539 3.164 4.539 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.199 4.205 calculate_rho_elec 119 8.7 0.117 0.120 4.198 4.204 sum_up_and_integrate 119 10.3 0.001 0.001 3.846 3.853 integrate_v_rspace 119 11.3 0.004 0.004 3.835 3.842 qs_ot_get_p 119 10.4 0.002 0.002 3.364 3.498 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.798 3.197 mp_waitall_1 94719 16.7 2.153 3.155 2.153 3.155 pw_transfer 1439 11.6 0.068 0.069 3.019 3.024 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.921 2.926 dbcsr_complete_redistribute 329 12.2 0.281 0.287 1.950 2.765 fft_wrap_pw1pw2_140 487 13.2 0.650 0.652 2.493 2.499 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.177 2.477 apply_single 119 13.6 0.000 0.000 2.177 2.477 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.650 2.460 make_m2s 4572 13.5 0.037 0.038 2.308 2.452 make_images 4572 14.5 0.356 0.390 2.188 2.332 calculate_dm_sparse 119 9.5 0.000 0.000 2.240 2.253 mp_alltoall_i22 627 13.8 1.400 2.244 1.400 2.244 density_rs2pw 119 9.7 0.003 0.003 2.223 2.243 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.421 2.223 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.271 2.199 grid_integrate_task_list 119 12.3 2.076 2.119 2.076 2.119 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.031 2.082 ot_diis_step 108 11.5 0.014 0.014 2.007 2.007 qs_ot_p2m_diag 50 11.0 0.043 0.043 1.990 1.992 fft3d_ps 1201 14.6 0.868 0.886 1.841 1.845 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.825 1.829 mp_sum_l 7287 12.8 0.990 1.735 0.990 1.735 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.637 1.638 grid_collocate_task_list 119 9.7 1.518 1.538 1.518 1.538 cp_fm_cholesky_invert 11 10.9 1.412 1.416 1.412 1.416 wfi_extrapolate 11 7.9 0.001 0.001 1.390 1.390 potential_pw2rs 119 12.3 0.014 0.014 1.358 1.359 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.353 1.353 cp_fm_diag_elpa_base 50 14.0 1.207 1.260 1.351 1.351 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.311 1.334 hybrid_alltoall_any 4725 16.4 0.092 0.151 1.047 1.250 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.237 1.243 mp_alltoall_d11v 2130 13.8 1.161 1.188 1.161 1.188 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.155 1.175 make_images_data 4572 15.5 0.048 0.051 0.960 1.140 acc_transpose_blocks_sync 27432 16.5 1.076 1.106 1.076 1.106 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.992 1.011 qs_create_task_list 11 7.9 0.011 0.017 0.944 0.958 generate_qs_task_list 11 8.9 0.367 0.387 0.933 0.945 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=47.158000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=734.454545, yerr=15.773186 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 502.534144E+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 493743. 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 0.031 0.080 86.653 86.658 qs_mol_dyn_low 1 2.0 0.003 0.003 86.042 86.118 qs_forces 11 3.9 0.003 0.004 85.919 85.920 qs_energies 11 4.9 0.002 0.003 83.023 83.045 scf_env_do_scf 11 5.9 0.000 0.001 70.013 70.016 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 64.594 64.594 dbcsr_multiply_generic 2055 12.4 0.104 0.107 55.410 55.722 multiply_cannon 2055 13.4 0.184 0.189 46.875 47.889 qs_scf_new_mos 99 7.5 0.000 0.001 47.517 47.643 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.516 47.643 multiply_cannon_loop 2055 14.4 1.773 1.815 45.919 46.980 ot_scf_mini 99 9.5 0.002 0.002 45.151 45.240 velocity_verlet 10 3.0 0.004 0.015 43.713 43.714 ot_mini 99 10.5 0.001 0.001 26.865 26.970 qs_ot_get_derivative 99 11.5 0.001 0.001 20.044 20.152 multiply_cannon_multrec 49320 15.4 11.370 12.038 17.583 18.266 rebuild_ks_matrix 110 8.3 0.000 0.000 14.537 14.708 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.024 14.537 14.708 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.721 12.873 mp_waitall_1 220248 16.4 10.936 11.898 10.936 11.898 init_scf_run 11 5.9 0.000 0.001 10.957 10.957 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.957 10.957 multiply_cannon_sync_h2d 49320 15.4 9.631 10.112 9.631 10.112 qs_ot_get_p 110 10.4 0.001 0.002 9.764 9.884 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.933 8.476 multiply_cannon_metrocomm3 49320 15.4 0.083 0.086 6.651 8.013 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 7.771 7.775 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.211 7.669 apply_single 110 13.6 0.000 0.001 7.211 7.668 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 7.426 7.467 sum_up_and_integrate 110 10.3 0.001 0.002 7.034 7.051 integrate_v_rspace 110 11.3 0.002 0.003 7.009 7.031 calculate_first_density_matrix 1 7.0 0.000 0.001 6.825 6.833 acc_transpose_blocks 49320 15.4 0.212 0.220 6.548 6.606 ot_diis_step 99 11.5 0.006 0.006 6.595 6.595 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.359 6.509 calculate_rho_elec 110 8.6 0.020 0.025 6.358 6.508 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.490 6.504 dbcsr_mm_accdrv_process 87628 16.1 3.040 3.119 6.082 6.336 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.692 5.692 init_scf_loop 11 6.9 0.000 0.000 5.390 5.391 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.265 5.314 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.121 5.146 cp_fm_diag_elpa_base 48 14.0 5.108 5.132 5.119 5.145 acc_transpose_blocks_kernels 49320 16.4 0.423 0.436 5.041 5.060 mp_sum_l 6594 12.7 4.172 4.974 4.172 4.974 jit_kernel_transpose 5 15.6 4.618 4.630 4.618 4.630 make_m2s 4110 13.4 0.060 0.065 3.927 4.051 wfi_extrapolate 11 7.9 0.001 0.001 4.027 4.027 make_images 4110 14.4 0.178 0.190 3.832 3.960 calculate_dm_sparse 110 9.5 0.001 0.001 3.735 3.821 density_rs2pw 110 9.6 0.004 0.004 3.375 3.582 grid_integrate_task_list 110 12.3 3.263 3.464 3.263 3.464 multiply_cannon_metrocomm1 49320 15.4 0.065 0.067 2.421 3.449 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.230 3.283 prepare_preconditioner 11 7.9 0.000 0.000 3.200 3.221 make_preconditioner 11 8.9 0.000 0.001 3.200 3.221 pw_transfer 1331 11.6 0.055 0.069 3.122 3.209 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.034 3.124 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.987 3.035 jit_kernel_multiply 13 15.9 2.761 2.783 2.761 2.783 fft_wrap_pw1pw2_140 451 13.1 0.310 0.332 2.585 2.685 potential_pw2rs 110 12.3 0.005 0.006 2.618 2.642 mp_alltoall_d11v 2046 13.8 2.098 2.588 2.098 2.588 fft3d_ps 1111 14.6 0.799 0.889 2.327 2.397 grid_collocate_task_list 110 9.6 2.159 2.306 2.159 2.306 transfer_rs2pw 451 10.6 0.005 0.006 1.901 2.116 mp_waitany 14300 13.8 1.820 2.048 1.820 2.048 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.945 1.963 mp_sum_d 3891 11.9 1.401 1.934 1.401 1.934 make_images_data 4110 15.4 0.045 0.048 1.782 1.904 hybrid_alltoall_any 4261 16.3 0.084 0.485 1.551 1.814 transfer_pw2rs 451 13.1 0.006 0.007 1.768 1.786 cp_fm_cholesky_invert 11 10.9 1.740 1.744 1.740 1.744 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=86.658000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.000000, yerr=2.923261 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 594.587648E+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 1315189. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.059 72.375 72.376 qs_mol_dyn_low 1 2.0 0.003 0.003 71.971 71.982 qs_forces 11 3.9 0.003 0.003 68.699 68.700 qs_energies 11 4.9 0.002 0.006 65.377 65.381 scf_env_do_scf 11 5.9 0.000 0.001 56.511 56.514 scf_env_do_scf_inner_loop 99 6.5 0.003 0.009 48.961 48.961 velocity_verlet 10 3.0 0.001 0.002 39.122 39.190 dbcsr_multiply_generic 2055 12.4 0.114 0.118 38.168 38.391 qs_scf_new_mos 99 7.5 0.001 0.001 33.143 33.283 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.143 33.282 multiply_cannon 2055 13.4 0.224 0.243 31.554 32.859 ot_scf_mini 99 9.5 0.003 0.003 31.474 31.605 multiply_cannon_loop 2055 14.4 1.165 1.200 30.289 31.276 ot_mini 99 10.5 0.001 0.001 18.409 18.539 multiply_cannon_multrec 24660 15.4 6.990 8.522 14.181 15.771 rebuild_ks_matrix 110 8.3 0.000 0.001 13.508 13.643 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.024 13.508 13.643 qs_ot_get_derivative 99 11.5 0.001 0.001 12.628 12.760 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.879 12.000 mp_waitall_1 176588 16.5 7.590 10.009 7.590 10.009 multiply_cannon_metrocomm3 24660 15.4 0.075 0.078 5.177 8.150 multiply_cannon_sync_h2d 24660 15.4 6.397 7.656 6.397 7.656 init_scf_loop 11 6.9 0.000 0.000 7.517 7.518 dbcsr_mm_accdrv_process 52282 16.1 5.554 6.569 7.019 7.346 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.470 7.167 apply_single 110 13.6 0.000 0.001 6.470 7.166 qs_ot_get_p 110 10.4 0.006 0.016 6.420 6.581 init_scf_run 11 5.9 0.000 0.001 6.489 6.490 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.489 6.489 sum_up_and_integrate 110 10.3 0.002 0.002 6.340 6.354 integrate_v_rspace 110 11.3 0.002 0.003 6.313 6.328 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.877 5.887 calculate_rho_elec 110 8.6 0.039 0.047 5.876 5.887 ot_diis_step 99 11.5 0.010 0.011 5.735 5.735 prepare_preconditioner 11 7.9 0.000 0.000 5.492 5.520 make_preconditioner 11 8.9 0.004 0.016 5.492 5.520 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.741 5.466 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.059 5.212 make_m2s 4110 13.4 0.057 0.059 4.221 4.681 make_images 4110 14.4 0.408 0.467 4.110 4.564 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.500 4.521 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.039 4.039 pw_transfer 1331 11.6 0.066 0.073 3.528 3.676 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.422 3.572 wfi_extrapolate 11 7.9 0.001 0.001 3.521 3.521 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.516 3.517 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.472 3.485 cp_fm_diag_elpa_base 48 14.0 3.425 3.445 3.470 3.481 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.289 3.360 mp_sum_dm 438 4.9 3.230 3.320 3.230 3.320 density_rs2pw 110 9.6 0.004 0.004 3.125 3.303 grid_integrate_task_list 110 12.3 3.165 3.300 3.165 