=== 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: ba7248420c2af75c068fa8700be06de812d89520 ################# 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 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.3.2, 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. # # Author: Matthias Krack (12.01.2023) # # \ 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 ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_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)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib 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__HDF5 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 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_prefixed_scalapack.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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/01 job id: 44419204 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/02 job id: 44419290 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/03 job id: 44419292 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/04 job id: 44419293 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/05 job id: 44419297 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/06 job id: 44419298 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/07 job id: 44419299 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/08 job id: 44419300 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/09 job id: 44419303 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/10 job id: 44419304 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/11 job id: 44419308 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/12 job id: 44419310 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/13 job id: 44419321 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/14 job id: 44419323 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/15 job id: 44419324 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/16 job id: 44419325 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/17 job id: 44419326 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/18 job id: 44419328 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/19 job id: 44419329 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/20 job id: 44419330 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/21 job id: 44419331 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/22 job id: 44419332 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/23 job id: 44419333 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/24 job id: 44419334 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/25 job id: 44419335 --- 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/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/26 job id: 44419336 --- 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 ~~~~~~~ === 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]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_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 344 9. 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 0.021 0.034 134.570 134.571 farming_run 1 2.0 133.813 133.814 134.543 134.548 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.472692E+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 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 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 0.011 0.025 116.018 116.018 qs_energies 1 2.0 0.000 0.000 115.743 115.746 mp2_main 1 3.0 0.000 0.000 113.842 113.845 mp2_gpw_main 1 4.0 0.034 0.042 112.971 112.975 mp2_ri_gpw_compute_in 1 5.0 0.201 0.207 94.177 94.281 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.404 55.507 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.167 41.710 46.657 get_2c_integrals 1 6.0 0.000 0.000 37.854 38.548 integrate_v_rspace 273 8.0 0.438 0.454 25.103 29.699 pw_transfer 6555 10.6 0.375 0.392 27.427 28.128 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 26.084 26.599 grid_integrate_task_list 273 9.0 20.916 25.952 20.916 25.952 fft_wrap_pw1pw2_100 2178 12.4 1.214 1.456 23.603 24.136 compute_2c_integrals 1 7.0 0.003 0.003 19.885 19.931 compute_2c_integrals_loop_lm 1 8.0 0.007 0.008 18.948 19.596 mp2_eri_2c_integrate_gpw 1 9.0 2.378 2.434 18.941 19.588 rpa_ri_compute_en 1 5.0 0.000 0.000 18.673 18.797 cp_fm_cholesky_decompose 12 8.2 17.909 18.616 17.909 18.616 cholesky_decomp 1 7.0 0.000 0.000 16.807 17.519 fft3d_s 5443 13.4 16.135 16.382 16.158 16.405 ao_to_mo_and_store_B_mult_1 272 7.0 10.849 15.565 10.849 15.565 calculate_wavefunction 272 8.0 5.418 5.508 12.538 13.184 rpa_num_int 1 6.0 0.001 0.011 10.615 10.616 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.542 10.605 calc_mat_Q 8 8.0 0.000 0.000 9.388 9.516 contract_S_to_Q 8 9.0 0.000 0.000 8.808 8.936 calc_potential_gpw 544 9.5 0.005 0.006 8.261 8.714 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.253 8.653 parallel_gemm_fm 14 9.1 0.000 0.000 8.393 8.512 parallel_gemm_fm_cosma 14 10.1 8.393 8.512 8.393 8.512 potential_pw2rs 545 10.0 0.106 0.108 7.708 8.394 create_integ_mat 1 6.0 0.036 0.044 7.856 7.857 collocate_single_gaussian 272 10.0 0.039 0.042 7.487 7.821 array2fm 1 7.0 0.000 0.000 6.826 7.291 pw_scatter_s 2720 13.7 4.434 4.581 4.434 4.581 pw_gather_s 2722 13.2 3.890 4.223 3.890 4.223 array2fm_buffer_send 1 8.0 2.992 3.180 2.992 3.180 pw_poisson_solve 545 10.5 1.127 1.177 2.177 2.368 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.971731, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2744.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_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 344 10. 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 0.029 0.041 397.048 397.051 farming_run 1 2.0 396.249 396.260 397.003 397.007 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.223315E+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 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 2403167. 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 0.013 0.037 209.911 209.913 qs_energies 1 2.0 0.000 0.000 209.688 209.696 scf_env_do_scf 1 3.0 0.000 0.000 106.799 106.800 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.908 105.915 rebuild_ks_matrix 4 6.0 0.000 0.000 105.907 105.914 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.063 105.907 105.914 hfx_ks_matrix 4 8.0 0.001 0.001 105.524 105.528 integrate_four_center 4 9.0 0.144 0.461 105.524 105.527 mp2_main 1 3.0 0.002 0.017 102.596 102.604 mp2_gpw_main 1 4.0 0.037 0.082 101.760 101.771 integrate_four_center_main 4 10.0 0.068 0.476 96.967 99.872 integrate_four_center_bin 263 11.0 96.899 99.870 96.899 99.870 init_scf_loop 1 4.0 0.000 0.000 92.417 92.417 mp2_ri_gpw_compute_in 1 5.0 0.068 0.102 74.938 75.956 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.028 54.537 55.557 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.160 42.239 47.335 integrate_v_rspace 95 8.0 0.398 0.569 28.613 33.535 pw_transfer 2240 10.6 0.146 0.172 29.939 30.411 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.941 29.419 grid_integrate_task_list 95 9.0 23.898 29.026 23.898 29.026 mp2_ri_gpw_compute_en 1 5.0 0.055 0.063 26.638 28.324 ao_to_mo_and_store_B_mult_1 91 7.0 10.612 27.737 10.612 27.737 fft_wrap_pw1pw2_100 730 12.4 1.273 1.442 26.646 27.109 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.840 1.897 24.904 24.914 get_2c_integrals 1 6.0 0.004 0.033 20.286 20.337 compute_2c_integrals 1 7.0 0.003 0.005 19.265 19.274 compute_2c_integrals_loop_lm 1 8.0 0.004 0.026 18.879 19.140 mp2_eri_2c_integrate_gpw 1 9.0 1.730 1.845 18.875 19.139 fft3d_s 1823 13.4 18.414 18.693 18.428 18.706 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.381 14.381 calculate_wavefunction 91 8.0 2.016 2.054 9.744 10.002 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.578 8.788 9.223 potential_pw2rs 186 10.0 0.033 0.035 8.619 9.195 local_gemm 172 8.0 8.232 8.662 8.232 8.662 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.247 8.533 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.524 7.891 8.319 calc_potential_gpw 182 9.5 0.002 0.002 7.939 8.241 collocate_single_gaussian 91 10.0 0.017 0.037 7.893 8.090 mp2_ri_gpw_compute_en_ener 172 7.0 6.340 6.424 6.340 6.424 mp_sendrecv_dm3 2068 8.0 5.922 6.358 5.922 6.358 mp_sync 38 10.4 3.260 6.056 3.260 6.056 pw_gather_s 912 13.2 4.915 5.448 4.915 5.448 pw_scatter_s 910 13.7 3.942 4.399 3.942 4.399 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.753550, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.354048E+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 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 578578. 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 0.016 0.033 55.059 55.061 qs_mol_dyn_low 1 2.0 0.003 0.004 54.760 54.769 qs_forces 11 3.9 0.002 0.002 54.699 54.700 qs_energies 11 4.9 0.001 0.003 53.139 53.160 scf_env_do_scf 11 5.9 0.001 0.006 47.106 47.106 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.900 44.901 dbcsr_multiply_generic 2286 12.5 0.095 0.100 35.289 35.799 qs_scf_new_mos 108 7.5 0.000 0.001 34.315 34.609 qs_scf_loop_do_ot 108 8.5 0.001 0.001 34.314 34.609 ot_scf_mini 108 9.5 0.002 0.003 32.516 32.716 multiply_cannon 2286 13.5 0.178 0.184 26.812 28.368 multiply_cannon_loop 2286 14.5 1.491 1.551 26.065 27.639 velocity_verlet 10 3.0 0.002 0.002 27.594 27.595 ot_mini 108 10.5 0.001 0.001 20.750 20.989 qs_ot_get_derivative 108 11.5 0.001 0.001 17.655 17.850 mp_waitall_1 245248 16.5 9.645 16.292 9.645 16.292 multiply_cannon_metrocomm3 54864 15.5 0.067 0.072 6.661 13.833 multiply_cannon_multrec 54864 15.5 4.134 6.432 7.442 10.802 rebuild_ks_matrix 119 8.3 0.000 0.000 8.143 8.298 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 8.142 8.297 mp_sum_l 7207 12.9 6.348 8.074 6.348 8.074 multiply_cannon_sync_h2d 54864 15.5 5.773 7.397 5.773 7.397 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.166 7.306 qs_ot_get_p 119 10.4 0.001 0.001 6.941 7.203 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.138 6.612 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.859 6.009 init_scf_run 11 5.9 0.000 0.001 4.732 4.733 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.732 4.732 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.483 4.707 calculate_rho_elec 119 8.7 0.011 0.017 4.483 4.706 sum_up_and_integrate 119 10.3 0.012 0.015 4.658 4.666 integrate_v_rspace 119 11.3 0.002 0.002 4.646 4.655 dbcsr_mm_accdrv_process 76910 16.1 1.167 1.799 3.231 4.487 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.806 3.871 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 2.231 3.744 calculate_dm_sparse 119 9.5 0.000 0.000 3.018 3.167 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.932 3.138 apply_single 119 13.6 0.000 0.000 2.932 3.138 rs_pw_transfer 974 11.9 0.011 0.012 2.908 3.032 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.905 2.905 ot_diis_step 108 11.5 0.006 0.006 2.773 2.773 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.676 2.677 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.613 2.672 cp_fm_redistribute_end 50 14.0 2.432 2.650 2.437 2.651 jit_kernel_multiply 13 15.8 2.002 2.608 2.002 2.608 density_rs2pw 119 9.7 0.004 0.004 2.413 2.577 cp_fm_diag_elpa_base 50 14.0 0.212 2.558 0.213 2.568 calculate_first_density_matrix 1 7.0 0.000 0.001 2.497 2.501 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.297 2.299 mp_sum_d 4125 12.0 1.525 2.194 1.525 2.194 acc_transpose_blocks 54864 15.5 0.230 0.248 1.767 2.193 init_scf_loop 11 6.9 0.000 0.003 2.180 2.184 wfi_extrapolate 11 7.9 0.001 0.001 2.172 2.172 grid_integrate_task_list 119 12.3 2.020 2.123 2.020 2.123 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.007 2.050 pw_transfer 1439 11.6 0.053 0.059 1.966 2.046 potential_pw2rs 119 12.3 0.004 0.004 1.985 2.004 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.889 1.973 make_m2s 4572 13.5 0.054 0.057 1.816 1.877 make_images 4572 14.5 0.133 0.138 1.733 1.793 fft3d_ps 1201 14.6 0.372 0.474 1.660 1.735 mp_alltoall_d11v 2130 13.8 1.427 1.655 1.427 1.655 fft_wrap_pw1pw2_140 487 13.2 0.082 0.096 1.439 1.525 mp_waitany 12084 13.8 1.296 1.503 1.296 1.503 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.395 1.422 grid_collocate_task_list 119 9.7 1.289 1.359 1.289 1.359 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.861 1.290 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.061000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.272727, yerr=1.052349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_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 487.882752E+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 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 1030387. 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 0.014 0.038 39.610 39.612 qs_mol_dyn_low 1 2.0 0.005 0.019 39.359 39.366 qs_forces 11 3.9 0.002 0.003 39.295 39.301 qs_energies 11 4.9 0.001 0.001 37.583 37.591 scf_env_do_scf 11 5.9 0.001 0.006 32.243 32.243 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.664 29.664 dbcsr_multiply_generic 2286 12.5 0.100 0.104 21.544 21.892 qs_scf_new_mos 108 7.5 0.001 0.001 20.167 20.423 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.167 20.422 ot_scf_mini 108 9.5 0.002 0.003 19.272 19.453 velocity_verlet 10 3.0 0.001 0.002 18.607 18.608 multiply_cannon 2286 13.5 0.206 0.215 16.537 18.102 multiply_cannon_loop 2286 14.5 0.901 0.981 15.411 16.737 ot_mini 108 10.5 0.001 0.001 11.843 12.084 mp_waitall_1 200699 16.5 5.822 11.175 5.822 11.175 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.219 9.692 qs_ot_get_derivative 108 11.5 0.001 0.001 9.382 9.565 multiply_cannon_multrec 27432 15.5 1.960 4.474 5.981 8.824 rebuild_ks_matrix 119 8.3 0.000 0.000 7.397 7.540 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.396 7.540 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.537 6.666 dbcsr_mm_accdrv_process 47894 16.0 3.088 5.145 3.953 5.830 qs_ot_get_p 119 10.4 0.001 0.001 4.540 4.780 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.653 4.505 sum_up_and_integrate 119 10.3 0.024 0.026 4.365 4.370 integrate_v_rspace 119 11.3 0.002 0.002 4.341 4.347 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.049 4.121 apply_single 119 13.6 0.000 0.000 3.048 4.121 init_scf_run 11 5.9 0.000 0.001 4.069 4.069 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.069 4.069 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.983 4.018 calculate_rho_elec 119 8.7 0.021 0.024 3.982 4.017 mp_sum_l 7207 12.9 2.041 3.973 2.041 3.973 rs_pw_transfer 974 11.9 0.010 0.011 2.824 3.237 qs_ot_p2m_diag 50 11.0 0.008 0.012 3.050 3.066 multiply_cannon_sync_h2d 27432 15.5 2.210 2.877 2.210 2.877 make_m2s 4572 13.5 0.052 0.054 2.521 2.730 density_rs2pw 119 9.7 0.004 0.004 2.306 2.704 make_images 4572 14.5 0.199 0.236 2.432 2.640 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.618 2.618 init_scf_loop 11 6.9 0.000 0.000 2.546 2.547 calculate_first_density_matrix 1 7.0 0.000 0.000 2.500 2.501 ot_diis_step 108 11.5 0.011 0.011 2.411 2.412 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.254 2.255 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.150 2.244 cp_fm_redistribute_end 50 14.0 1.861 2.224 1.866 2.225 calculate_dm_sparse 119 9.5 0.000 0.000 2.089 2.163 cp_fm_diag_elpa_base 50 14.0 0.344 2.110 0.357 2.163 pw_transfer 1439 11.6 0.065 0.072 2.097 2.133 potential_pw2rs 119 12.3 0.006 0.006 2.083 2.097 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.048 2.051 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.005 2.043 jit_kernel_multiply 10 16.1 0.813 2.024 0.813 2.024 grid_integrate_task_list 119 12.3 1.835 1.931 1.835 1.931 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.825 1.865 fft3d_ps 1201 14.6 0.512 0.564 1.705 1.736 prepare_preconditioner 11 7.9 0.000 0.000 1.608 1.637 make_preconditioner 11 8.9 0.000 0.000 1.608 1.637 fft_wrap_pw1pw2_140 487 13.2 0.079 0.086 1.525 1.566 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.504 1.559 acc_transpose_blocks 27432 15.5 0.108 0.114 1.223 1.557 make_images_data 4572 15.5 0.045 0.051 1.141 1.529 wfi_extrapolate 11 7.9 0.001 0.001 1.511 1.511 hybrid_alltoall_any 4725 16.4 0.051 0.112 0.995 1.457 grid_collocate_task_list 119 9.7 1.230 1.367 1.230 1.367 mp_alltoall_d11v 2130 13.8 1.231 1.335 1.231 1.335 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.306 1.314 mp_allgather_i34 2286 14.5 0.570 1.309 0.570 1.309 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.165 1.211 mp_sum_d 4125 12.0 0.577 1.015 0.577 1.015 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.996 1.002 mp_waitany 5720 13.7 0.550 0.992 0.550 0.992 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.148 0.576 0.968 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.919 0.934 acc_transpose_blocks_kernels 27432 16.5 0.181 0.271 0.696 0.931 mp_alltoall_z22v 1201 16.6 0.762 0.873 0.762 0.873 rs_pw_transfer_PW2RS_50 119 14.3 0.589 0.608 0.778 0.847 make_images_sizes 4572 15.5 0.005 0.005 0.571 0.819 mp_alltoall_i44 4572 16.5 0.566 0.814 0.566 0.814 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.612000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.727273, yerr=1.762793 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_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 524.107776E+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 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 2569535. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 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 0.019 0.033 33.312 33.314 qs_mol_dyn_low 1 2.0 0.003 0.004 32.932 32.940 qs_forces 11 3.9 0.002 0.004 32.866 32.867 qs_energies 11 4.9 0.001 0.003 31.261 31.263 scf_env_do_scf 11 5.9 0.001 0.002 26.217 26.217 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.562 23.562 dbcsr_multiply_generic 2286 12.5 0.098 0.103 16.746 16.819 velocity_verlet 10 3.0 0.001 0.002 15.705 15.706 qs_scf_new_mos 108 7.5 0.001 0.001 15.066 15.085 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.066 15.084 ot_scf_mini 108 9.5 0.002 0.003 14.341 14.360 multiply_cannon 2286 13.5 0.194 0.200 13.355 14.163 multiply_cannon_loop 2286 14.5 0.637 0.671 12.546 13.366 ot_mini 108 10.5 0.001 0.001 8.845 8.869 qs_ot_get_derivative 108 11.5 0.001 0.001 7.344 7.363 multiply_cannon_multrec 18288 15.5 1.933 2.826 6.971 7.232 rebuild_ks_matrix 119 8.3 0.000 0.000 6.563 6.577 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 6.563 6.577 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.802 5.815 dbcsr_mm_accdrv_process 38222 16.0 4.275 5.698 4.956 5.752 sum_up_and_integrate 119 10.3 0.029 0.030 4.103 4.119 integrate_v_rspace 119 11.3 0.002 0.003 4.073 4.089 mp_waitall_1 158411 16.6 2.717 3.899 2.717 3.899 init_scf_run 11 5.9 0.000 0.001 3.780 3.780 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.780 3.780 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.627 3.630 calculate_rho_elec 119 8.7 0.030 0.031 3.626 3.629 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.790 3.413 qs_ot_get_p 119 10.4 0.001 0.001 3.323 3.334 rs_pw_transfer 974 11.9 0.009 0.010 2.538 2.824 init_scf_loop 11 6.9 0.001 0.003 2.634 2.634 calculate_first_density_matrix 1 7.0 0.000 0.002 2.522 2.523 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.469 2.437 density_rs2pw 119 9.7 0.004 0.004 2.149 2.394 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.042 2.359 apply_single 119 13.6 0.000 0.000 2.041 2.359 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.223 2.228 jit_kernel_multiply 10 15.9 0.630 2.076 0.630 2.076 pw_transfer 1439 11.6 0.066 0.071 2.023 2.035 make_m2s 4572 13.5 0.044 0.045 1.868 2.021 cp_dbcsr_syevd 50 12.0 0.003 0.004 1.958 1.959 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.930 