3.300 md_write_output 11 3.9 0.068 3.230 0.069 3.257 md_output 10 3.0 0.000 0.000 0.071 3.257 update_particle_set 20 4.0 0.000 0.000 3.185 3.253 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.119 3.163 fft_wrap_pw1pw2_140 451 13.1 0.358 0.377 2.937 3.087 calculate_dm_sparse 110 9.5 0.001 0.001 2.973 2.999 calculate_first_density_matrix 1 7.0 0.000 0.001 2.872 2.874 hybrid_alltoall_any 4261 16.3 0.105 0.447 2.016 2.805 make_images_data 4110 15.4 0.050 0.054 2.301 2.776 fft3d_ps 1111 14.6 1.116 1.360 2.513 2.677 cp_fm_cholesky_invert 11 10.9 2.515 2.522 2.515 2.522 mp_sum_l 6594 12.7 1.769 2.424 1.769 2.424 grid_collocate_task_list 110 9.6 2.170 2.292 2.170 2.292 potential_pw2rs 110 12.3 0.008 0.009 2.212 2.223 jit_kernel_multiply 10 16.2 1.096 2.063 1.096 2.063 acc_transpose_blocks 24660 15.4 0.118 0.121 1.985 2.047 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.961 1.980 mp_alltoall_d11v 2046 13.8 1.768 1.936 1.768 1.936 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.868 1.869 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.785 1.800 mp_allgather_i34 2055 14.4 0.574 1.588 0.574 1.588 transfer_rs2pw 451 10.6 0.006 0.007 1.434 1.586 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.538 1.550 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.386 1.491 dbcsr_complete_redistribute 325 12.2 0.252 0.334 1.202 1.463 multiply_cannon_metrocomm4 22605 15.4 0.074 0.080 0.752 1.456 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=72.376000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=561.454545, yerr=7.548192 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 667.308032E+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 0.132 0.189 62.907 62.955 qs_mol_dyn_low 1 2.0 0.003 0.004 60.546 61.831 qs_forces 11 3.9 0.015 0.018 60.351 60.353 qs_energies 11 4.9 0.008 0.051 57.134 57.173 scf_env_do_scf 11 5.9 0.001 0.003 48.675 48.676 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.127 40.128 velocity_verlet 10 3.0 0.001 0.002 32.622 32.626 dbcsr_multiply_generic 2055 12.4 0.110 0.113 28.741 29.008 qs_scf_new_mos 99 7.5 0.001 0.001 25.428 25.521 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.428 25.521 ot_scf_mini 99 9.5 0.002 0.003 24.175 24.283 multiply_cannon 2055 13.4 0.211 0.220 22.507 23.726 multiply_cannon_loop 2055 14.4 0.818 0.851 21.329 22.346 ot_mini 99 10.5 0.001 0.001 13.688 13.796 rebuild_ks_matrix 110 8.3 0.000 0.000 12.172 12.291 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.017 12.172 12.291 multiply_cannon_multrec 16440 15.4 3.695 4.692 9.922 10.852 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.716 10.821 mp_waitall_1 139946 16.5 6.809 10.044 6.809 10.044 qs_ot_get_derivative 99 11.5 0.001 0.001 9.268 9.376 init_scf_loop 11 6.9 0.000 0.001 8.512 8.514 multiply_cannon_metrocomm3 16440 15.4 0.046 0.047 4.233 7.288 prepare_preconditioner 11 7.9 0.000 0.000 6.743 6.762 make_preconditioner 11 8.9 0.000 0.000 6.743 6.762 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.085 6.427 sum_up_and_integrate 110 10.3 0.002 0.003 6.317 6.333 integrate_v_rspace 110 11.3 0.003 0.003 6.291 6.308 dbcsr_mm_accdrv_process 34862 16.1 5.379 5.791 6.070 6.212 init_scf_run 11 5.9 0.000 0.001 5.883 5.883 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.883 5.883 qs_ot_get_p 110 10.4 0.001 0.003 5.647 5.811 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.780 5.787 calculate_rho_elec 110 8.6 0.059 0.059 5.780 5.787 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.915 5.385 apply_single 110 13.6 0.000 0.000 4.914 5.384 make_m2s 4110 13.4 0.050 0.052 4.175 4.557 make_images 4110 14.4 0.401 0.525 4.058 4.436 ot_diis_step 99 11.5 0.011 0.011 4.390 4.390 qs_ot_p2m_diag 48 11.0 0.041 0.044 3.981 3.986 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.163 3.806 multiply_cannon_sync_h2d 16440 15.4 3.244 3.772 3.244 3.772 pw_transfer 1331 11.6 0.065 0.070 3.622 3.637 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.610 3.610 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.515 3.533 grid_integrate_task_list 110 12.3 3.196 3.435 3.196 3.435 density_rs2pw 110 9.6 0.004 0.004 3.013 3.156 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.053 3.063 cp_fm_diag_elpa_base 48 14.0 2.986 3.016 3.051 3.061 fft_wrap_pw1pw2_140 451 13.1 0.461 0.472 3.029 3.048 wfi_extrapolate 11 7.9 0.001 0.001 2.971 2.971 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.873 2.875 calculate_first_density_matrix 1 7.0 0.000 0.002 2.822 2.824 make_images_data 4110 15.4 0.048 0.052 2.317 2.811 hybrid_alltoall_any 4261 16.3 0.109 0.384 2.125 2.792 cp_fm_cholesky_invert 11 10.9 2.630 2.636 2.630 2.636 calculate_dm_sparse 110 9.5 0.001 0.001 2.568 2.605 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.480 2.525 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.440 2.510 mp_sum_l 6594 12.7 1.771 2.453 1.771 2.453 fft3d_ps 1111 14.6 1.118 1.131 2.432 2.446 grid_collocate_task_list 110 9.6 2.221 2.414 2.221 2.414 multiply_cannon_metrocomm4 14385 15.4 0.048 0.053 0.874 2.348 mp_irecv_dv 48980 15.7 0.798 2.212 0.798 2.212 potential_pw2rs 110 12.3 0.011 0.011 2.155 2.159 qs_energies_init_hamiltonians 11 5.9 0.013 0.016 2.099 2.112 mp_alltoall_d11v 2046 13.8 1.816 2.056 1.816 2.056 dbcsr_complete_redistribute 325 12.2 0.368 0.416 1.506 1.961 cp_fm_upper_to_full 70 14.2 1.393 1.752 1.393 1.752 acc_transpose_blocks 16440 15.4 0.080 0.083 1.628 1.693 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.649 1.663 cp_fm_cholesky_decompose 22 10.9 1.590 1.614 1.590 1.614 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.500 1.513 transfer_rs2pw 451 10.6 0.005 0.005 1.335 1.500 mp_allgather_i34 2055 14.4 0.492 1.475 0.492 1.475 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.031 1.471 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.364 1.461 md_write_output 11 3.9 0.045 1.417 0.047 1.457 md_output 10 3.0 0.000 0.000 0.085 1.456 mp_max_l 33 2.8 1.284 1.364 1.284 1.364 mp_waitany 17072 13.8 1.134 1.301 1.134 1.301 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.274 1.284 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.169 1.264 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=62.955000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=632.727273, yerr=8.378445 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 732.282880E+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 0.033 0.065 73.736 73.746 qs_mol_dyn_low 1 2.0 0.003 0.003 69.673 73.383 qs_forces 11 3.9 0.003 0.003 69.358 69.360 qs_energies 11 4.9 0.002 0.002 65.976 65.981 scf_env_do_scf 11 5.9 0.000 0.001 53.607 53.610 scf_env_do_scf_inner_loop 99 6.5 0.009 0.015 42.035 42.035 velocity_verlet 10 3.0 0.002 0.002 36.970 36.975 dbcsr_multiply_generic 2055 12.4 0.116 0.121 34.626 34.862 multiply_cannon 2055 13.4 0.237 0.251 27.621 28.388 qs_scf_new_mos 99 7.5 0.001 0.001 27.529 27.650 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.528 27.650 multiply_cannon_loop 2055 14.4 1.416 1.484 26.230 26.875 ot_scf_mini 99 9.5 0.002 0.003 25.847 25.943 ot_mini 99 10.5 0.001 0.001 14.986 15.112 multiply_cannon_multrec 24660 15.4 4.078 6.687 13.066 14.319 rebuild_ks_matrix 110 8.3 0.000 0.000 11.877 11.983 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.877 11.983 init_scf_loop 11 6.9 0.000 0.000 11.532 11.533 qs_ot_get_derivative 99 11.5 0.001 0.001 10.816 10.922 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.493 10.589 dbcsr_mm_accdrv_process 52304 16.0 7.807 9.368 8.830 9.967 prepare_preconditioner 11 7.9 0.000 0.000 9.703 9.718 make_preconditioner 11 8.9 0.000 0.001 9.703 9.718 init_scf_run 11 5.9 0.000 0.001 9.643 9.643 scf_env_initial_rho_setup 11 6.9 0.001 0.001 9.642 9.643 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.978 9.400 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 6.748 6.749 calculate_first_density_matrix 1 7.0 0.000 0.000 6.591 6.593 acc_transpose_blocks 24660 15.4 0.114 0.116 6.312 6.453 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.383 6.422 sum_up_and_integrate 110 10.3 0.001 0.002 6.176 6.189 integrate_v_rspace 110 11.3 0.003 0.003 6.150 6.164 mp_waitall_1 121746 16.5 4.164 6.102 4.164 6.102 qs_ot_get_p 110 10.4 0.006 0.008 5.886 6.036 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.808 5.819 calculate_rho_elec 110 8.6 0.078 0.081 5.807 5.819 make_m2s 4110 13.4 0.059 0.061 5.278 5.601 make_images 4110 14.4 0.578 0.700 5.137 5.456 cp_fm_upper_to_full 70 14.2 3.307 4.785 3.307 4.785 acc_transpose_blocks_kernels 24660 16.4 0.308 0.315 4.762 4.773 jit_kernel_transpose 5 15.6 4.454 4.458 4.454 4.458 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.138 4.351 apply_single 110 13.6 0.000 0.000 4.138 4.351 ot_diis_step 99 11.5 0.011 0.011 4.132 4.132 md_output 10 3.0 0.000 0.000 0.170 4.001 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.925 3.944 mp_max_l 33 2.8 3.711 3.873 3.711 3.873 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.671 3.802 pw_transfer 1331 11.6 0.065 0.074 3.737 3.772 dbcsr_complete_redistribute 325 12.2 0.412 0.465 2.624 3.739 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.631 3.667 grid_integrate_task_list 110 12.3 3.303 3.453 3.303 3.453 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.345 3.402 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.362 3.363 multiply_cannon_metrocomm3 24660 15.4 0.037 0.039 1.441 3.303 md_write_output 11 3.9 0.135 3.225 0.138 3.278 fft_wrap_pw1pw2_140 451 13.1 0.531 0.543 3.135 3.175 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.072 3.169 density_rs2pw 110 9.6 0.004 0.004 2.941 3.065 calculate_dm_sparse 110 9.5 0.001 0.001 2.974 3.008 wfi_extrapolate 11 7.9 0.001 0.001 2.957 2.957 hybrid_alltoall_any 4261 16.3 0.124 0.460 2.279 2.892 make_images_data 4110 15.4 0.050 0.055 2.558 2.882 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.820 2.829 cp_fm_diag_elpa_base 48 14.0 2.676 2.735 2.818 2.827 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.716 2.800 mp_alltoall_i22 605 13.7 1.629 2.761 1.629 2.761 cp_fm_cholesky_invert 11 10.9 2.615 2.623 2.615 2.623 multiply_cannon_sync_h2d 24660 15.4 2.394 2.532 2.394 2.532 fft3d_ps 1111 14.6 1.154 1.192 2.456 2.469 grid_collocate_task_list 110 9.6 2.265 2.433 2.265 2.433 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.238 2.240 potential_pw2rs 110 12.3 0.012 0.013 2.012 2.017 mp_alltoall_d11v 2046 13.8 1.692 1.899 1.692 1.899 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.798 1.832 mp_sum_l 6594 12.7 1.053 1.817 1.053 1.817 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.610 1.711 cp_fm_cholesky_decompose 22 10.9 1.596 1.654 1.596 1.654 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.625 1.639 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.533 1.555 acc_transpose_blocks_sync 73980 16.4 1.395 1.535 1.395 1.535 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=73.746000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.727273, yerr=8.012387 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 865.857536E+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 0.019 0.043 59.893 59.898 qs_mol_dyn_low 1 2.0 0.003 0.003 59.593 59.611 qs_forces 11 3.9 0.007 0.008 59.527 59.527 qs_energies 11 4.9 0.016 0.018 55.878 55.883 scf_env_do_scf 11 5.9 0.000 0.001 43.105 43.105 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.412 35.413 velocity_verlet 10 3.0 0.002 0.002 31.122 31.125 dbcsr_multiply_generic 2055 12.4 0.107 0.109 27.684 27.895 multiply_cannon 2055 13.4 0.239 0.249 22.130 23.495 multiply_cannon_loop 2055 14.4 0.606 0.620 20.875 21.203 qs_scf_new_mos 99 7.5 0.001 0.001 20.773 20.830 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.772 20.830 ot_scf_mini 99 9.5 0.002 0.002 19.526 19.579 rebuild_ks_matrix 110 8.3 0.000 0.000 11.663 11.723 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.020 11.663 11.723 ot_mini 99 10.5 0.001 0.001 10.842 10.895 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.369 10.423 init_scf_run 11 5.9 0.000 0.001 9.559 9.560 scf_env_initial_rho_setup 11 6.9 0.044 0.068 9.559 9.559 multiply_cannon_multrec 8220 15.4 3.151 4.397 7.528 8.731 mp_waitall_1 103326 16.6 5.905 7.655 5.905 7.655 init_scf_loop 11 6.9 0.000 0.000 7.647 7.648 qs_ot_get_derivative 99 11.5 0.001 0.001 7.107 7.158 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 6.925 6.927 calculate_first_density_matrix 1 7.0 0.000 0.000 6.725 6.769 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.580 6.600 sum_up_and_integrate 110 10.3 0.001 0.002 6.219 6.233 integrate_v_rspace 110 11.3 0.003 0.003 6.192 6.206 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.036 6.048 calculate_rho_elec 110 8.6 0.113 0.113 6.035 6.047 prepare_preconditioner 11 7.9 0.000 0.000 6.004 6.014 make_preconditioner 11 8.9 0.000 0.001 6.004 6.014 acc_transpose_blocks 8220 15.4 0.042 0.043 5.838 5.863 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.579 5.653 dbcsr_mm_accdrv_process 17442 15.9 3.069 4.082 4.238 5.162 acc_transpose_blocks_kernels 8220 16.4 0.114 0.117 5.116 5.123 jit_kernel_transpose 5 15.6 5.001 5.008 5.001 5.008 qs_ot_get_p 110 10.4 0.001 0.002 4.735 4.790 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 3.150 4.517 make_m2s 4110 13.4 0.037 0.039 4.190 4.433 make_images 4110 14.4 0.647 0.708 4.059 4.300 pw_transfer 1331 11.6 0.066 0.071 4.086 4.099 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.978 3.995 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.776 3.818 apply_single 110 13.6 0.000 0.000 3.775 3.818 ot_diis_step 99 11.5 0.012 0.012 3.712 3.712 grid_integrate_task_list 110 12.3 3.381 3.475 3.381 3.475 fft_wrap_pw1pw2_140 451 13.1 0.719 0.734 3.456 3.470 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.378 3.382 density_rs2pw 110 9.6 0.004 0.004 3.013 3.123 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.040 3.041 cp_fm_cholesky_invert 11 10.9 2.869 2.873 2.869 2.873 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.748 2.752 wfi_extrapolate 11 7.9 0.001 0.001 2.694 2.694 hybrid_alltoall_any 4261 16.3 0.201 0.850 2.233 2.634 make_images_data 4110 15.4 0.042 0.048 2.224 2.553 fft3d_ps 1111 14.6 1.290 1.299 2.522 2.537 grid_collocate_task_list 110 9.6 2.362 2.533 2.362 2.533 multiply_cannon_sync_h2d 8220 15.4 2.344 2.515 2.344 2.515 calculate_dm_sparse 110 9.5 0.001 0.001 2.465 2.509 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.482 2.490 cp_fm_diag_elpa_base 48 14.0 2.427 2.448 2.480 2.488 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.027 2.073 potential_pw2rs 110 12.3 0.015 0.015 2.009 2.011 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.785 1.994 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.923 1.953 mp_alltoall_d11v 2046 13.8 1.730 1.901 1.730 1.901 cp_fm_cholesky_decompose 22 10.9 1.677 1.700 1.677 1.700 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.568 1.691 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.620 1.627 dbcsr_complete_redistribute 325 12.2 0.550 0.566 1.406 1.492 mp_allgather_i34 2055 14.4 0.427 1.479 0.427 1.479 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.478 1.450 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.420 1.430 mp_irecv_dv 24056 15.7 0.451 1.401 0.451 1.401 jit_kernel_multiply 8 15.5 0.848 1.368 0.848 1.368 qs_create_task_list 11 7.9 0.009 0.022 1.224 1.326 generate_qs_task_list 11 8.9 0.375 0.441 1.215 1.316 transfer_rs2pw 451 10.6 0.005 0.005 1.100 1.227 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.179 1.202 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=59.898000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=807.454545, yerr=12.175711 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 1.366655E+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 0.052 0.089 94.909 94.917 qs_mol_dyn_low 1 2.0 0.003 0.003 94.407 94.417 qs_forces 11 3.9 0.003 0.004 91.424 91.425 qs_energies 11 4.9 0.002 0.002 87.300 87.302 scf_env_do_scf 11 5.9 0.000 0.001 73.201 73.201 velocity_verlet 10 3.0 0.021 0.023 58.695 59.064 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.725 44.727 dbcsr_multiply_generic 2055 12.4 0.122 0.126 33.476 33.521 init_scf_loop 11 6.9 0.000 0.000 28.407 28.409 multiply_cannon 2055 13.4 0.333 0.347 26.246 26.917 qs_scf_new_mos 99 7.5 0.001 0.001 26.780 26.811 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.779 26.811 prepare_preconditioner 11 7.9 0.000 0.000 26.359 26.366 make_preconditioner 11 8.9 0.001 0.001 26.359 26.366 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.473 25.836 ot_scf_mini 99 9.5 0.002 0.002 24.970 25.001 multiply_cannon_loop 2055 14.4 0.810 0.828 24.468 24.908 cp_fm_upper_to_full 70 14.2 12.778 18.392 12.778 18.392 ot_mini 99 10.5 0.001 0.001 14.079 14.102 rebuild_ks_matrix 110 8.3 0.001 0.001 14.018 14.057 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.017 14.056 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.708 12.744 dbcsr_complete_redistribute 325 12.2 1.008 1.034 7.439 10.731 init_scf_run 11 5.9 0.000 0.001 9.824 9.824 scf_env_initial_rho_setup 11 6.9 0.001 0.001 9.824 9.824 multiply_cannon_multrec 8220 15.4 4.054 4.225 9.723 9.818 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.425 9.722 qs_ot_get_derivative 99 11.5 0.001 0.001 9.440 9.472 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.869 9.127 mp_alltoall_i22 605 13.7 5.493 8.809 5.493 8.809 mp_waitall_1 84994 16.7 7.651 8.435 7.651 8.435 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.453 7.489 calculate_rho_elec 110 8.6 0.223 0.223 7.453 7.489 sum_up_and_integrate 110 10.3 0.002 0.002 6.927 6.943 integrate_v_rspace 110 11.3 0.004 0.004 6.899 6.915 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 6.779 6.785 calculate_first_density_matrix 1 7.0 0.000 0.000 6.378 6.379 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.289 6.303 make_m2s 4110 13.4 0.043 0.043 5.505 5.985 pw_transfer 1331 11.6 0.074 0.075 5.820 5.825 make_images 4110 14.4 0.877 0.918 5.316 5.795 dbcsr_mm_accdrv_process 11614 15.7 3.904 4.105 5.522 5.734 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.703 5.708 acc_transpose_blocks 8220 15.4 0.042 0.043 5.606 5.631 qs_ot_get_p 110 10.4 0.003 0.003 5.556 5.581 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.823 5.290 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.802 5.243 apply_single 110 13.6 0.000 0.000 4.802 5.243 cp_fm_cholesky_invert 11 10.9 5.236 5.241 5.236 5.241 fft_wrap_pw1pw2_140 451 13.1 1.370 1.377 5.011 5.014 ot_diis_step 99 11.5 0.015 0.016 4.616 4.616 acc_transpose_blocks_kernels 8220 16.4 0.116 0.118 4.554 4.557 jit_kernel_transpose 5 15.6 4.438 4.440 4.438 4.440 qs_ot_p2m_diag 48 11.0 0.151 0.155 3.973 3.981 density_rs2pw 110 9.6 0.004 0.004 3.938 3.958 grid_integrate_task_list 110 12.3 3.683 3.746 3.683 3.746 hybrid_alltoall_any 4261 16.3 0.265 0.564 2.959 3.702 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.665 3.666 make_images_data 4110 15.4 0.047 0.050 3.000 3.613 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.996 3.472 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.459 3.460 fft3d_ps 1111 14.6 1.883 1.887 3.406 3.417 wfi_extrapolate 11 7.9 0.001 0.001 3.329 3.329 calculate_dm_sparse 110 9.5 0.001 0.001 3.199 3.223 multiply_cannon_sync_h2d 8220 15.4 3.150 3.173 3.150 3.173 mp_sum_dm 438 4.9 2.590 2.960 2.590 2.960 md_output 10 3.0 0.000 0.000 0.373 2.959 md_write_output 11 3.9 0.142 1.132 0.371 2.958 update_particle_set 20 4.0 0.000 0.000 2.576 2.945 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.880 2.880 cp_fm_diag_elpa_base 48 14.0 2.338 2.542 2.879 2.879 grid_collocate_task_list 110 9.6 2.672 2.682 2.672 2.682 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.616 2.634 potential_pw2rs 110 12.3 0.021 0.021 2.455 2.460 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.186 2.256 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.101 2.183 mp_alltoall_d11v 2046 13.8 1.966 2.027 1.966 2.027 cp_fm_cholesky_decompose 