1.943 calculate_dm_sparse 119 9.5 0.000 0.000 1.927 1.937 make_images 4572 14.5 0.190 0.201 1.783 1.934 grid_integrate_task_list 119 12.3 1.802 1.889 1.802 1.889 potential_pw2rs 119 12.3 0.007 0.008 1.860 1.868 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.866 1.868 prepare_preconditioner 11 7.9 0.000 0.001 1.821 1.823 make_preconditioner 11 8.9 0.000 0.001 1.821 1.823 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.673 1.755 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.710 1.715 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.705 1.706 cp_fm_redistribute_end 50 14.0 1.271 1.683 1.272 1.683 cp_fm_diag_elpa_base 50 14.0 0.394 1.593 0.409 1.642 fft3d_ps 1201 14.6 0.523 0.539 1.610 1.624 multiply_cannon_sync_h2d 18288 15.5 1.380 1.595 1.380 1.595 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.585 1.595 fft_wrap_pw1pw2_140 487 13.2 0.089 0.093 1.512 1.525 mp_sum_l 7207 12.9 1.141 1.483 1.141 1.483 ot_diis_step 108 11.5 0.011 0.011 1.480 1.481 grid_collocate_task_list 119 9.7 1.214 1.364 1.214 1.364 acc_transpose_blocks 18288 15.5 0.075 0.077 1.326 1.353 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.301 1.305 wfi_extrapolate 11 7.9 0.001 0.001 1.210 1.210 qs_energies_init_hamiltonians 11 5.9 0.003 0.019 1.017 1.018 make_images_data 4572 15.5 0.045 0.049 0.814 0.986 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.890 0.914 acc_transpose_blocks_kernels 18288 16.5 0.210 0.221 0.888 0.912 mp_waitany 9880 13.7 0.592 0.895 0.592 0.895 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.696 0.878 rs_pw_transfer_RS2PW_140 130 11.5 0.120 0.123 0.569 0.858 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.833 0.835 cp_fm_cholesky_invert 11 10.9 0.784 0.789 0.784 0.789 mp_alltoall_d11v 2130 13.8 0.692 0.788 0.692 0.788 mp_alltoall_z22v 1201 16.6 0.682 0.775 0.682 0.775 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.662 0.736 jit_kernel_transpose 5 15.6 0.677 0.694 0.677 0.694 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.314000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.000000, yerr=2.860388 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_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 555.929600E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.032 36.700 36.702 qs_mol_dyn_low 1 2.0 0.003 0.004 36.366 36.374 qs_forces 11 3.9 0.001 0.002 36.304 36.305 qs_energies 11 4.9 0.019 0.022 34.541 34.548 scf_env_do_scf 11 5.9 0.001 0.002 29.141 29.142 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.524 25.525 dbcsr_multiply_generic 2286 12.5 0.101 0.103 18.756 18.924 velocity_verlet 10 3.0 0.004 0.005 18.547 18.549 qs_scf_new_mos 108 7.5 0.001 0.001 16.719 16.779 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.719 16.779 ot_scf_mini 108 9.5 0.002 0.003 15.759 15.816 multiply_cannon 2286 13.5 0.228 0.259 14.888 15.467 multiply_cannon_loop 2286 14.5 0.937 0.967 13.902 14.401 ot_mini 108 10.5 0.001 0.001 9.625 9.696 multiply_cannon_multrec 27432 15.5 2.343 3.017 8.892 9.242 qs_ot_get_derivative 108 11.5 0.001 0.001 7.783 7.842 dbcsr_mm_accdrv_process 47916 15.9 5.556 7.429 6.457 7.672 rebuild_ks_matrix 119 8.3 0.000 0.000 6.788 6.855 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.788 6.854 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.026 6.078 init_scf_run 11 5.9 0.000 0.001 4.029 4.029 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.029 4.029 sum_up_and_integrate 119 10.3 0.035 0.038 3.972 3.978 integrate_v_rspace 119 11.3 0.003 0.010 3.936 3.944 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.739 3.772 calculate_rho_elec 119 8.7 0.040 0.046 3.738 3.772 init_scf_loop 11 6.9 0.000 0.000 3.593 3.593 qs_ot_get_p 119 10.4 0.001 0.001 3.496 3.576 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.756 3.250 prepare_preconditioner 11 7.9 0.000 0.000 2.716 2.723 make_preconditioner 11 8.9 0.000 0.000 2.716 2.723 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.322 2.646 mp_waitall_1 137007 16.6 1.974 2.625 1.974 2.625 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.150 2.611 apply_single 119 13.6 0.000 0.000 2.150 2.611 calculate_first_density_matrix 1 7.0 0.000 0.000 2.592 2.595 rs_pw_transfer 974 11.9 0.009 0.009 2.210 2.389 make_m2s 4572 13.5 0.054 0.056 2.283 2.381 density_rs2pw 119 9.7 0.004 0.004 2.131 2.314 make_images 4572 14.5 0.268 0.329 2.176 2.272 calculate_dm_sparse 119 9.5 0.000 0.000 2.167 2.234 pw_transfer 1439 11.6 0.066 0.072 2.171 2.212 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.160 2.170 jit_kernel_multiply 10 15.8 0.841 2.135 0.841 2.135 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.078 2.124 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.079 2.080 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.963 1.992 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.887 1.904 grid_integrate_task_list 119 12.3 1.820 1.903 1.820 1.903 cp_dbcsr_syevd 50 12.0 0.005 0.014 1.839 1.840 ot_diis_step 108 11.5 0.012 0.012 1.803 1.804 fft3d_ps 1201 14.6 0.559 0.609 1.753 1.791 fft_wrap_pw1pw2_140 487 13.2 0.088 0.095 1.688 1.739 potential_pw2rs 119 12.3 0.008 0.009 1.725 1.730 mp_sum_l 7207 12.9 1.157 1.725 1.157 1.725 acc_transpose_blocks 27432 15.5 0.111 0.115 1.552 1.582 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.569 1.573 cp_fm_redistribute_end 50 14.0 1.034 1.534 1.035 1.535 cp_fm_diag_elpa_base 50 14.0 0.477 1.459 0.496 1.500 wfi_extrapolate 11 7.9 0.001 0.001 1.382 1.383 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.329 1.343 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.850 1.335 grid_collocate_task_list 119 9.7 1.220 1.327 1.220 1.327 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.128 1.147 cp_fm_upper_to_full 72 13.5 0.800 1.109 0.800 1.109 dbcsr_complete_redistribute 329 12.2 0.125 0.154 0.834 1.108 qs_energies_init_hamiltonians 11 5.9 0.009 0.010 1.086 1.088 multiply_cannon_sync_h2d 27432 15.5 0.993 1.068 0.993 1.068 make_images_data 4572 15.5 0.045 0.050 0.889 1.008 acc_transpose_blocks_kernels 27432 16.5 0.266 0.276 0.931 0.945 mp_alltoall_d11v 2130 13.8 0.797 0.918 0.797 0.918 hybrid_alltoall_any 4725 16.4 0.062 0.153 0.766 0.909 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.791 0.872 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.854 0.867 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.591 0.855 cp_fm_cholesky_invert 11 10.9 0.849 0.853 0.849 0.853 mp_alltoall_z22v 1201 16.6 0.776 0.816 0.776 0.816 mp_alltoall_i22 627 13.8 0.444 0.748 0.444 0.748 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.702000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.000000, yerr=3.490246 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_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 612.057088E+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 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 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 0.016 0.039 28.477 28.478 qs_mol_dyn_low 1 2.0 0.003 0.003 28.281 28.290 qs_forces 11 3.9 0.002 0.002 28.205 28.206 qs_energies 11 4.9 0.001 0.001 26.476 26.479 scf_env_do_scf 11 5.9 0.001 0.002 21.741 21.741 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.148 19.149 velocity_verlet 10 3.0 0.002 0.002 14.821 14.825 dbcsr_multiply_generic 2286 12.5 0.094 0.097 12.097 12.163 qs_scf_new_mos 108 7.5 0.001 0.001 10.941 10.961 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.941 10.961 ot_scf_mini 108 9.5 0.002 0.002 10.259 10.285 multiply_cannon 2286 13.5 0.229 0.238 9.541 10.030 multiply_cannon_loop 2286 14.5 0.330 0.342 8.628 8.790 rebuild_ks_matrix 119 8.3 0.000 0.000 6.107 6.126 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.107 6.126 multiply_cannon_multrec 9144 15.5 1.582 1.812 5.732 5.975 ot_mini 108 10.5 0.001 0.001 5.668 5.699 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.444 5.461 qs_ot_get_derivative 108 11.5 0.001 0.001 4.355 4.382 dbcsr_mm_accdrv_process 12550 15.8 3.025 4.060 4.050 4.149 sum_up_and_integrate 119 10.3 0.038 0.041 3.759 3.763 integrate_v_rspace 119 11.3 0.003 0.003 3.721 3.726 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.657 3.664 calculate_rho_elec 119 8.7 0.060 0.061 3.656 3.663 init_scf_run 11 5.9 0.000 0.001 3.285 3.285 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.285 3.285 qs_ot_get_p 119 10.4 0.001 0.001 2.799 2.846 init_scf_loop 11 6.9 0.000 0.000 2.569 2.571 pw_transfer 1439 11.6 0.066 0.069 2.236 2.245 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.142 2.152 density_rs2pw 119 9.7 0.004 0.004 1.980 2.121 calculate_first_density_matrix 1 7.0 0.000 0.000 2.105 2.107 mp_waitall_1 115863 16.7 1.479 1.978 1.479 1.978 make_m2s 4572 13.5 0.034 0.035 1.765 1.942 grid_integrate_task_list 119 12.3 1.854 1.928 1.854 1.928 rs_pw_transfer 974 11.9 0.008 0.009 1.740 1.892 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.874 1.876 make_images 4572 14.5 0.267 0.300 1.675 1.852 fft3d_ps 1201 14.6 0.565 0.577 1.802 1.814 prepare_preconditioner 11 7.9 0.000 0.000 1.810 1.814 make_preconditioner 11 8.9 0.000 0.000 1.810 1.814 fft_wrap_pw1pw2_140 487 13.2 0.086 0.090 1.752 1.764 jit_kernel_multiply 10 15.5 0.987 1.747 0.987 1.747 calculate_dm_sparse 119 9.5 0.000 0.000 1.710 1.729 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.696 1.724 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.679 1.679 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.558 1.571 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.570 1.571 potential_pw2rs 119 12.3 0.010 0.011 1.471 1.475 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.417 1.418 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.403 1.411 cp_fm_redistribute_end 50 14.0 0.708 1.396 0.709 1.397 grid_collocate_task_list 119 9.7 1.273 1.390 1.273 1.390 cp_fm_diag_elpa_base 50 14.0 0.643 1.323 0.687 1.381 ot_diis_step 108 11.5 0.012 0.013 1.300 1.300 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.226 1.227 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.171 1.194 apply_single 119 13.6 0.000 0.000 1.171 1.194 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.127 1.140 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.130 1.135 wfi_extrapolate 11 7.9 0.001 0.001 1.127 1.127 make_images_data 4572 15.5 0.039 0.042 0.827 1.051 hybrid_alltoall_any 4725 16.4 0.063 0.177 0.797 1.051 acc_transpose_blocks 9144 15.5 0.038 0.039 0.937 0.944 cp_fm_cholesky_invert 11 10.9 0.925 0.928 0.925 0.928 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.918 mp_alltoall_d11v 2130 13.8 0.778 0.875 0.778 0.875 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.340 0.794 multiply_cannon_sync_h2d 9144 15.5 0.717 0.793 0.717 0.793 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.787 0.789 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.739 0.747 mp_alltoall_z22v 1201 16.6 0.684 0.745 0.684 0.745 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.680 0.730 acc_transpose_blocks_kernels 9144 16.5 0.117 0.121 0.698 0.702 mp_allgather_i34 2286 14.5 0.246 0.657 0.246 0.657 yz_to_x 606 15.1 0.265 0.276 0.621 0.642 x_to_yz 595 16.2 0.276 0.288 0.604 0.608 jit_kernel_transpose 5 15.6 0.581 0.584 0.581 0.584 mp_waitany 5200 13.7 0.449 0.576 0.449 0.576 qs_create_task_list 11 7.9 0.000 0.001 0.547 0.573 generate_qs_task_list 11 8.9 0.194 0.214 0.547 0.572 dbcsr_complete_redistribute 329 12.2 0.169 0.177 0.544 0.572 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.478000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=581.909091, yerr=3.629539 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_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 754.327552E+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 63729. MP_Allreduce 10074 433. 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 0.014 0.036 41.324 41.325 qs_mol_dyn_low 1 2.0 0.003 0.003 41.125 41.132 qs_forces 11 3.9 0.002 0.004 41.063 41.064 qs_energies 11 4.9 0.001 0.001 39.075 39.079 scf_env_do_scf 11 5.9 0.001 0.002 33.353 33.353 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.541 25.543 velocity_verlet 10 3.0 0.002 0.002 23.425 23.430 dbcsr_multiply_generic 2286 12.5 0.102 0.103 17.319 17.485 qs_scf_new_mos 108 7.5 0.001 0.001 15.636 15.733 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.636 15.732 ot_scf_mini 108 9.5 0.002 0.002 14.549 14.647 multiply_cannon 2286 13.5 0.302 0.313 13.480 14.389 multiply_cannon_loop 2286 14.5 0.345 0.351 12.249 13.150 ot_mini 108 10.5 0.001 0.001 8.701 8.815 multiply_cannon_multrec 9144 15.5 3.413 4.737 8.519 8.610 init_scf_loop 11 6.9 0.000 0.000 7.785 7.787 rebuild_ks_matrix 119 8.3 0.000 0.000 7.385 7.522 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.014 7.385 7.521 prepare_preconditioner 11 7.9 0.000 0.000 6.804 6.817 make_preconditioner 11 8.9 0.000 0.000 6.804 6.817 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.676 6.799 qs_ot_get_derivative 108 11.5 0.001 0.001 6.689 6.785 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.369 6.694 dbcsr_mm_accdrv_process 12550 15.8 4.045 5.555 4.983 6.351 cp_fm_upper_to_full 72 14.2 3.152 4.528 3.152 4.528 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.322 4.327 calculate_rho_elec 119 8.7 0.118 0.121 4.322 4.327 sum_up_and_integrate 119 10.3 0.065 0.066 4.064 4.070 integrate_v_rspace 119 11.3 0.003 0.003 3.999 4.006 init_scf_run 11 5.9 0.000 0.001 3.680 3.680 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.680 3.680 qs_ot_get_p 119 10.4 0.001 0.001 3.133 3.268 mp_waitall_1 94719 16.7 2.306 3.242 2.306 3.242 pw_transfer 1439 11.6 0.069 0.069 2.849 2.854 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.403 2.853 dbcsr_complete_redistribute 329 12.2 0.288 0.292 1.969 2.786 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.750 2.756 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.656 2.454 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.207 2.437 apply_single 119 13.6 0.000 0.000 2.207 2.437 make_m2s 4572 13.5 0.038 0.038 2.273 2.423 fft3d_ps 1201 14.6 0.595 0.603 2.375 2.380 fft_wrap_pw1pw2_140 487 13.2 0.095 0.097 2.324 2.331 make_images 4572 14.5 0.351 0.382 2.152 2.302 mp_alltoall_i22 627 13.8 1.463 2.284 1.463 2.284 density_rs2pw 119 9.7 0.004 0.004 2.246 2.263 calculate_first_density_matrix 1 7.0 0.000 0.000 2.223 2.224 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.429 2.220 calculate_dm_sparse 119 9.5 0.000 0.000 2.153 2.171 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.250 2.156 grid_integrate_task_list 119 12.3 2.055 2.068 2.055 2.068 ot_diis_step 108 11.5 0.014 0.014 1.987 1.987 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.945 1.946 qs_ot_p2m_diag 50 11.0 0.042 0.043 1.900 1.902 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.781 1.782 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.675 1.725 mp_sum_l 7207 12.9 1.026 1.695 1.026 1.695 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.635 1.660 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.630 1.631 rs_pw_transfer 974 11.9 0.009 0.010 1.477 1.502 grid_collocate_task_list 119 9.7 1.487 1.494 1.487 1.494 potential_pw2rs 119 12.3 0.014 0.015 1.457 1.460 cp_fm_cholesky_invert 11 10.9 1.452 1.455 1.452 1.455 jit_kernel_multiply 8 15.5 0.911 1.411 0.911 1.411 wfi_extrapolate 11 7.9 0.001 0.001 1.388 1.389 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.147 1.352 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.343 1.343 cp_fm_diag_elpa_base 50 14.0 1.199 1.251 1.341 1.341 make_images_data 4572 15.5 0.043 0.046 1.094 1.284 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.212 1.221 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.129 1.148 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.118 1.132 mp_alltoall_d11v 2130 13.8 1.079 1.109 1.079 1.109 multiply_cannon_sync_h2d 9144 15.5 1.043 1.048 1.043 1.048 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.990 1.031 qs_create_task_list 11 7.9 0.001 0.001 0.953 0.968 generate_qs_task_list 11 8.9 0.383 0.402 0.952 0.968 yz_to_x 606 15.1 0.459 0.470 0.949 0.957 acc_transpose_blocks 9144 15.5 0.038 0.038 0.941 0.946 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.921 0.933 mp_alltoall_z22v 1201 16.6 0.829 0.870 0.829 0.870 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.325000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=705.272727, yerr=10.677852 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 503.623680E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.031 83.785 83.786 qs_mol_dyn_low 1 2.0 0.003 0.003 83.541 83.552 qs_forces 11 3.9 0.002 0.002 83.474 83.475 qs_energies 11 4.9 0.001 0.002 80.612 80.632 scf_env_do_scf 11 5.9 0.001 0.001 71.843 71.846 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 66.253 66.254 dbcsr_multiply_generic 2055 12.4 0.108 0.113 52.607 52.888 qs_scf_new_mos 99 7.5 0.000 0.001 48.852 48.979 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.851 48.979 ot_scf_mini 99 9.5 0.002 0.002 46.417 46.530 multiply_cannon 2055 13.4 0.173 0.180 42.800 43.691 multiply_cannon_loop 2055 14.4 1.543 1.587 41.877 42.793 velocity_verlet 10 3.0 0.002 0.003 42.160 42.161 ot_mini 99 10.5 0.001 0.001 28.051 28.179 qs_ot_get_derivative 99 11.5 0.001 0.001 21.170 21.300 multiply_cannon_multrec 49320 15.4 12.062 12.900 17.080 17.641 rebuild_ks_matrix 110 8.3 0.000 0.001 14.610 14.775 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.014 14.609 14.775 mp_waitall_1 220248 16.4 11.840 13.019 11.840 13.019 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.814 12.960 multiply_cannon_sync_h2d 49320 15.4 9.900 10.645 9.900 10.645 qs_ot_get_p 110 10.4 0.001 0.001 9.734 9.833 multiply_cannon_metrocomm3 49320 15.4 0.079 0.083 7.215 8.609 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.128 7.648 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.170 7.622 apply_single 110 13.6 0.000 0.001 7.169 7.622 sum_up_and_integrate 110 10.3 0.036 0.043 7.122 7.136 integrate_v_rspace 110 11.3 0.003 0.003 7.085 7.108 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.984 7.061 init_scf_run 11 5.9 0.000 0.001 6.753 6.753 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.753 6.753 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.598 6.746 calculate_rho_elec 110 8.6 0.021 0.026 6.597 6.745 ot_diis_step 99 11.5 0.005 0.006 6.547 6.547 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.483 6.514 mp_sum_l 6514 12.8 5.531 6.322 5.531 6.322 init_scf_loop 11 6.9 0.000 0.000 5.560 5.560 cp_dbcsr_syevd 48 12.0 0.003 0.004 5.401 5.402 dbcsr_mm_accdrv_process 87628 16.1 2.081 2.182 4.899 5.191 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.918 4.920 cp_fm_redistribute_end 48 14.0 4.293 4.893 4.296 4.894 cp_fm_diag_elpa_base 48 14.0 0.592 4.761 0.595 4.784 rs_pw_transfer 902 11.9 0.011 0.013 3.663 4.149 wfi_extrapolate 11 7.9 0.001 0.001 4.018 4.019 make_m2s 4110 13.4 0.061 0.066 3.813 3.908 calculate_dm_sparse 110 9.5 0.000 0.001 3.794 3.904 multiply_cannon_metrocomm1 49320 15.4 0.060 0.063 2.834 3.855 density_rs2pw 110 9.6 0.004 0.005 3.398 3.844 make_images 4110 14.4 0.177 0.190 3.717 3.817 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.585 3.589 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.385 3.465 prepare_preconditioner 11 7.9 0.000 0.000 3.402 3.420 make_preconditioner 11 8.9 0.000 0.000 3.402 3.420 grid_integrate_task_list 110 12.3 3.252 3.406 3.252 3.406 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.275 3.331 pw_transfer 1331 11.6 0.055 0.069 3.254 3.317 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.191 3.232 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.166 3.232 fft_wrap_pw1pw2_140 451 13.1 0.171 0.191 2.682 2.750 fft3d_ps 1111 14.6 0.784 0.892 2.694 2.747 potential_pw2rs 110 12.3 0.006 0.007 2.672 2.691 calculate_first_density_matrix 1 7.0 0.000 0.000 2.645 2.649 jit_kernel_multiply 13 15.9 2.534 2.550 2.534 2.550 mp_alltoall_d11v 2046 13.8 2.020 2.357 2.020 2.357 acc_transpose_blocks 49320 15.4 0.223 0.234 2.242 2.308 grid_collocate_task_list 110 9.6 2.086 2.289 2.086 2.289 mp_waitany 14300 13.8 1.786 2.266 1.786 2.266 mp_sum_d 3879 11.9 1.522 2.044 1.522 2.044 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.936 1.956 cp_fm_cholesky_invert 11 10.9 1.854 1.858 1.854 1.858 make_images_data 4110 15.4 0.043 0.046 1.698 1.826 hybrid_alltoall_any 4261 16.3 0.082 0.484 1.485 1.739 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.786000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=478.000000, yerr=3.104249 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_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 588.460032E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 2462973. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.037 68.957 68.958 qs_mol_dyn_low 1 2.0 0.003 0.003 68.735 68.745 qs_forces 11 3.9 0.002 0.002 68.668 68.668 qs_energies 11 4.9 0.001 0.001 65.328 65.332 scf_env_do_scf 11 5.9 0.000 0.001 56.800 56.803 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.070 49.071 dbcsr_multiply_generic 2055 12.4 0.115 0.120 37.647 37.865 velocity_verlet 10 3.0 0.001 0.002 36.460 36.461 qs_scf_new_mos 99 7.5 0.001 0.001 32.596 32.752 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.595 32.752 multiply_cannon 2055 13.4 0.223 0.247 31.005 32.361 ot_scf_mini 99 9.5 0.003 0.003 30.939 31.083 multiply_cannon_loop 2055 14.4 0.927 0.961 29.694 30.605 ot_mini 99 10.5 0.001 0.001 18.178 18.326 multiply_cannon_multrec 24660 15.4 7.653 9.171 13.832 15.418 rebuild_ks_matrix 110 8.3 0.000 0.001 13.905 14.026 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.904 14.026 qs_ot_get_derivative 99 11.5 0.001 0.001 12.380 12.529 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.256 12.366 mp_waitall_1 176588 16.5 7.610 10.119 7.610 10.119 multiply_cannon_sync_h2d 24660 15.4 7.001 7.998 7.001 7.998 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 5.133 7.716 init_scf_loop 11 6.9 0.000 0.000 7.693 7.693 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.482 7.164 apply_single 110 13.6 0.000 0.001 6.482 7.164 sum_up_and_integrate 110 10.3 0.051 0.058 6.705 6.717 integrate_v_rspace 110 11.3 0.002 0.003 6.653 6.666 dbcsr_mm_accdrv_process 52282 16.1 4.682 5.459 6.021 6.308 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.232 6.242 qs_ot_get_p 110 10.4 0.001 0.001 6.115 6.242 calculate_rho_elec 110 8.6 0.039 0.047 6.231 6.242 init_scf_run 11 5.9 0.000 0.001 6.179 6.179 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.179 6.179 ot_diis_step 99 11.5 0.010 0.010 5.738 5.739 prepare_preconditioner 11 7.9 0.000 0.000 5.653 5.670 make_preconditioner 11 8.9 0.000 0.000 5.653 5.670 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.242 5.404 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.649 5.328 make_m2s 4110 13.4 0.057 0.060 4.173 4.648 make_images 4110 14.4 0.396 0.441 4.063 4.535 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.214 4.236 density_rs2pw 110 9.6 0.004 0.004 3.385 3.953 pw_transfer 1331 11.6 0.067 0.074 3.792 3.938 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.685 3.833 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.785 3.785 rs_pw_transfer 902 11.9 0.012 0.013 3.023 3.643 wfi_extrapolate 11 7.9 0.001 0.001 3.556 3.556 grid_integrate_task_list 110 12.3 3.146 3.320 3.146 3.320 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.313 3.314 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.205 3.281 fft_wrap_pw1pw2_140 451 13.1 0.203 0.219 3.110 3.262 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.246 3.248 cp_fm_redistribute_end 48 14.0 2.428 3.220 2.430 3.221 fft3d_ps 1111 14.6 1.101 1.320 3.033 3.163 cp_fm_diag_elpa_base 48 14.0 0.757 3.076 0.788 3.163 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.963 3.012 calculate_dm_sparse 110 9.5 0.001 0.001 2.981 3.012 make_images_data 4110 15.4 0.047 0.051 2.332 2.804 hybrid_alltoall_any 4261 16.3 0.102 0.441 2.023 2.781 cp_fm_cholesky_invert 11 10.9 2.652 2.659 2.652 2.659 calculate_first_density_matrix 1 7.0 0.000 0.000 2.535 2.538 potential_pw2rs 110 12.3 0.008 0.009 2.506 2.520 grid_collocate_task_list 110 9.6 2.049 2.486 2.049 2.486 mp_sum_l 6514 12.8 1.845 2.425 1.845 2.425 jit_kernel_multiply 11 16.3 0.991 2.128 0.991 2.128 mp_alltoall_d11v 2046 13.8 1.696 2.071 1.696 2.071 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.966 1.989 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.846 1.847 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.814 1.825 mp_waitany 10164 13.8 1.205 1.821 1.205 1.821 multiply_cannon_metrocomm4 22605 15.4 0.072 0.076 0.786 1.653 mp_allgather_i34 2055 14.4 0.640 1.649 0.640 1.649 acc_transpose_blocks 24660 15.4 0.109 0.112 1.578 1.615 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.557 1.569 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.217 0.953 1.555 mp_irecv_dv 57340 16.2 0.662 1.538 0.662 1.538 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.367 1.481 dbcsr_complete_redistribute 325 12.2 0.241 0.296 1.127 1.401 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.958000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.545455, yerr=7.315782 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_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 659.763200E+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 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 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 0.015 0.031 61.106 61.107 qs_mol_dyn_low 1 2.0 0.003 0.004 60.830 60.841 qs_forces 11 3.9 0.002 0.003 60.728 60.729 qs_energies 11 4.9 0.001 0.003 57.508 57.511 scf_env_do_scf 11 5.9 0.001 0.001 49.494 49.494 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.775 40.776 velocity_verlet 10 3.0 0.001 0.002 33.256 33.259 dbcsr_multiply_generic 2055 12.4 0.109 0.113 28.363 28.602 qs_scf_new_mos 99 7.5 0.001 0.001 25.281 25.359 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.281 25.359 ot_scf_mini 99 9.5 0.003 0.004 24.061 24.159 multiply_cannon 2055 13.4 0.209 0.218 21.983 23.145 multiply_cannon_loop 2055 14.4 0.617 0.631 20.794 21.759 ot_mini 99 10.5 0.001 0.001 13.556 13.655 rebuild_ks_matrix 110 8.3 0.000 0.000 12.623 12.741 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 12.622 12.741 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.126 11.232 multiply_cannon_multrec 16440 15.4 3.973 5.237 9.706 10.884 mp_waitall_1 139946 16.5 6.838 10.143 6.838 10.143 qs_ot_get_derivative 99 11.5 0.001 0.001 9.132 9.229 init_scf_loop 11 6.9 0.001 0.004 8.680 8.682 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.113 7.263 prepare_preconditioner 11 7.9 0.000 0.000 6.869 6.885 make_preconditioner 11 8.9 0.000 0.001 6.869 6.885 sum_up_and_integrate 110 10.3 0.060 0.061 6.711 6.732 integrate_v_rspace 110 11.3 0.002 0.003 6.651 6.674 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.193 6.548 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.263 6.272 calculate_rho_elec 110 8.6 0.058 0.059 6.262 6.272 qs_ot_get_p 110 10.4 0.001 0.001 5.706 5.865 dbcsr_mm_accdrv_process 34862 16.1 4.622 5.248 5.588 5.781 init_scf_run 11 5.9 0.000 0.001 5.573 5.573 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.572 5.573 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.935 5.370 apply_single 110 13.6 0.000 0.000 4.934 5.370 make_m2s 4110 13.4 0.050 0.052 4.244 4.616 density_rs2pw 110 9.6 0.004 0.005 3.343 4.551 make_images 4110 14.4 0.390 0.507 4.129 4.501 multiply_cannon_sync_h2d 16440 15.4 3.711 4.458 3.711 4.458 ot_diis_step 99 11.5 0.010 0.011 4.395 4.395 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.060 4.064 rs_pw_transfer 902 11.9 0.010 0.011 2.866 4.046 pw_transfer 1331 11.6 0.066 0.073 3.887 3.899 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.779 3.793 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.127 3.745 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.715 3.716 grid_integrate_task_list 110 12.3 3.157 3.408 3.157 3.408 fft_wrap_pw1pw2_140 451 13.1 0.212 0.215 3.210 3.232 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.175 3.176 cp_fm_redistribute_end 48 14.0 1.980 3.145 1.982 3.145 cp_fm_diag_elpa_base 48 14.0 1.098 2.990 1.158 3.100 fft3d_ps 1111 14.6 1.089 1.100 3.079 3.095 wfi_extrapolate 11 7.9 0.001 0.001 3.032 3.032 make_images_data 4110 15.4 0.043 0.048 2.401 2.894 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.824 2.825 hybrid_alltoall_any 4261 16.3 0.105 0.373 2.134 2.743 cp_fm_cholesky_invert 11 10.9 2.714 2.720 2.714 2.720 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.490 2.540 calculate_dm_sparse 110 9.5 0.001 0.001 2.494 2.530 mp_waitany 17072 13.8 1.210 2.470 1.210 2.470 calculate_first_density_matrix 1 7.0 0.000 0.001 2.450 2.451 mp_alltoall_d11v 2046 13.8 1.846 2.438 1.846 2.438 potential_pw2rs 110 12.3 0.011 0.011 2.421 2.436 grid_collocate_task_list 110 9.6 2.087 2.417 2.087 2.417 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.344 2.412 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.860 2.311 mp_irecv_dv 48980 15.7 0.789 2.189 0.789 2.189 mp_sum_l 6514 12.8 1.522 2.120 1.522 2.120 rs_pw_transfer_RS2PW_140 121 11.5 0.177 0.184 0.921 2.116 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.993 1.994 dbcsr_complete_redistribute 325 12.2 0.322 0.351 1.480 1.936 cp_fm_upper_to_full 70 13.6 1.416 1.904 1.416 1.904 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.696 1.709 cp_fm_cholesky_decompose 22 10.9 1.631 1.650 1.631 1.650 mp_allgather_i34 2055 14.4 0.517 1.548 0.517 1.548 rs_gather_matrices 110 12.3 0.234 0.259 