22 10.9 1.957 1.969 1.957 1.969 qs_create_task_list 11 7.9 0.001 0.002 1.878 1.922 generate_qs_task_list 11 8.9 0.730 0.784 1.877 1.921 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.906 1.909 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=94.917000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1238.454545, yerr=49.365729 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 1.103326E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 12044928 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.1 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 632.340480E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10431456 MPI messages size (bytes): total size 4.526012E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.881156E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3605352 96403587072 32768 < size <= 131072 1305088 74666999808 131072 < size <= 4194304 5190114 3200148350408 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57450. MP_Allreduce 11139 794. MP_Sync 87 MP_Alltoall 2242 2530001. MP_SendRecv 24510 18752. MP_ISendRecv 24510 18752. MP_Wait 42802 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.044 211.549 211.550 qs_mol_dyn_low 1 2.0 0.003 0.003 210.653 210.667 qs_forces 11 3.9 0.005 0.005 210.560 210.561 qs_energies 11 4.9 0.002 0.003 204.936 204.955 scf_env_do_scf 11 5.9 0.001 0.001 187.977 187.981 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 167.023 167.025 dbcsr_multiply_generic 2527 12.6 0.218 0.224 126.957 127.675 qs_scf_new_mos 118 7.6 0.001 0.001 126.769 127.089 qs_scf_loop_do_ot 118 8.6 0.001 0.001 126.768 127.089 velocity_verlet 10 3.0 0.005 0.012 126.753 126.754 ot_scf_mini 118 9.6 0.003 0.003 119.997 120.350 multiply_cannon 2527 13.6 0.238 0.249 102.435 104.640 multiply_cannon_loop 2527 14.6 2.412 2.461 100.190 101.998 ot_mini 118 10.6 0.001 0.001 67.199 67.524 multiply_cannon_multrec 60648 15.6 31.931 33.576 41.787 43.691 qs_ot_get_derivative 118 11.6 0.001 0.001 42.209 42.526 rebuild_ks_matrix 129 8.3 0.001 0.001 34.022 34.375 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.027 34.022 34.375 mp_waitall_1 269254 16.5 29.432 32.819 29.432 32.819 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.571 30.896 qs_ot_get_p 129 10.4 0.001 0.002 30.342 30.630 multiply_cannon_sync_h2d 60648 15.6 26.478 28.688 26.478 28.688 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.450 25.268 apply_single 129 13.6 0.001 0.001 24.450 25.268 ot_diis_step 118 11.6 0.008 0.008 24.739 24.740 qs_ot_p2m_diag 83 11.4 0.079 0.092 23.724 23.796 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.892 20.893 init_scf_loop 11 6.9 0.000 0.000 20.879 20.880 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.188 20.416 multiply_cannon_metrocomm3 60648 15.6 0.119 0.124 16.224 18.860 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.852 17.886 cp_fm_diag_elpa_base 83 14.4 17.776 17.820 17.848 17.880 prepare_preconditioner 11 7.9 0.000 0.000 16.306 16.343 make_preconditioner 11 8.9 0.000 0.001 16.306 16.343 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.514 15.717 make_m2s 5054 13.6 0.105 0.117 14.047 14.446 sum_up_and_integrate 129 10.3 0.002 0.004 14.310 14.329 integrate_v_rspace 129 11.3 0.003 0.004 14.251 14.271 make_images 5054 14.6 0.403 0.421 13.862 14.270 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.583 13.753 calculate_rho_elec 129 8.7 0.045 0.064 13.582 13.752 init_scf_run 11 5.9 0.000 0.001 12.784 12.785 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.784 12.784 mp_sum_l 8010 12.9 9.165 10.813 9.165 10.813 dbcsr_mm_accdrv_process 125468 16.2 4.769 4.911 9.423 9.984 cp_fm_cholesky_invert 11 10.9 9.430 9.437 9.430 9.437 wfi_extrapolate 11 7.9 0.001 0.001 9.170 9.171 calculate_dm_sparse 129 9.5 0.001 0.001 8.579 8.679 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 8.398 8.530 multiply_cannon_metrocomm1 60648 15.6 0.095 0.098 6.476 8.469 qs_ot_get_orbitals 118 10.6 0.001 0.001 8.205 8.297 pw_transfer 1559 11.6 0.074 0.086 7.855 8.055 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 7.652 7.850 make_images_data 5054 15.6 0.068 0.074 6.868 7.675 grid_integrate_task_list 129 12.3 7.119 7.619 7.119 7.619 density_rs2pw 129 9.7 0.006 0.006 7.116 7.541 hybrid_alltoall_any 5240 16.5 0.297 2.296 6.009 7.289 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.941 6.953 fft_wrap_pw1pw2_140 527 13.2 0.862 0.902 6.631 6.831 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.939 6.104 fft3d_ps 1301 14.7 2.214 2.866 5.647 5.986 mp_alltoall_d11v 2423 14.1 4.467 5.756 4.467 5.756 grid_collocate_task_list 129 9.7 4.879 5.213 4.879 5.213 potential_pw2rs 129 12.3 0.009 0.010 4.720 4.750 cp_fm_cholesky_decompose 22 10.9 4.667 4.680 4.667 4.680 mp_sum_d 4501 12.1 3.450 4.482 3.450 4.482 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=211.550000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.090909, yerr=6.302302 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 2.200017E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 6024768 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.7 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 841.310208E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2427840 MPI messages size (bytes): total size 4.132587E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.702166E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71532 2339635200 32768 < size <= 131072 729952 56049532928 131072 < size <= 4194304 1387568 1410045313024 4194304 < size <= 16777216 155760 1473827979536 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57636. MP_Allreduce 11185 956. MP_Sync 88 MP_Alltoall 1983 5044185. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.052 0.072 190.256 190.259 qs_mol_dyn_low 1 2.0 0.003 0.003 189.636 189.694 qs_forces 11 3.9 0.004 0.005 189.551 189.552 qs_energies 11 4.9 0.003 0.004 182.785 182.796 scf_env_do_scf 11 5.9 0.001 0.001 165.828 165.838 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 133.214 133.217 velocity_verlet 10 3.0 0.006 0.012 119.616 119.618 dbcsr_multiply_generic 2529 12.6 0.190 0.197 97.614 98.982 qs_scf_new_mos 118 7.6 0.001 0.001 94.610 95.362 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.610 95.361 ot_scf_mini 118 9.6 0.004 0.004 89.768 90.602 multiply_cannon 2529 13.6 0.506 0.565 77.307 82.084 multiply_cannon_loop 2529 14.6 1.584 1.660 74.410 77.463 ot_mini 118 10.6 0.001 0.001 50.460 51.185 mp_waitall_1 216598 16.6 24.146 37.466 24.146 37.466 multiply_cannon_multrec 30348 15.6 21.045 25.781 31.804 37.012 rebuild_ks_matrix 129 8.3 0.001 0.001 32.292 33.068 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.025 32.292 33.067 init_scf_loop 11 6.9 0.000 0.000 32.527 32.528 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.039 29.754 qs_ot_get_derivative 118 11.6 0.001 0.002 28.269 29.081 prepare_preconditioner 11 7.9 0.000 0.000 28.258 28.332 make_preconditioner 11 8.9 0.001 0.009 28.258 28.332 multiply_cannon_metrocomm3 30348 15.6 0.101 0.108 15.718 28.264 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.903 27.456 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.268 23.434 apply_single 129 13.6 0.001 0.001 22.268 23.433 qs_ot_get_p 129 10.4 0.001 0.002 21.371 22.092 ot_diis_step 118 11.6 0.015 0.015 22.003 22.005 multiply_cannon_sync_h2d 30348 15.6 18.093 20.644 18.093 20.644 cp_fm_cholesky_invert 11 10.9 16.484 16.496 16.484 16.496 qs_ot_p2m_diag 84 11.4 0.190 0.219 16.442 16.478 make_m2s 5058 13.6 0.087 0.093 14.241 15.425 cp_dbcsr_syevd 84 12.4 0.006 0.006 15.224 15.225 make_images 5058 14.6 1.177 1.380 14.023 15.207 sum_up_and_integrate 129 10.3 0.002 0.004 14.046 14.073 integrate_v_rspace 129 11.3 0.003 0.004 13.985 14.015 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.381 13.413 calculate_rho_elec 129 8.7 0.087 0.104 13.381 13.412 init_scf_run 11 5.9 0.000 0.001 12.136 12.137 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.135 12.137 cp_fm_diag_elpa 84 13.4 0.000 0.000 12.024 12.061 cp_fm_diag_elpa_base 84 14.4 11.770 11.865 12.019 12.051 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 11.306 11.900 multiply_cannon_metrocomm4 27819 15.6 0.102 0.119 3.914 11.150 dbcsr_mm_accdrv_process 62758 16.2 5.504 6.069 10.205 10.836 mp_irecv_dv 70084 16.3 3.710 10.738 3.710 10.738 make_images_data 5058 15.6 0.069 0.078 8.436 9.869 hybrid_alltoall_any 5245 16.5 0.353 1.551 7.120 9.131 pw_transfer 1559 11.6 0.085 0.100 8.589 8.641 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.364 8.419 wfi_extrapolate 11 7.9 0.001 0.001 8.325 8.325 grid_integrate_task_list 129 12.3 7.251 7.605 7.251 7.605 density_rs2pw 129 9.7 0.006 0.006 7.048 7.567 fft_wrap_pw1pw2_140 527 13.2 0.941 0.965 7.412 7.482 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.226 7.048 cp_fm_cholesky_decompose 22 10.9 6.851 6.927 6.851 6.927 calculate_dm_sparse 129 9.5 0.001 0.001 6.558 6.712 mp_sum_l 8016 12.9 4.263 6.441 4.263 6.441 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.262 6.273 fft3d_ps 1301 14.7 2.850 3.024 5.904 5.948 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.401 5.475 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.243 5.456 grid_collocate_task_list 129 9.7 5.043 5.376 5.043 5.376 mp_alltoall_d11v 2429 14.1 4.179 4.865 4.179 4.865 mp_allgather_i34 2529 14.6 1.411 4.562 1.411 4.562 potential_pw2rs 129 12.3 0.016 0.018 4.463 4.481 dbcsr_complete_redistribute 397 12.7 0.793 0.993 3.271 4.117 mp_sum_d 4511 12.1 2.655 4.044 2.655 4.044 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.259000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=802.454545, yerr=1.157084 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 2.951022E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 4017216 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.3 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 955.797504E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1052064 MPI messages size (bytes): total size 2.737208E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.601750E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 284544 37295751168 131072 < size <= 4194304 665856 1004519030784 4194304 < size <= 16777216 66080 937891119088 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57629. MP_Allreduce 11184 996. MP_Sync 88 MP_Alltoall 1724 9394234. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.033 179.855 179.856 qs_mol_dyn_low 1 2.0 0.003 0.003 179.424 179.438 qs_forces 11 3.9 0.006 0.008 178.779 178.781 qs_energies 11 4.9 0.002 0.003 172.285 172.299 scf_env_do_scf 11 5.9 0.001 0.001 152.443 152.444 scf_env_do_scf_inner_loop 118 6.6 0.015 0.041 117.954 117.954 velocity_verlet 10 3.0 0.001 0.002 112.714 112.733 dbcsr_multiply_generic 2529 12.6 0.186 0.192 86.804 87.941 qs_scf_new_mos 118 7.6 0.001 0.001 80.762 81.065 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.762 81.064 ot_scf_mini 118 9.6 0.003 0.004 76.500 76.853 multiply_cannon 2529 13.6 0.510 0.527 66.957 72.118 multiply_cannon_loop 2529 14.6 1.139 1.220 64.102 66.636 ot_mini 118 10.6 0.001 0.001 43.044 43.421 init_scf_loop 11 6.9 0.000 0.000 34.395 34.396 mp_waitall_1 172006 16.6 24.668 34.181 24.668 34.181 rebuild_ks_matrix 129 8.3 0.001 0.001 30.212 30.652 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.017 30.211 30.651 prepare_preconditioner 11 7.9 0.000 0.000 30.407 30.457 make_preconditioner 11 8.9 0.000 0.001 30.407 30.457 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.034 29.405 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.260 27.662 multiply_cannon_multrec 20232 15.6 13.275 16.105 22.704 25.573 multiply_cannon_metrocomm3 20232 15.6 0.063 0.066 15.253 24.602 qs_ot_get_derivative 118 11.6 0.002 0.002 23.391 23.757 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.705 20.788 apply_single 129 13.6 0.001 0.001 19.705 20.787 ot_diis_step 118 11.6 0.018 0.018 19.543 19.544 qs_ot_get_p 129 10.4 0.009 0.021 18.831 19.295 make_m2s 5058 13.6 0.079 0.084 14.672 15.791 make_images 5058 14.6 1.163 1.258 14.433 15.547 multiply_cannon_sync_h2d 20232 15.6 13.759 15.292 13.759 15.292 init_scf_run 11 5.9 0.000 0.001 14.862 14.863 scf_env_initial_rho_setup 11 6.9 0.001 0.001 14.861 14.862 qs_ot_p2m_diag 84 11.4 0.268 0.275 14.458 14.466 cp_fm_cholesky_invert 11 10.9 14.086 14.095 14.086 14.095 sum_up_and_integrate 129 10.3 0.002 0.003 13.974 14.001 integrate_v_rspace 129 11.3 0.003 0.004 13.914 13.943 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.591 13.623 calculate_rho_elec 129 8.7 0.129 0.144 13.590 13.623 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.395 13.396 make_images_data 5058 15.6 0.066 0.076 8.798 10.297 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.200 10.227 cp_fm_diag_elpa_base 84 14.4 9.790 9.949 10.197 10.224 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 9.855 9.864 hybrid_alltoall_any 5245 16.5 0.452 2.051 7.584 9.853 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.263 9.531 multiply_cannon_metrocomm4 17703 15.6 0.067 0.077 3.448 9.459 mp_irecv_dv 50659 16.2 3.317 9.187 3.317 9.187 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 8.985 9.125 dbcsr_mm_accdrv_process 41846 16.2 5.653 5.946 8.880 9.031 pw_transfer 1559 11.6 0.085 0.103 8.802 8.899 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.577 8.682 grid_integrate_task_list 129 12.3 7.369 7.771 7.369 7.771 fft_wrap_pw1pw2_140 527 13.2 1.065 1.108 7.633 7.740 wfi_extrapolate 11 7.9 0.001 0.001 7.385 7.385 calculate_first_density_matrix 1 7.0 0.000 0.000 7.292 7.295 cp_fm_upper_to_full 106 14.8 5.749 7.229 5.749 7.229 density_rs2pw 129 9.7 0.006 0.006 6.910 7.209 cp_fm_cholesky_decompose 22 10.9 7.140 7.153 7.140 7.153 acc_transpose_blocks 20232 15.6 0.105 0.109 6.341 6.473 dbcsr_complete_redistribute 397 12.7 1.155 1.188 4.532 6.276 fft3d_ps 1301 14.7 2.840 3.069 5.874 5.967 calculate_dm_sparse 129 9.5 0.001 0.001 5.856 5.953 grid_collocate_task_list 129 9.7 5.202 5.571 5.202 5.571 acc_transpose_blocks_kernels 20232 16.6 0.287 0.293 5.310 5.422 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.746 5.357 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.405 5.143 jit_kernel_transpose 5 15.6 5.023 5.135 5.023 5.135 mp_sum_l 8016 12.9 3.330 4.905 3.330 4.905 mp_alltoall_d11v 2429 14.1 4.319 4.824 4.319 4.824 mp_allgather_i34 2529 14.6 1.348 4.358 1.348 4.358 potential_pw2rs 129 12.3 0.020 0.022 4.332 4.341 transfer_fm_to_dbcsr 11 9.9 0.019 0.025 2.354 4.078 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.934 3.959 mp_alltoall_i22 720 14.1 1.924 3.826 1.924 3.826 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 3.716 3.718 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=179.856000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=907.181818, yerr=8.715902 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410024443904 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444712984576 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796586E+12 0.0% 0.0% 100.0% flops max/rank 4.320340E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705502176 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.144881E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837893360 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58189. MP_Allreduce 11085 1083. MP_Sync 86 MP_Alltoall 1700 12496381. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.028 190.225 190.226 qs_mol_dyn_low 1 2.0 0.003 0.003 189.874 189.888 qs_forces 11 3.9 0.004 0.005 188.755 188.765 qs_energies 11 4.9 0.002 0.003 181.711 181.723 scf_env_do_scf 11 5.9 0.001 0.001 161.341 161.355 velocity_verlet 10 3.0 0.493 1.028 123.112 123.157 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.522 115.523 dbcsr_multiply_generic 2485 12.5 0.191 0.197 83.185 84.034 qs_scf_new_mos 116 7.6 0.001 0.001 79.583 79.911 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.582 79.910 ot_scf_mini 116 9.6 0.003 0.003 75.053 75.399 multiply_cannon 2485 13.5 0.545 0.583 59.110 62.239 multiply_cannon_loop 2485 14.5 1.814 1.867 55.644 57.674 init_scf_loop 11 6.9 0.000 0.000 45.700 45.700 ot_mini 116 10.6 0.001 0.001 42.265 42.595 prepare_preconditioner 11 7.9 0.000 0.000 41.621 41.657 make_preconditioner 11 8.9 0.000 0.001 41.621 41.657 make_full_inverse_cholesky 11 9.9 0.010 0.023 35.246 40.189 multiply_cannon_multrec 29820 15.5 13.534 18.761 26.434 31.318 rebuild_ks_matrix 127 8.3 0.001 0.001 29.021 29.326 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 29.021 29.325 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.192 26.475 mp_waitall_1 146592 16.7 16.801 26.119 16.801 26.119 qs_ot_get_derivative 116 11.6 0.001 0.002 23.209 23.562 make_m2s 4970 13.5 0.092 0.097 19.645 20.339 make_images 4970 14.5 1.912 2.153 19.338 20.030 qs_ot_get_p 127 10.4 0.001 0.002 19.156 19.517 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.452 18.984 apply_single 127 13.6 0.001 0.001 18.452 18.984 ot_diis_step 116 11.6 0.017 0.018 18.932 18.933 cp_fm_upper_to_full 104 14.8 11.477 16.814 11.477 16.814 cp_fm_cholesky_invert 11 10.9 15.695 15.703 15.695 15.703 multiply_cannon_metrocomm3 29820 15.5 0.049 0.052 6.345 15.580 qs_ot_p2m_diag 82 11.4 0.338 0.384 14.959 15.009 init_scf_run 11 5.9 0.000 0.001 14.476 14.478 scf_env_initial_rho_setup 11 6.9 0.001 0.001 14.476 14.477 sum_up_and_integrate 127 10.3 0.002 0.003 13.867 13.889 integrate_v_rspace 127 11.3 0.003 0.004 13.808 13.831 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.651 13.700 calculate_rho_elec 127 8.7 0.168 0.183 13.650 13.699 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.578 13.579 dbcsr_mm_accdrv_process 61748 16.2 8.350 9.158 12.475 12.976 dbcsr_complete_redistribute 393 12.7 1.476 1.624 9.079 12.839 make_images_data 4970 15.5 0.067 0.073 10.473 11.776 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.719 11.486 multiply_cannon_sync_h2d 29820 15.5 10.360 11.148 10.360 11.148 hybrid_alltoall_any 5155 16.4 0.522 2.167 9.320 10.867 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.415 10.432 cp_fm_diag_elpa_base 82 14.4 9.465 9.770 10.409 10.424 transfer_fm_to_dbcsr 11 9.9 0.000 0.004 6.353 10.019 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.702 9.947 mp_alltoall_i22 712 14.1 5.551 9.416 5.551 9.416 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 9.206 9.274 pw_transfer 1535 11.6 0.085 0.104 9.085 9.152 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.861 8.935 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 8.131 8.220 fft_wrap_pw1pw2_140 519 13.2 1.239 1.292 7.898 7.989 grid_integrate_task_list 127 12.3 7.488 7.725 7.488 7.725 wfi_extrapolate 11 7.9 0.001 0.001 7.608 7.608 cp_fm_cholesky_decompose 22 10.9 7.405 7.474 7.405 7.474 density_rs2pw 127 9.7 0.005 0.006 6.754 7.060 acc_transpose_blocks 29820 15.5 0.149 0.155 6.696 6.822 multiply_cannon_metrocomm4 24850 15.5 0.081 0.089 2.764 6.758 calculate_first_density_matrix 1 7.0 0.000 0.000 6.667 6.674 mp_irecv_dv 75445 16.2 2.610 6.486 2.610 6.486 calculate_dm_sparse 127 9.5 0.001 0.001 6.213 6.315 fft3d_ps 1281 14.7 2.966 3.002 5.929 5.974 grid_collocate_task_list 127 9.7 5.241 5.682 5.241 5.682 mp_alltoall_d11v 2401 14.1 4.914 5.198 4.914 5.198 acc_transpose_blocks_kernels 29820 16.5 0.398 0.406 4.751 4.781 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.506 4.603 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.417 4.418 jit_kernel_transpose 5 15.6 4.353 4.378 4.353 4.378 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.184 4.250 potential_pw2rs 127 12.3 0.022 0.023 4.216 4.228 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=190.226000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1081.181818, yerr=17.988977 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_performance_tests/19/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 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 92.796578E+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 6705500608 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 1.578242E+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 716108623056 ------------------------------------------------------------------------------- - - - 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 4003 58187. MP_Allreduce 11085 1167. MP_Sync 86 MP_Alltoall 1700 18828148. 