0.992 1.532 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 1.460 1.472 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.006 1.471 jit_kernel_multiply 8 16.6 0.580 1.467 0.580 1.467 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.374 1.467 acc_transpose_blocks 16440 15.4 0.070 0.073 1.260 1.277 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.267 1.277 mp_alltoall_z22v 1111 16.6 1.223 1.262 1.223 1.262 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.107000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.000000, yerr=7.687061 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 723.070976E+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 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 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 0.026 0.047 69.836 69.838 qs_mol_dyn_low 1 2.0 0.003 0.004 69.544 69.553 qs_forces 11 3.9 0.002 0.004 69.132 69.133 qs_energies 11 4.9 0.002 0.005 65.685 65.690 scf_env_do_scf 11 5.9 0.001 0.002 54.735 54.738 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.868 42.868 velocity_verlet 10 3.0 0.002 0.002 37.750 37.766 dbcsr_multiply_generic 2055 12.4 0.117 0.119 33.002 33.189 qs_scf_new_mos 99 7.5 0.001 0.001 27.554 27.664 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.553 27.664 multiply_cannon 2055 13.4 0.243 0.263 26.030 26.755 ot_scf_mini 99 9.5 0.003 0.004 25.920 26.026 multiply_cannon_loop 2055 14.4 0.887 0.904 24.689 25.209 ot_mini 99 10.5 0.001 0.001 15.204 15.331 multiply_cannon_multrec 24660 15.4 4.255 7.098 12.858 14.117 rebuild_ks_matrix 110 8.3 0.000 0.001 12.356 12.442 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.022 12.356 12.441 init_scf_loop 11 6.9 0.001 0.005 11.824 11.824 qs_ot_get_derivative 99 11.5 0.001 0.001 11.044 11.158 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.905 10.979 prepare_preconditioner 11 7.9 0.000 0.000 10.051 10.064 make_preconditioner 11 8.9 0.000 0.001 10.051 10.064 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.303 9.748 dbcsr_mm_accdrv_process 52304 16.0 6.955 8.196 8.457 9.423 init_scf_run 11 5.9 0.001 0.003 8.150 8.150 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.149 8.150 mp_waitall_1 121746 16.5 4.585 6.816 4.585 6.816 sum_up_and_integrate 110 10.3 0.066 0.069 6.651 6.663 integrate_v_rspace 110 11.3 0.003 0.003 6.584 6.596 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.285 6.296 calculate_rho_elec 110 8.6 0.077 0.081 6.285 6.295 qs_ot_get_p 110 10.4 0.001 0.001 5.762 5.941 make_m2s 4110 13.4 0.059 0.061 5.336 5.603 make_images 4110 14.4 0.571 0.695 5.195 5.457 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 5.248 5.249 calculate_first_density_matrix 1 7.0 0.001 0.004 5.045 5.048 acc_transpose_blocks 24660 15.4 0.105 0.108 4.636 4.981 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.854 4.884 cp_fm_upper_to_full 70 13.8 3.259 4.636 3.259 4.636 acc_transpose_blocks_kernels 24660 16.4 0.302 0.310 3.999 4.330 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.148 4.199 ot_diis_step 99 11.5 0.011 0.011 4.122 4.122 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.031 4.119 apply_single 110 13.6 0.000 0.000 4.030 4.119 jit_kernel_transpose 5 15.6 3.696 4.026 3.696 4.026 pw_transfer 1331 11.6 0.065 0.074 3.986 4.015 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.935 3.949 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.800 3.940 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.880 3.913 dbcsr_complete_redistribute 325 12.2 0.425 0.480 2.691 3.827 density_rs2pw 110 9.6 0.004 0.004 3.206 3.717 grid_integrate_task_list 110 12.3 3.261 3.474 3.261 3.474 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.450 3.450 multiply_cannon_sync_h2d 24660 15.4 3.207 3.381 3.207 3.381 fft_wrap_pw1pw2_140 451 13.1 0.202 0.211 3.320 3.354 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.140 3.264 fft3d_ps 1111 14.6 1.093 1.129 3.171 3.195 calculate_dm_sparse 110 9.5 0.001 0.001 3.097 3.128 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.045 3.108 rs_pw_transfer 902 11.9 0.010 0.011 2.532 3.069 wfi_extrapolate 11 7.9 0.001 0.001 3.010 3.010 make_images_data 4110 15.4 0.047 0.050 2.644 2.952 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.885 2.888 cp_fm_cholesky_invert 11 10.9 2.859 2.869 2.859 2.869 cp_fm_redistribute_end 48 14.0 1.442 2.852 1.443 2.853 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.738 2.841 hybrid_alltoall_any 4261 16.3 0.120 0.458 2.398 2.838 mp_alltoall_i22 605 13.7 1.647 2.827 1.647 2.827 cp_fm_diag_elpa_base 48 14.0 1.328 2.717 1.406 2.823 jit_kernel_multiply 12 15.8 1.175 2.449 1.175 2.449 grid_collocate_task_list 110 9.6 2.183 2.441 2.183 2.441 potential_pw2rs 110 12.3 0.012 0.013 2.304 2.313 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.297 2.298 mp_alltoall_d11v 2046 13.8 1.772 2.231 1.772 2.231 cp_fm_cholesky_decompose 22 10.9 1.710 1.753 1.710 1.753 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.720 1.752 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.706 1.724 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.699 1.710 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.598 1.695 mp_waitany 13376 13.8 1.111 1.608 1.111 1.608 mp_sum_l 6514 12.8 0.957 1.493 0.957 1.493 multiply_cannon_metrocomm4 20550 15.4 0.056 0.060 0.841 1.462 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=69.838000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=685.818182, yerr=7.929441 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_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 828.436480E+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 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 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 0.014 0.033 60.091 60.092 qs_mol_dyn_low 1 2.0 0.003 0.004 59.835 59.844 qs_forces 11 3.9 0.002 0.002 59.641 59.641 qs_energies 11 4.9 0.001 0.001 55.938 55.943 scf_env_do_scf 11 5.9 0.001 0.001 44.926 44.927 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 37.033 37.034 velocity_verlet 10 3.0 0.002 0.002 31.665 31.676 dbcsr_multiply_generic 2055 12.4 0.106 0.107 26.626 26.777 multiply_cannon 2055 13.4 0.247 0.254 21.124 22.451 qs_scf_new_mos 99 7.5 0.001 0.001 21.678 21.736 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.677 21.736 ot_scf_mini 99 9.5 0.002 0.002 20.437 20.486 multiply_cannon_loop 2055 14.4 0.322 0.336 19.785 20.038 rebuild_ks_matrix 110 8.3 0.000 0.000 12.099 12.148 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.099 12.147 ot_mini 99 10.5 0.001 0.001 11.455 11.505 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.764 10.804 multiply_cannon_multrec 8220 15.4 3.485 5.067 7.899 9.231 mp_waitall_1 103326 16.6 6.198 7.912 6.198 7.912 init_scf_run 11 5.9 0.000 0.001 7.856 7.856 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.856 7.856 init_scf_loop 11 6.9 0.000 0.000 7.844 7.846 qs_ot_get_derivative 99 11.5 0.001 0.001 7.624 7.676 sum_up_and_integrate 110 10.3 0.079 0.081 6.622 6.634 integrate_v_rspace 110 11.3 0.003 0.003 6.542 6.554 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.436 6.447 calculate_rho_elec 110 8.6 0.115 0.116 6.435 6.446 prepare_preconditioner 11 7.9 0.000 0.000 6.165 6.176 make_preconditioner 11 8.9 0.000 0.000 6.165 6.176 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.741 5.814 dbcsr_mm_accdrv_process 17442 15.9 2.944 4.409 4.284 5.519 qs_ot_get_p 110 10.4 0.001 0.001 5.036 5.084 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 5.027 5.028 calculate_first_density_matrix 1 7.0 0.000 0.000 5.001 5.002 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.134 4.779 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.676 4.694 make_m2s 4110 13.4 0.038 0.040 4.199 4.460 make_images 4110 14.4 0.636 0.695 4.069 4.333 acc_transpose_blocks 8220 15.4 0.036 0.037 4.131 4.247 pw_transfer 1331 11.6 0.066 0.071 4.203 4.212 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.096 4.107 acc_transpose_blocks_kernels 8220 16.4 0.109 0.112 3.901 4.013 jit_kernel_transpose 5 15.6 3.792 3.901 3.792 3.901 ot_diis_step 99 11.5 0.012 0.012 3.810 3.810 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.715 3.764 apply_single 110 13.6 0.000 0.000 3.715 3.764 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.660 3.663 fft_wrap_pw1pw2_140 451 13.1 0.216 0.223 3.536 3.559 density_rs2pw 110 9.6 0.004 0.004 3.180 3.555 grid_integrate_task_list 110 12.3 3.350 3.524 3.350 3.524 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.348 3.348 fft3d_ps 1111 14.6 1.149 1.176 3.338 3.348 multiply_cannon_sync_h2d 8220 15.4 2.897 3.045 2.897 3.045 cp_fm_cholesky_invert 11 10.9 2.995 3.000 2.995 3.000 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.812 2.813 make_images_data 4110 15.4 0.038 0.043 2.393 2.807 cp_fm_redistribute_end 48 14.0 0.717 2.780 0.722 2.781 calculate_dm_sparse 110 9.5 0.001 0.001 2.713 2.759 wfi_extrapolate 11 7.9 0.001 0.001 2.759 2.759 cp_fm_diag_elpa_base 48 14.0 1.871 2.579 2.050 2.742 hybrid_alltoall_any 4261 16.3 0.200 0.866 2.333 2.718 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 2.695 2.695 rs_pw_transfer 902 11.9 0.010 0.010 2.220 2.625 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.530 2.562 grid_collocate_task_list 110 9.6 2.277 2.537 2.277 2.537 potential_pw2rs 110 12.3 0.015 0.016 2.164 2.168 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.773 1.998 mp_alltoall_d11v 2046 13.8 1.676 1.918 1.676 1.918 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.834 1.856 cp_fm_cholesky_decompose 22 10.9 1.694 1.705 1.694 1.705 jit_kernel_multiply 7 15.8 1.033 1.693 1.033 1.693 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.674 1.686 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.671 1.679 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.531 1.656 mp_allgather_i34 2055 14.4 0.524 1.617 0.524 1.617 dbcsr_complete_redistribute 325 12.2 0.561 0.580 1.470 1.551 mp_waitany 9240 13.8 1.061 1.497 1.061 1.497 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.482 1.382 qs_create_task_list 11 7.9 0.001 0.001 1.238 1.344 generate_qs_task_list 11 8.9 0.388 0.452 1.237 1.344 mp_irecv_dv 24056 15.7 0.457 1.343 0.457 1.343 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.820 1.238 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.165 0.796 1.208 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.180 1.