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 0.026 0.044 172.482 172.484 qs_mol_dyn_low 1 2.0 0.003 0.003 172.085 172.098 qs_forces 11 3.9 0.004 0.005 171.991 171.995 qs_energies 11 4.9 0.036 0.037 164.600 164.608 scf_env_do_scf 11 5.9 0.001 0.001 142.842 142.855 velocity_verlet 10 3.0 0.011 0.011 110.088 110.091 scf_env_do_scf_inner_loop 116 6.6 0.011 0.021 107.287 107.288 dbcsr_multiply_generic 2485 12.5 0.182 0.189 75.950 76.464 qs_scf_new_mos 116 7.6 0.001 0.001 70.594 70.711 qs_scf_loop_do_ot 116 8.6 0.001 0.001 70.593 70.710 ot_scf_mini 116 9.6 0.003 0.004 66.237 66.405 multiply_cannon 2485 13.5 0.552 0.573 57.330 60.465 multiply_cannon_loop 2485 14.5 0.803 0.834 54.427 55.613 ot_mini 116 10.6 0.001 0.001 37.319 37.492 init_scf_loop 11 6.9 0.000 0.000 35.407 35.409 mp_waitall_1 124680 16.7 24.895 31.828 24.895 31.828 prepare_preconditioner 11 7.9 0.000 0.000 31.549 31.580 make_preconditioner 11 8.9 0.000 0.001 31.549 31.580 make_full_inverse_cholesky 11 9.9 0.013 0.025 29.430 29.684 rebuild_ks_matrix 127 8.3 0.001 0.001 28.965 29.126 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 28.964 29.126 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.382 26.531 multiply_cannon_multrec 9940 15.5 10.159 14.712 17.689 22.445 multiply_cannon_metrocomm3 9940 15.5 0.024 0.026 12.799 20.729 qs_ot_get_derivative 116 11.6 0.002 0.002 20.302 20.479 cp_fm_cholesky_invert 11 10.9 18.051 18.057 18.051 18.057 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.065 17.301 apply_single 127 13.6 0.001 0.001 17.065 17.300 ot_diis_step 116 11.6 0.020 0.020 16.945 16.945 qs_ot_get_p 127 10.4 0.002 0.008 16.294 16.484 make_m2s 4970 13.5 0.063 0.069 14.764 15.606 make_images 4970 14.5 2.142 2.562 14.459 15.303 init_scf_run 11 5.9 0.000 0.001 14.912 14.912 scf_env_initial_rho_setup 11 6.9 0.026 0.027 14.912 14.912 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.085 14.135 calculate_rho_elec 127 8.7 0.251 0.262 14.085 14.135 sum_up_and_integrate 127 10.3 0.002 0.002 14.019 14.063 integrate_v_rspace 127 11.3 0.004 0.004 13.959 14.004 qs_ot_p2m_diag 82 11.4 0.489 0.495 12.606 12.623 cp_dbcsr_syevd 82 12.4 0.005 0.005 11.464 11.464 multiply_cannon_sync_h2d 9940 15.5 10.565 11.137 10.565 11.137 pw_transfer 1535 11.6 0.084 0.090 9.752 9.789 hybrid_alltoall_any 5155 16.4 0.835 3.751 8.389 9.782 make_images_data 4970 15.5 0.055 0.066 8.577 9.779 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 9.533 9.574 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 9.554 9.571 cp_fm_diag_elpa 82 13.4 0.000 0.000 8.659 8.669 cp_fm_diag_elpa_base 82 14.4 8.427 8.499 8.655 8.666 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 8.487 8.556 fft_wrap_pw1pw2_140 519 13.2 1.671 1.707 8.467 8.517 grid_integrate_task_list 127 12.3 7.719 8.033 7.719 8.033 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.903 8.030 cp_fm_cholesky_decompose 22 10.9 7.810 7.910 7.810 7.910 dbcsr_mm_accdrv_process 20590 16.1 3.133 4.397 7.185 7.893 multiply_cannon_metrocomm1 9940 15.5 0.030 0.031 4.310 7.840 wfi_extrapolate 11 7.9 0.001 0.001 7.318 7.318 calculate_first_density_matrix 1 7.0 0.000 0.000 7.316 7.317 density_rs2pw 127 9.7 0.005 0.005 6.833 7.144 acc_transpose_blocks 9940 15.5 0.055 0.057 6.086 6.108 calculate_dm_sparse 127 9.5 0.001 0.001 6.003 6.103 fft3d_ps 1281 14.7 3.120 3.231 6.004 6.015 multiply_cannon_metrocomm4 7455 15.5 0.026 0.030 1.856 5.924 mp_irecv_dv 28618 15.9 1.813 5.829 1.813 5.829 grid_collocate_task_list 127 9.7 5.531 5.815 5.531 5.815 dbcsr_complete_redistribute 393 12.7 2.073 2.143 5.215 5.574 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.266 5.268 acc_transpose_blocks_kernels 9940 16.5 0.149 0.153 5.128 5.164 mp_alltoall_d11v 2401 14.1 4.609 5.023 4.609 5.023 jit_kernel_transpose 5 15.6 4.979 5.019 4.979 5.019 mp_allgather_i34 2485 14.5 1.119 4.611 1.119 4.611 potential_pw2rs 127 12.3 0.026 0.026 4.279 4.283 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.591 3.873 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.605 3.691 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.355 3.661 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.579 3.621 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.427 3.545 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.472 3.487 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=172.484000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1481.363636, yerr=34.717752 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 3.240542E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3955 59814. MP_Allreduce 10965 1506. MP_Sync 86 MP_Alltoall 1700 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 0.020 0.038 292.414 292.416 qs_mol_dyn_low 1 2.0 0.003 0.003 291.897 291.910 qs_forces 11 3.9 0.005 0.005 291.804 291.807 qs_energies 11 4.9 0.201 0.219 282.941 282.948 scf_env_do_scf 11 5.9 0.001 0.001 256.247 256.256 velocity_verlet 10 3.0 0.002 0.002 208.811 208.818 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 132.314 132.316 init_scf_loop 11 6.9 0.000 0.000 123.662 123.664 prepare_preconditioner 11 7.9 0.000 0.000 118.815 118.833 make_preconditioner 11 8.9 0.000 0.001 118.815 118.833 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.164 115.992 qs_scf_new_mos 116 7.6 0.001 0.001 88.127 88.185 qs_scf_loop_do_ot 116 8.6 0.001 0.001 88.126 88.184 dbcsr_multiply_generic 2485 12.5 0.215 0.225 84.348 84.808 ot_scf_mini 116 9.6 0.004 0.004 83.271 83.304 cp_fm_upper_to_full 104 14.8 52.964 75.599 52.964 75.599 multiply_cannon 2485 13.5 0.678 0.724 61.991 63.043 multiply_cannon_loop 2485 14.5 1.023 1.056 58.091 59.376 ot_mini 116 10.6 0.001 0.001 43.565 43.603 dbcsr_complete_redistribute 393 12.7 3.988 4.064 29.719 42.318 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.146 38.772 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.609 36.168 rebuild_ks_matrix 127 8.3 0.001 0.001 34.322 34.341 qs_ks_build_kohn_sham_matrix 127 9.3 0.020 0.023 34.321 34.340 mp_alltoall_i22 712 14.1 21.379 33.892 21.379 33.892 cp_fm_cholesky_invert 11 10.9 33.136 33.142 33.136 33.142 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.821 31.846 mp_waitall_1 102768 16.8 27.499 31.696 27.499 31.696 qs_ot_get_p 127 10.4 0.002 0.002 24.822 24.894 qs_ot_get_derivative 116 11.6 0.002 0.002 24.214 24.247 qs_ot_p2m_diag 82 11.4 0.870 0.875 20.899 20.928 multiply_cannon_metrocomm3 9940 15.5 0.025 0.026 18.929 20.191 ot_diis_step 116 11.6 0.021 0.022 19.323 19.323 make_m2s 4970 13.5 0.073 0.077 17.875 19.281 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.134 19.136 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.714 18.810 apply_single 127 13.6 0.001 0.001 18.714 18.809 make_images 4970 14.5 3.031 3.224 17.396 18.805 multiply_cannon_multrec 9940 15.5 10.192 11.997 17.931 18.024 qs_rho_update_rho_low 127 7.7 0.001 0.001 17.075 17.105 calculate_rho_elec 127 8.7 0.475 0.475 17.074 17.104 sum_up_and_integrate 127 10.3 0.002 0.002 16.227 16.316 integrate_v_rspace 127 11.3 0.005 0.005 16.165 16.255 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.019 16.020 init_scf_run 11 5.9 0.000 0.001 16.016 16.016 scf_env_initial_rho_setup 11 6.9 0.001 0.001 16.016 16.016 cp_fm_diag_elpa_base 82 14.4 11.666 13.252 16.015 16.015 multiply_cannon_sync_h2d 9940 15.5 14.221 14.270 14.221 14.270 pw_transfer 1535 11.6 0.095 0.096 13.034 13.039 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 12.797 12.802 hybrid_alltoall_any 5155 16.4 1.299 3.027 10.156 12.227 make_images_data 4970 15.5 0.066 0.071 9.882 11.964 fft_wrap_pw1pw2_140 519 13.2 3.196 3.290 11.398 11.406 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 10.024 10.098 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.487 9.513 dbcsr_mm_accdrv_process 20590 16.0 4.232 6.143 7.491 9.397 wfi_extrapolate 11 7.9 0.001 0.001 9.237 9.237 cp_fm_cholesky_decompose 22 10.9 8.877 8.904 8.877 8.904 grid_integrate_task_list 127 12.3 8.498 8.696 8.498 8.696 density_rs2pw 127 9.7 0.005 0.005 8.213 8.262 qs_energies_init_hamiltonians 11 5.9 0.008 0.017 8.242 8.242 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 7.858 7.885 fft3d_ps 1281 14.7 3.942 3.965 7.298 7.342 mp_alltoall_d11v 2401 14.1 6.884 6.995 6.884 6.995 calculate_dm_sparse 127 9.5 0.001 0.001 6.734 6.813 calculate_first_density_matrix 1 7.0 0.000 0.000 6.508 6.511 grid_collocate_task_list 127 9.7 6.365 6.444 6.365 6.444 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.286 6.371 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=292.416000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2919.272727, yerr=163.510406 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 1.260941E+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 13810295. 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 0.039 0.048 85.331 85.333 qs_energies 1 2.0 0.000 0.000 84.682 84.689 ls_scf 1 3.0 0.000 0.000 83.777 83.784 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.678 72.798 multiply_cannon 111 7.7 0.017 0.020 55.996 57.042 multiply_cannon_loop 111 8.7 0.227 0.245 52.571 53.809 ls_scf_main 1 4.0 0.000 0.000 52.362 52.373 density_matrix_trs4 2 5.0 0.002 0.003 46.806 46.891 ls_scf_init_scf 1 4.0 0.000 0.000 28.384 28.386 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.257 27.313 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.158 25.174 mp_waitall_1 11031 10.9 22.329 24.919 22.329 24.919 multiply_cannon_multrec 2664 9.7 8.164 8.969 15.605 17.450 multiply_cannon_sync_h2d 2664 9.7 13.648 15.237 13.648 15.237 make_m2s 222 7.7 0.009 0.011 13.078 13.528 make_images 222 8.7 0.098 0.108 13.056 13.507 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.663 11.467 make_images_data 222 9.7 0.004 0.005 7.614 8.198 dbcsr_mm_accdrv_process 4760 10.4 0.589 0.711 7.056 8.037 hybrid_alltoall_any 227 10.6 0.216 1.838 6.536 7.872 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.381 7.797 dbcsr_mm_accdrv_process_sort 4760 11.4 6.268 7.159 6.268 7.159 calculate_norms 4752 9.8 5.510 6.285 5.510 6.285 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.035 5.169 mp_sum_l 887 5.1 3.069 4.747 3.069 4.747 make_images_sizes 222 9.7 0.000 0.000 0.748 3.623 mp_alltoall_i44 222 10.7 0.748 3.623 0.748 3.623 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.074 3.498 mp_irecv_dv 6231 10.9 2.057 3.475 2.057 3.475 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.283 3.344 arnoldi_extremal 4 6.8 0.000 0.000 3.235 3.254 arnoldi_normal_ev 4 7.8 0.001 0.002 3.235 3.253 build_subspace 16 8.4 0.009 0.012 3.140 3.143 ls_scf_post 1 4.0 0.000 0.000 3.030 3.037 ls_scf_store_result 1 5.0 0.000 0.000 2.851 2.893 dbcsr_special_finalize 555 9.7 0.005 0.006 2.365 2.873 dbcsr_merge_single_wm 555 10.7 0.464 0.592 2.357 2.865 make_images_pack 222 9.7 2.205 2.626 2.207 2.628 dbcsr_sort_data 658 11.4 2.150 2.613 2.150 2.613 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.336 2.577 dbcsr_matrix_vector_mult_local 304 10.0 2.071 2.466 2.073 2.467 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.311 2.382 buffer_matrices_ensure_size 222 8.7 1.753 2.051 1.753 2.051 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.768 1.770 rebuild_ks_matrix 3 7.3 0.000 0.000 1.759 1.760 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.759 1.760 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.333000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 2.164523E+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 0.049 0.146 93.335 93.339 qs_energies 1 2.0 0.000 0.005 92.478 92.496 ls_scf 1 3.0 0.010 0.021 91.053 91.073 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.918 75.917 ls_scf_main 1 4.0 0.601 1.653 57.384 57.392 multiply_cannon 111 7.7 0.027 0.039 52.591 56.553 multiply_cannon_loop 111 8.7 0.135 0.146 49.802 52.757 density_matrix_trs4 2 5.0 0.653 1.802 49.167 50.432 ls_scf_init_scf 1 4.0 0.000 0.002 29.936 29.939 mp_waitall_1 9105 10.9 20.474 28.817 20.474 28.817 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.695 28.789 multiply_cannon_multrec 1332 9.7 13.299 16.688 22.658 27.186 matrix_sqrt_Newton_Schulz 2 6.5 0.006 0.013 26.319 26.330 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.316 19.439 make_m2s 222 7.7 0.006 0.007 14.961 15.474 make_images 222 8.7 1.368 1.692 14.930 15.445 dbcsr_mm_accdrv_process 4041 10.4 0.323 0.524 8.950 10.493 dbcsr_mm_accdrv_process_sort 4041 11.4 8.467 9.965 8.467 9.965 make_images_data 222 9.7 0.004 0.005 8.684 9.557 hybrid_alltoall_any 227 10.6 0.543 2.530 8.069 9.243 mp_sum_l 887 5.1 5.922 9.239 5.922 9.239 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.201 7.732 mp_irecv_dv 3311 11.0 3.181 7.683 3.181 7.683 calculate_norms 2376 9.8 6.070 6.884 6.070 6.884 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.898 6.784 multiply_cannon_sync_h2d 1332 9.7 4.937 6.145 4.937 6.145 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.680 5.921 arnoldi_extremal 4 6.8 0.000 0.000 4.643 4.660 arnoldi_normal_ev 4 7.8 0.001 0.004 4.643 4.660 build_subspace 16 8.4 0.014 0.021 4.372 4.375 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.559 4.264 ls_scf_post 1 4.0 0.002 0.006 3.723 3.733 ls_scf_store_result 1 5.0 0.000 0.000 3.399 3.546 dbcsr_matrix_vector_mult 304 9.0 0.010 0.023 3.118 3.349 dbcsr_matrix_vector_mult_local 304 10.0 2.737 3.214 2.739 3.216 mp_allgather_i34 111 8.7 0.782 3.009 0.782 3.009 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.125 2.809 dbcsr_data_new 4174 10.1 2.109 2.389 2.109 2.389 dbcsr_complete_redistribute 5 7.6 0.239 0.253 1.579 2.286 make_images_pack 222 9.7 1.816 2.143 1.818 2.145 matrix_ls_to_qs 2 6.0 0.000 0.000 1.452 2.144 dbcsr_sort_data 436 11.2 1.857 2.119 1.857 2.119 matrix_decluster 2 7.0 0.000 0.000 1.387 2.079 mp_alltoall_i22 14 8.6 1.217 1.954 1.217 1.954 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.895 1.898 rebuild_ks_matrix 3 7.3 0.000 0.000 1.882 1.885 qs_ks_build_kohn_sham_matrix 3 8.3 0.008 0.034 1.882 1.885 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=93.339000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1807.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 2.940744E+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 0.131 0.154 95.431 95.436 qs_energies 1 2.0 0.000 0.001 94.322 94.325 ls_scf 1 3.0 0.000 0.002 92.906 92.910 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.477 75.901 ls_scf_main 1 4.0 0.214 1.726 58.876 58.882 multiply_cannon 111 7.7 0.036 0.081 51.283 56.155 multiply_cannon_loop 111 8.7 0.116 0.132 48.576 51.911 density_matrix_trs4 2 5.0 0.205 1.665 49.948 51.343 mp_waitall_1 7281 11.0 22.751 31.254 22.751 31.254 ls_scf_init_scf 1 4.0 0.000 0.001 30.361 30.368 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.949 29.021 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.553 26.567 multiply_cannon_multrec 888 9.7 12.769 15.236 21.420 24.667 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.285 20.246 make_m2s 222 7.7 0.006 0.007 16.484 17.153 make_images 222 8.7 1.582 1.853 16.446 17.115 make_images_data 222 9.7 0.004 0.004 9.609 10.617 hybrid_alltoall_any 227 10.6 0.641 2.960 9.089 10.351 mp_sum_l 887 5.1 6.029 9.515 6.029 9.515 dbcsr_mm_accdrv_process 3754 10.4 0.305 0.493 8.177 9.468 dbcsr_mm_accdrv_process_sort 3754 11.4 7.742 8.975 7.742 8.975 multiply_cannon_sync_h2d 888 9.7 6.063 8.120 6.063 8.120 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.482 6.864 mp_irecv_dv 2335 11.1 2.464 6.805 2.464 6.805 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.501 6.649 apply_matrix_preconditioner 6 5.3 0.000 0.000 6.091 6.497 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.561 6.385 arnoldi_extremal 4 6.8 0.000 0.000 5.069 5.088 arnoldi_normal_ev 4 7.8 0.001 0.004 5.069 5.087 build_subspace 16 8.4 0.014 0.020 4.763 4.770 calculate_norms 1584 9.8 4.339 4.765 4.339 4.765 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.321 4.623 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.435 3.769 mp_allgather_i34 111 8.7 0.817 3.744 0.817 3.744 ls_scf_post 1 4.0 0.000 0.001 3.669 3.673 dbcsr_matrix_vector_mult_local 304 10.0 3.019 3.596 3.022 3.598 ls_scf_store_result 1 5.0 0.000 0.000 3.388 3.492 dbcsr_complete_redistribute 5 7.6 0.361 0.383 2.314 2.618 matrix_ls_to_qs 2 6.0 0.000 0.000 2.158 2.477 dbcsr_data_new 4116 9.9 2.106 2.453 2.106 2.453 matrix_decluster 2 7.0 0.000 0.000 2.046 2.360 dbcsr_sort_data 325 11.1 1.888 2.156 1.888 2.156 make_images_sizes 222 9.7 0.000 0.000 1.033 2.152 mp_alltoall_i44 222 10.7 1.032 2.151 1.032 2.151 mp_alltoall_i22 14 8.6 1.755 2.137 1.755 2.137 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.111 2.117 rebuild_ks_matrix 3 7.3 0.000 0.000 2.093 2.099 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.022 2.093 2.099 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.436000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2253.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 3.318088E+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 0.046 102.914 102.916 qs_energies 1 2.0 0.000 0.000 102.280 102.286 ls_scf 1 3.0 0.000 0.000 100.620 100.626 dbcsr_multiply_generic 111 6.7 0.017 0.021 84.237 84.544 multiply_cannon 111 7.7 0.044 0.099 56.057 60.647 ls_scf_main 1 4.0 0.000 0.000 58.715 58.722 multiply_cannon_loop 111 8.7 0.152 0.168 50.985 54.470 density_matrix_trs4 2 5.0 0.002 0.003 52.588 52.709 ls_scf_init_scf 1 4.0 0.000 0.000 38.707 38.710 ls_scf_init_matrix_S 1 5.0 0.000 0.000 37.492 37.554 matrix_sqrt_Newton_Schulz 2 6.5 0.005 0.011 34.973 34.987 mp_waitall_1 6369 11.0 23.147 34.338 23.147 34.338 multiply_cannon_multrec 1332 9.7 14.187 17.167 22.170 25.701 make_m2s 222 7.7 0.006 0.007 21.080 22.507 make_images 222 8.7 3.139 3.602 21.030 22.459 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.722 20.783 make_images_data 222 9.7 0.004 0.004 11.749 13.350 hybrid_alltoall_any 227 10.6 0.796 3.779 11.101 12.918 dbcsr_mm_accdrv_process 3641 10.4 0.279 0.478 7.618 9.155 mp_sum_l 887 5.1 5.523 9.033 5.523 9.033 dbcsr_mm_accdrv_process_sort 3641 11.4 7.176 8.656 7.176 8.656 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.555 7.484 compute_matrix_preconditioner 1 6.0 0.001 0.001 6.621 6.627 multiply_cannon_sync_h2d 1332 9.7 5.454 6.110 5.454 6.110 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.080 6.110 mp_irecv_dv 3229 10.9 2.055 6.025 2.055 6.025 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.506 5.874 arnoldi_extremal 4 6.8 0.000 0.000 5.287 5.305 arnoldi_normal_ev 4 7.8 0.001 0.005 5.287 5.305 build_subspace 16 8.4 0.014 0.021 4.912 4.919 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.566 4.719 mp_allgather_i34 111 8.7 2.195 4.609 2.195 4.609 calculate_norms 2376 9.8 4.160 4.527 4.160 4.527 acc_transpose_blocks 1332 9.7 0.008 0.008 4.049 4.521 acc_transpose_blocks_kernels 1332 10.7 0.018 0.018 3.937 4.408 jit_kernel_transpose 1 13.0 3.919 4.391 3.919 4.391 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 3.582 3.875 dbcsr_matrix_vector_mult_local 304 10.0 3.190 3.693 3.192 3.695 dbcsr_sort_data 658 11.4 3.075 3.502 3.075 3.502 dbcsr_special_finalize 555 9.7 0.006 0.007 2.831 3.233 dbcsr_merge_single_wm 555 10.7 0.541 0.658 2.823 3.224 ls_scf_post 1 4.0 0.000 0.000 3.198 3.204 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.003 3.057 ls_scf_store_result 1 5.0 0.000 0.000 2.944 2.999 dbcsr_data_release 10477 10.7 1.583 2.407 1.583 2.407 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=102.916000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2722.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 4.750426E+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 0.037 0.072 94.947 94.955 qs_energies 1 2.0 0.000 0.000 94.096 94.103 ls_scf 1 3.0 0.000 0.000 92.020 92.027 dbcsr_multiply_generic 111 6.7 0.017 0.018 73.547 73.707 multiply_cannon 111 7.7 0.072 0.154 55.322 58.899 ls_scf_main 1 4.0 0.000 0.000 56.058 56.062 multiply_cannon_loop 111 8.7 0.088 0.094 52.801 54.352 density_matrix_trs4 2 5.0 0.002 0.003 49.105 49.197 ls_scf_init_scf 1 4.0 0.000 0.000 32.528 32.529 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.272 31.299 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.172 29.189 mp_waitall_1 5436 11.0 23.918 28.417 23.918 28.417 multiply_cannon_multrec 444 9.7 13.813 16.326 20.951 23.532 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.514 14.817 make_m2s 222 7.7 0.005 0.005 13.548 14.439 make_images 222 8.7 2.036 2.469 13.481 14.369 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 5.935 13.490 hybrid_alltoall_any 227 10.6 0.801 3.816 8.110 9.706 make_images_data 222 9.7 0.003 0.004 8.285 9.521 multiply_cannon_sync_h2d 444 9.7 6.716 8.448 6.716 8.448 dbcsr_mm_accdrv_process 3003 10.4 0.353 0.406 6.834 7.948 dbcsr_mm_accdrv_process_sort 3003 11.4 6.471 7.555 6.471 7.555 arnoldi_extremal 4 6.8 0.000 0.000 5.823 5.840 arnoldi_normal_ev 4 7.8 0.001 0.005 5.823 5.840 build_subspace 16 8.4 0.015 0.020 5.429 5.438 mp_sum_l 887 5.1 2.679 4.694 2.679 4.694 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.408 4.572 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.171 4.372 dbcsr_matrix_vector_mult_local 304 10.0 3.697 4.160 3.699 4.163 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.523 3.984 mp_irecv_dv 1241 11.2 1.508 3.953 1.508 3.953 compute_matrix_preconditioner 1 6.0 0.001 0.002 3.843 3.853 calculate_norms 792 9.8 3.600 3.742 3.600 3.742 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.844 3.605 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.529 3.580 mp_allgather_i34 111 8.7 1.100 3.559 1.100 3.559 ls_scf_post 1 4.0 0.000 0.000 3.434 3.441 acc_transpose_blocks 444 9.7 0.003 0.003 3.419 3.424 acc_transpose_blocks_kernels 444 10.7 0.007 0.008 3.354 3.358 jit_kernel_transpose 1 13.0 3.347 3.351 3.347 3.351 make_images_sizes 222 9.7 0.000 0.000 0.882 3.342 mp_alltoall_i44 222 10.7 0.881 3.342 0.881 3.342 ls_scf_store_result 1 5.0 0.000 0.000 3.195 3.225 dbcsr_finalize 304 7.8 0.062 0.077 2.203 2.274 dbcsr_data_new 4608 9.7 1.803 2.242 1.803 2.242 dbcsr_merge_all 275 8.9 0.482 0.524 2.058 2.114 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.060 2.060 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.040 2.042 rebuild_ks_matrix 3 7.3 0.000 0.000 2.008 2.009 qs_ks_build_kohn_sham_matrix 3 8.3 0.006 0.007 2.008 2.009 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=94.955000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3764.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 8.869163E+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 0.070 0.089 106.839 106.840 qs_energies 1 2.0 0.000 0.000 105.489 105.520 ls_scf 1 3.0 0.000 0.000 102.497 102.527 dbcsr_multiply_generic 111 6.7 0.024 0.027 76.116 76.256 ls_scf_main 1 4.0 0.000 0.000 63.706 63.706 density_matrix_trs4 2 5.0 0.002 0.003 54.676 54.728 multiply_cannon 111 7.7 0.125 0.195 49.931 52.344 multiply_cannon_loop 111 8.7 0.099 0.106 46.813 47.404 ls_scf_init_scf 1 4.0 0.000 0.000 34.981 34.982 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.423 33.447 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 30.715 30.723 mp_waitall_1 4527 11.1 22.084 25.790 22.084 25.790 make_m2s 222 7.7 0.005 0.005 22.546 23.742 make_images 222 8.7 3.565 3.883 22.439 23.633 multiply_cannon_multrec 444 9.7 17.762 18.376 22.441 23.020 hybrid_alltoall_any 227 10.6 1.651 3.611 12.765 15.661 make_images_data 222 9.7 0.003 0.004 12.977 15.224 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.511 11.870 multiply_cannon_sync_h2d 444 9.7 8.789 8.874 8.789 8.874 arnoldi_extremal 4 6.8 0.000 0.000 7.378 7.388 arnoldi_normal_ev 4 7.8 0.002 0.009 7.377 7.388 build_subspace 16 8.4 0.026 0.036 6.837 6.850 dbcsr_matrix_vector_mult 304 9.0 0.017 0.035 5.461 5.622 dbcsr_matrix_vector_mult_local 304 10.0 5.044 5.363 5.046 5.366 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.155 5.247 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.882 5.133 dbcsr_mm_accdrv_process 1814 10.4 0.264 0.341 4.495 4.631 dbcsr_mm_accdrv_process_sort 1814 11.4 4.163 4.302 4.163 4.302 mp_allgather_i34 111 8.7 1.230 3.938 1.230 3.938 ls_scf_post 1 4.0 0.000 0.000 3.811 3.842 make_images_sizes 222 9.7 0.000 0.000 1.441 3.580 mp_alltoall_i44 222 10.7 1.440 3.579 1.440 3.579 ls_scf_store_result 1 5.0 0.000 0.000 3.517 3.556 calculate_norms 792 9.8 3.235 3.316 3.235 3.316 dbcsr_finalize 304 7.8 0.082 0.089 3.073 3.188 dbcsr_merge_all 275 8.9 0.885 0.922 2.858 2.964 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.962 2.962 dbcsr_complete_redistribute 5 7.6 1.429 1.470 2.728 2.834 dbcsr_sort_data 325 11.1 2.435 2.509 2.435 2.509 matrix_ls_to_qs 2 6.0 0.000 0.000 2.387 2.504 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.461 2.463 rebuild_ks_matrix 3 7.3 0.000 0.000 2.395 2.397 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.002 2.395 2.397 dbcsr_data_new 6591 9.6 1.880 2.362 1.880 2.362 dbcsr_new_transposed 4 7.5 0.241 0.254 2.304 2.329 dbcsr_frobenius_norm 74 6.6 2.057 2.141 2.205 2.237 dbcsr_add_d 103 6.2 0.000 0.000 2.124 2.208 dbcsr_add_anytype 103 7.2 0.858 0.890 2.124 2.208 dbcsr_data_release 12724 10.6 1.967 2.193 1.967 2.193 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.840000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6984.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/f5d7817782ccd5e60ecd1b01248fcc43fdeed73e_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 589.496320E+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 4805924. 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 0.090 0.155 240.157 240.176 qs_mol_dyn_low 1 2.0 0.005 0.014 238.765 238.846 qs_forces 5 3.8 0.004 0.005 238.655 238.659 qs_energies 5 4.8 0.009 0.100 235.362 235.393 scf_env_do_scf 5 5.8 0.001 0.017 219.559 219.561 scf_env_do_scf_inner_loop 105 6.6 0.005 0.047 191.718 191.723 qs_scf_new_mos 105 7.6 0.000 0.001 149.425 149.608 qs_scf_loop_do_ot 105 8.6 0.001 0.001 149.425 149.607 ot_scf_mini 105 9.6 0.005 0.043 139.498 139.647 dbcsr_multiply_generic 1445 12.2 0.129 0.138 134.718 135.153 multiply_cannon 1445 13.2 0.274 0.283 115.141 117.101 multiply_cannon_loop 1445 14.2 2.818 2.974 113.362 114.632 velocity_verlet 4 3.0 0.001 0.001 108.385 108.386 ot_mini 105 10.6 0.001 0.006 60.649 60.796 qs_ot_get_p 112 10.4 0.003 0.033 48.836 49.149 multiply_cannon_multrec 69360 15.2 29.821 34.362 39.698 44.695 mp_waitall_1 488190 16.1 35.535 42.230 35.535 42.230 qs_ot_get_derivative 55 11.6 0.001 0.001 38.895 39.031 qs_ot_p2m_diag 40 11.0 0.021 0.041 37.617 37.699 cp_dbcsr_syevd 40 12.0 0.002 0.019 34.229 34.232 multiply_cannon_sync_h2d 69360 15.2 29.015 33.914 29.015 33.914 multiply_cannon_metrocomm3 69360 15.2 0.203 0.214 25.973 33.019 rebuild_ks_matrix 110 8.4 0.000 0.000 30.632 30.804 qs_ks_build_kohn_sham_matrix 110 9.4 0.018 0.065 30.632 30.803 cp_fm_syevd 40 13.0 0.001 0.029 28.955 29.087 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.984 28.145 init_scf_loop 7 6.6 0.001 0.018 27.794 27.803 cp_fm_redistribute_end 40 14.0 12.054 24.059 12.059 24.060 cp_fm_syevd_base 40 14.0 11.992 24.000 11.992 24.000 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.089 23.385 apply_single 62 13.6 0.000 0.000 23.089 23.384 prepare_preconditioner 7 7.6 0.000 0.000 22.495 22.519 make_preconditioner 7 8.6 0.001 0.018 22.495 22.519 ot_new_cg_direction 55 11.6 0.001 0.001 21.094 21.095 qs_rho_update_rho_low 110 7.6 0.001 0.001 18.205 18.573 calculate_rho_elec 110 8.6 0.030 0.032 18.205 18.572 make_full_inverse_cholesky 7 9.6 0.000 0.011 15.092 15.159 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.767 14.948 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.980 14.075 pw_transfer 1645 12.4 0.079 0.101 13.301 13.557 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 13.162 13.418 mp_sum_l 4764 12.2 12.407 13.269 12.407 13.269 init_scf_run 5 5.8 0.000 0.001 13.039 13.040 scf_env_initial_rho_setup 5 6.8 0.001 0.003 13.039 13.040 density_rs2pw 110 9.6 0.005 0.015 12.320 12.798 fft_wrap_pw1pw2_240 915 15.0 0.820 0.909 11.643 11.916 calculate_dm_sparse 110 9.5 0.000 0.001 11.528 11.735 qs_vxc_create 110 10.4 0.002 0.004 11.237 11.283 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.540 10.622 dbcsr_mm_accdrv_process 154766 15.8 6.219 6.410 9.745 10.566 cp_fm_cholesky_invert 7 10.6 10.489 10.498 10.489 10.498 fft3d_pb 915 16.0 2.379 2.693 9.865 10.143 check_diag 80 13.5 8.684 8.936 9.620 9.751 sum_up_and_integrate 60 10.3 0.001 0.003 9.166 9.225 integrate_v_rspace 60 11.3 0.002 0.042 9.149 9.209 transfer_rs2pw 445 10.6 0.007 0.008 8.573 9.087 acc_transpose_blocks 69360 15.2 0.364 0.379 8.060 8.613 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 8.073 8.090 xc_rho_set_and_dset_create 110 12.4 0.075 0.096 7.777 8.033 calculate_first_density_matrix 1 7.0 0.000 0.004 7.955 7.969 multiply_cannon_metrocomm1 69360 15.2 0.095 0.102 4.697 7.779 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.578 7.644 xc_vxc_pw_create 60 11.3 0.039 0.050 7.582 7.626 make_m2s 2890 13.2 0.078 0.089 6.737 7.309 mp_alltoall_z22v 2340 17.7 6.931 7.213 6.931 7.213 make_images 2890 14.2 0.238 0.259 6.629 7.202 make_full_single_inverse 7 9.6 0.001 0.029 7.123 7.148 xc_pw_derive 510 13.4 0.005 0.007 6.903 6.978 acc_transpose_blocks_kernels 69360 16.2 0.849 0.899 5.312 5.759 mp_waitany 7680 13.5 4.532 5.380 4.532 5.380 potential_pw2rs 60 12.3 0.002 0.004 5.285 5.330 jit_kernel_transpose 5 15.0 4.463 4.877 4.463 4.877 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=240.176000, 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: f5d7817782ccd5e60ecd1b01248fcc43fdeed73e Summary: empty Status: OK