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=60.092000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=779.636364, yerr=14.436524 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_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 1.382912E+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 67104. MP_Allreduce 9672 819. 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 0.019 0.042 88.433 88.434 qs_mol_dyn_low 1 2.0 0.003 0.004 88.124 88.133 qs_forces 11 3.9 0.002 0.003 88.059 88.060 qs_energies 11 4.9 0.001 0.001 83.845 83.847 scf_env_do_scf 11 5.9 0.001 0.001 73.361 73.362 velocity_verlet 10 3.0 0.002 0.002 56.133 56.139 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 44.997 44.998 dbcsr_multiply_generic 2055 12.4 0.120 0.123 29.085 29.202 init_scf_loop 11 6.9 0.000 0.000 28.292 28.294 qs_scf_new_mos 99 7.5 0.001 0.001 26.370 26.460 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.369 26.460 prepare_preconditioner 11 7.9 0.000 0.000 26.245 26.256 make_preconditioner 11 8.9 0.000 0.000 26.245 26.256 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.573 25.713 ot_scf_mini 99 9.5 0.002 0.002 24.613 24.680 multiply_cannon 2055 13.4 0.343 0.361 22.054 22.767 multiply_cannon_loop 2055 14.4 0.341 0.345 20.245 20.559 cp_fm_upper_to_full 70 14.2 12.579 17.952 12.579 17.952 rebuild_ks_matrix 110 8.3 0.001 0.001 14.363 14.445 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.363 14.445 ot_mini 99 10.5 0.001 0.001 13.727 13.795 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.017 13.091 dbcsr_complete_redistribute 325 12.2 1.026 1.053 7.241 10.343 multiply_cannon_multrec 8220 15.4 4.422 4.578 9.771 9.866 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.216 9.323 qs_ot_get_derivative 99 11.5 0.001 0.001 9.097 9.165 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.656 8.743 mp_waitall_1 84994 16.7 7.567 8.640 7.567 8.640 mp_alltoall_i22 605 13.7 5.280 8.360 5.280 8.360 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.881 7.920 calculate_rho_elec 110 8.6 0.227 0.227 7.880 7.919 sum_up_and_integrate 110 10.3 0.150 0.153 7.503 7.520 integrate_v_rspace 110 11.3 0.004 0.004 7.353 7.370 init_scf_run 11 5.9 0.000 0.001 6.176 6.177 scf_env_initial_rho_setup 11 6.9 0.005 0.006 6.176 6.176 qs_ot_get_p 110 10.4 0.001 0.001 5.723 5.808 make_m2s 4110 13.4 0.043 0.044 5.237 5.759 make_images 4110 14.4 0.878 0.931 5.049 5.571 cp_fm_cholesky_invert 11 10.9 5.465 5.469 5.465 5.469 dbcsr_mm_accdrv_process 11614 15.7 3.220 3.683 5.207 5.438 pw_transfer 1331 11.6 0.075 0.076 5.372 5.379 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.255 5.263 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.713 5.224 apply_single 110 13.6 0.000 0.000 4.713 5.223 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.798 5.120 ot_diis_step 99 11.5 0.015 0.016 4.610 4.611 fft_wrap_pw1pw2_140 451 13.1 0.221 0.222 4.572 4.579 fft3d_ps 1111 14.6 1.295 1.306 4.451 4.461 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.147 4.154 multiply_cannon_sync_h2d 8220 15.4 3.951 3.959 3.951 3.959 density_rs2pw 110 9.6 0.004 0.004 3.878 3.904 grid_integrate_task_list 110 12.3 3.694 3.777 3.694 3.777 qs_energies_init_hamiltonians 11 5.9 0.010 0.011 3.712 3.713 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.704 3.704 hybrid_alltoall_any 4261 16.3 0.256 0.552 2.857 3.596 make_images_data 4110 15.4 0.042 0.045 2.866 3.503 calculate_dm_sparse 110 9.5 0.001 0.001 3.441 3.464 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.926 3.388 wfi_extrapolate 11 7.9 0.001 0.001 3.330 3.331 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.095 3.095 cp_fm_diag_elpa_base 48 14.0 2.569 2.762 3.093 3.093 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.007 3.011 calculate_first_density_matrix 1 7.0 0.000 0.000 2.719 2.721 grid_collocate_task_list 110 9.6 2.659 2.678 2.659 2.678 potential_pw2rs 110 12.3 0.021 0.021 2.562 2.568 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.503 2.532 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.415 2.456 rs_pw_transfer 902 11.9 0.011 0.011 2.332 2.361 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.235 2.298 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.092 2.176 mp_alltoall_d11v 2046 13.8 1.970 2.095 1.970 2.095 cp_fm_cholesky_decompose 22 10.9 2.005 2.013 2.005 2.013 jit_kernel_multiply 10 15.4 1.787 1.998 1.787 1.998 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.980 1.992 qs_create_task_list 11 7.9 0.001 0.001 1.922 1.971 generate_qs_task_list 11 8.9 0.760 0.813 1.921 1.970 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.952 1.960 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.763 1.802 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.434000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1229.090909, yerr=70.073279 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 633.253888E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 2520958. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.036 207.382 207.384 qs_mol_dyn_low 1 2.0 0.003 0.004 207.003 207.018 qs_forces 11 3.9 0.003 0.003 206.917 206.919 qs_energies 11 4.9 0.002 0.002 201.260 201.275 scf_env_do_scf 11 5.9 0.001 0.002 182.325 182.329 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 161.405 161.407 dbcsr_multiply_generic 2507 12.6 0.178 0.182 126.186 126.850 velocity_verlet 10 3.0 0.001 0.002 123.731 123.732 qs_scf_new_mos 117 7.6 0.001 0.001 121.007 121.298 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.006 121.297 ot_scf_mini 117 9.6 0.003 0.003 114.462 114.727 multiply_cannon 2507 13.6 0.236 0.244 103.087 104.729 multiply_cannon_loop 2507 14.6 2.119 2.188 100.900 102.598 ot_mini 117 10.6 0.001 0.001 65.552 65.801 multiply_cannon_multrec 60168 15.6 33.092 34.712 41.450 43.192 qs_ot_get_derivative 117 11.6 0.001 0.001 40.835 41.084 rebuild_ks_matrix 128 8.3 0.001 0.001 33.943 34.304 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.016 33.942 34.304 mp_waitall_1 267128 16.5 28.186 31.285 28.186 31.285 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.488 30.794 multiply_cannon_sync_h2d 60168 15.6 27.254 28.953 27.254 28.953 qs_ot_get_p 128 10.4 0.001 0.001 27.085 27.395 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.212 24.947 apply_single 128 13.6 0.001 0.001 24.212 24.947 ot_diis_step 117 11.6 0.007 0.008 24.373 24.374 init_scf_loop 11 6.9 0.000 0.001 20.838 20.839 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.396 20.485 multiply_cannon_metrocomm3 60168 15.6 0.116 0.122 15.607 19.300 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.919 19.120 cp_dbcsr_syevd 83 12.4 0.005 0.006 17.823 17.823 prepare_preconditioner 11 7.9 0.000 0.000 16.190 16.236 make_preconditioner 11 8.9 0.000 0.000 16.190 16.236 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.422 15.603 cp_fm_diag_elpa 83 13.4 0.000 0.000 14.950 14.956 cp_fm_redistribute_end 83 14.4 11.812 14.871 11.827 14.876 init_scf_run 11 5.9 0.000 0.001 14.790 14.791 scf_env_initial_rho_setup 11 6.9 0.001 0.002 14.790 14.790 cp_fm_diag_elpa_base 83 14.4 3.003 14.502 3.035 14.607 sum_up_and_integrate 128 10.3 0.089 0.105 14.492 14.509 integrate_v_rspace 128 11.3 0.003 0.004 14.403 14.422 make_m2s 5014 13.6 0.105 0.113 13.706 14.107 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.900 13.995 calculate_rho_elec 128 8.7 0.045 0.064 13.899 13.994 make_images 5014 14.6 0.402 0.421 13.524 13.930 density_rs2pw 128 9.7 0.006 0.007 7.314 10.623 cp_fm_cholesky_invert 11 10.9 9.346 9.354 9.346 9.354 rs_pw_transfer 1046 11.9 0.016 0.019 5.894 9.224 wfi_extrapolate 11 7.9 0.001 0.001 9.146 9.146 mp_sum_l 7870 13.0 8.126 9.137 8.126 9.137 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 9.070 9.077 calculate_dm_sparse 128 9.5 0.001 0.001 8.357 8.471 dbcsr_mm_accdrv_process 124484 16.2 3.318 3.460 7.922 8.420 pw_transfer 1547 11.6 0.077 0.099 8.112 8.361 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 8.247 8.353 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.040 8.181 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.906 8.159 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.814 7.908 multiply_cannon_metrocomm1 60168 15.6 0.089 0.093 6.096 7.808 grid_integrate_task_list 128 12.3 7.077 7.577 7.077 7.577 make_images_data 5014 15.6 0.067 0.074 6.674 7.518 hybrid_alltoall_any 5200 16.5 0.291 2.253 5.855 7.145 fft_wrap_pw1pw2_140 523 13.2 0.444 0.510 6.791 7.103 fft3d_ps 1291 14.7 2.106 2.889 6.693 6.857 mp_waitany 16020 13.9 2.658 5.970 2.658 5.970 grid_collocate_task_list 128 9.7 4.562 5.762 4.562 5.762 acc_transpose_blocks 60168 15.6 0.309 0.324 5.427 5.533 mp_alltoall_d11v 2415 14.1 4.159 5.523 4.159 5.523 calculate_first_density_matrix 1 7.0 0.000 0.000 5.477 5.485 rs_pw_transfer_RS2PW_140 139 11.5 0.281 0.302 2.101 5.452 potential_pw2rs 128 12.3 0.009 0.011 4.853 4.880 cp_fm_cholesky_decompose 22 10.9 4.676 4.691 4.676 4.691 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.384000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.545455, yerr=7.643936 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_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 828.719104E+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 4127 56635. MP_Allreduce 11357 942. MP_Sync 172 MP_Alltoall 1983 5621081. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_comm_split 84 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 0.020 0.045 195.836 195.838 qs_mol_dyn_low 1 2.0 0.003 0.004 195.460 195.473 qs_forces 11 3.9 0.003 0.007 195.378 195.379 qs_energies 11 4.9 0.002 0.006 188.546 188.558 scf_env_do_scf 11 5.9 0.001 0.002 169.853 169.863 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 136.579 136.581 velocity_verlet 10 3.0 0.001 0.002 122.192 122.194 dbcsr_multiply_generic 2529 12.6 0.187 0.193 100.079 101.316 qs_scf_new_mos 118 7.6 0.001 0.001 95.739 96.183 qs_scf_loop_do_ot 118 8.6 0.001 0.002 95.738 96.182 ot_scf_mini 118 9.6 0.004 0.005 90.886 91.388 multiply_cannon 2529 13.6 0.491 0.552 80.073 83.980 multiply_cannon_loop 2529 14.6 1.264 1.304 76.809 79.284 ot_mini 118 10.6 0.001 0.001 50.205 50.646 mp_waitall_1 216598 16.6 24.561 38.720 24.561 38.720 multiply_cannon_multrec 30348 15.6 22.156 25.829 31.785 35.821 rebuild_ks_matrix 129 8.3 0.001 0.001 33.858 34.384 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.026 33.857 34.383 init_scf_loop 11 6.9 0.001 0.006 33.181 33.182 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.496 30.975 prepare_preconditioner 11 7.9 0.000 0.000 28.717 28.778 make_preconditioner 11 8.9 0.000 0.002 28.717 28.778 qs_ot_get_derivative 118 11.6 0.001 0.002 28.026 28.531 multiply_cannon_metrocomm3 30348 15.6 0.095 0.100 15.828 28.503 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.422 27.962 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.197 23.313 apply_single 129 13.6 0.001 0.001 22.197 23.313 qs_ot_get_p 129 10.4 0.001 0.001 22.412 23.021 ot_diis_step 118 11.6 0.014 0.015 21.998 22.000 multiply_cannon_sync_h2d 30348 15.6 19.438 21.518 19.438 21.518 qs_ot_p2m_diag 84 11.4 0.190 0.219 17.524 17.562 cp_fm_cholesky_invert 11 10.9 16.845 16.858 16.845 16.858 cp_dbcsr_syevd 84 12.4 0.006 0.007 16.357 16.358 make_m2s 5058 13.6 0.091 0.097 14.198 15.681 make_images 5058 14.6 1.157 1.352 13.985 15.469 sum_up_and_integrate 129 10.3 0.115 0.135 15.071 15.103 integrate_v_rspace 129 11.3 0.003 0.004 14.956 14.992 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.411 14.440 calculate_rho_elec 129 8.7 0.089 0.106 14.410 14.439 init_scf_run 11 5.9 0.000 0.001 13.790 13.792 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.790 13.792 cp_fm_diag_elpa 84 13.4 0.000 0.001 13.176 13.183 cp_fm_redistribute_end 84 14.4 7.689 13.099 7.704 13.103 cp_fm_diag_elpa_base 84 14.4 5.158 12.581 5.380 12.948 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 11.088 11.468 multiply_cannon_metrocomm4 27819 15.6 0.095 0.109 3.828 10.592 make_images_data 5058 15.6 0.065 0.072 8.538 10.300 mp_irecv_dv 70084 16.3 3.633 10.209 3.633 10.209 density_rs2pw 129 9.7 0.006 0.006 7.739 10.153 dbcsr_mm_accdrv_process 62758 16.2 4.503 5.271 9.088 9.696 hybrid_alltoall_any 5245 16.5 0.345 1.517 7.329 9.646 pw_transfer 1559 11.6 0.089 0.113 9.192 9.260 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 8.963 9.034 rs_pw_transfer 1054 12.0 0.014 0.016 6.028 8.485 wfi_extrapolate 11 7.9 0.001 0.001 8.467 8.467 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 8.356 8.367 fft_wrap_pw1pw2_140 527 13.2 0.476 0.532 7.857 7.952 grid_integrate_task_list 129 12.3 7.194 7.673 7.194 7.673 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 7.478 7.664 fft3d_ps 1301 14.7 2.768 2.952 7.318 7.357 cp_fm_cholesky_decompose 22 10.9 6.982 7.064 6.982 7.064 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.202 6.957 calculate_dm_sparse 129 9.5 0.001 0.001 6.502 6.635 mp_sum_l 7936 13.1 4.023 6.083 4.023 6.083 grid_collocate_task_list 129 9.7 4.737 5.909 4.737 5.909 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.385 5.460 mp_alltoall_d11v 2429 14.1 4.161 5.356 4.161 5.356 potential_pw2rs 129 12.3 0.015 0.018 5.284 5.311 calculate_first_density_matrix 1 7.0 0.001 0.004 5.143 5.146 mp_waitany 11836 13.9 2.579 4.994 2.579 4.994 mp_allgather_i34 2529 14.6 1.817 4.763 1.817 4.763 rs_pw_transfer_RS2PW_140 140 11.5 0.355 0.386 2.160 4.596 mp_sum_d 4499 12.2 2.718 4.063 2.718 4.063 acc_transpose_blocks 30348 15.6 0.138 0.142 4.008 4.051 dbcsr_complete_redistribute 397 12.7 0.772 0.862 3.152 4.002 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=195.838000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.727273, yerr=4.534532 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 935.096320E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+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 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.079 178.440 178.445 qs_mol_dyn_low 1 2.0 0.003 0.004 177.921 177.934 qs_forces 11 3.9 0.003 0.005 177.801 177.810 qs_energies 11 4.9 0.002 0.009 171.130 171.140 scf_env_do_scf 11 5.9 0.001 0.002 155.290 155.291 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 119.803 119.804 velocity_verlet 10 3.0 0.001 0.002 114.315 114.318 dbcsr_multiply_generic 2507 12.6 0.181 0.185 81.879 82.982 qs_scf_new_mos 117 7.6 0.001 0.001 81.012 81.379 qs_scf_loop_do_ot 117 8.6 0.001 0.002 81.011 81.379 ot_scf_mini 117 9.6 0.004 0.004 76.887 77.264 multiply_cannon 2507 13.6 0.510 0.537 61.807 66.591 multiply_cannon_loop 2507 14.6 0.856 0.884 58.910 61.652 ot_mini 117 10.6 0.001 0.001 42.503 42.889 init_scf_loop 11 6.9 0.001 0.009 35.385 35.386 mp_waitall_1 170520 16.6 25.518 34.618 25.518 34.618 rebuild_ks_matrix 128 8.3 0.001 0.001 31.162 31.746 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.020 31.161 31.745 prepare_preconditioner 11 7.9 0.000 0.000 31.274 31.329 make_preconditioner 11 8.9 0.000 0.001 31.274 31.329 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.881 30.298 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.091 28.622 multiply_cannon_multrec 20056 15.6 13.297 16.160 21.995 24.765 multiply_cannon_metrocomm3 20056 15.6 0.061 0.065 15.602 24.610 qs_ot_get_derivative 117 11.6 0.001 0.002 22.712 23.094 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.916 20.858 apply_single 128 13.6 0.001 0.001 19.916 20.858 qs_ot_get_p 128 10.4 0.001 0.001 19.727 20.208 ot_diis_step 117 11.6 0.018 0.021 19.689 19.689 multiply_cannon_sync_h2d 20056 15.6 14.251 16.288 14.251 16.288 make_m2s 5014 13.6 0.081 0.087 15.006 15.773 make_images 5014 14.6 1.169 1.262 14.771 15.530 qs_ot_p2m_diag 83 11.4 0.265 0.272 15.506 15.515 sum_up_and_integrate 128 10.3 0.132 0.143 14.875 14.906 integrate_v_rspace 128 11.3 0.003 0.004 14.742 14.772 cp_fm_cholesky_invert 11 10.9 14.695 14.704 14.695 14.704 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.560 14.598 calculate_rho_elec 128 8.7 0.132 0.147 14.559 14.598 cp_dbcsr_syevd 83 12.4 0.006 0.007 14.489 14.490 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.343 11.343 cp_fm_redistribute_end 83 14.4 4.288 11.278 4.303 11.281 cp_fm_diag_elpa_base 83 14.4 6.549 10.677 6.956 11.159 init_scf_run 11 5.9 0.000 0.001 10.812 10.812 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.812 10.812 make_images_data 5014 15.6 0.061 0.068 9.325 10.465 hybrid_alltoall_any 5200 16.5 0.435 2.006 8.023 9.919 multiply_cannon_metrocomm4 17549 15.6 0.061 0.071 3.522 9.593 density_rs2pw 128 9.7 0.006 0.006 7.423 9.567 pw_transfer 1547 11.6 0.087 0.109 9.267 9.387 mp_irecv_dv 50230 16.2 3.398 9.337 3.398 9.337 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.933 9.221 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.042 9.166 dbcsr_mm_accdrv_process 41502 16.2 4.363 5.224 8.161 8.316 fft_wrap_pw1pw2_140 523 13.2 0.477 0.513 7.985 8.114 grid_integrate_task_list 128 12.3 7.289 7.858 7.289 7.858 cp_fm_upper_to_full 105 14.5 5.775 7.616 5.775 7.616 wfi_extrapolate 11 7.9 0.001 0.001 7.457 7.457 rs_pw_transfer 1046 11.9 0.014 0.015 5.280 7.431 cp_fm_cholesky_decompose 22 10.9 7.361 7.397 7.361 7.397 fft3d_ps 1291 14.7 2.669 2.894 7.303 7.395 dbcsr_complete_redistribute 395 12.7 1.175 1.222 4.675 6.467 calculate_dm_sparse 128 9.5 0.001 0.001 5.776 5.859 grid_collocate_task_list 128 9.7 4.890 5.802 4.890 5.802 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.764 5.769 mp_alltoall_d11v 2415 14.1 4.379 5.319 4.379 5.319 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.490 5.280 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.501 5.203 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.816 4.963 potential_pw2rs 128 12.3 0.020 0.022 4.940 4.957 mp_sum_l 7870 13.0 3.268 4.765 3.268 4.765 mp_waitany 11748 13.9 2.392 4.657 2.392 4.657 mp_allgather_i34 2507 14.6 1.383 4.616 1.383 4.616 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.373 4.148 rs_pw_transfer_RS2PW_140 139 11.5 0.332 0.350 1.949 4.091 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.046 4.081 mp_alltoall_i22 716 14.1 2.003 3.930 2.003 3.930 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.777 3.778 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.588 3.636 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.445000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=886.454545, yerr=10.849077 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 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 1.151660E+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 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4083 57219. MP_Allreduce 11246 1068. MP_Sync 168 MP_Alltoall 1700 12496371. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_comm_split 82 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 0.051 0.165 191.609 191.612 qs_mol_dyn_low 1 2.0 0.003 0.004 191.030 191.044 qs_forces 11 3.9 0.031 0.067 190.884 190.891 qs_energies 11 4.9 0.005 0.010 183.626 183.656 scf_env_do_scf 11 5.9 0.001 0.004 164.389 164.396 velocity_verlet 10 3.0 0.002 0.002 125.094 125.098 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 117.582 117.584 dbcsr_multiply_generic 2485 12.5 0.189 0.196 81.028 81.648 qs_scf_new_mos 116 7.6 0.001 0.001 79.576 79.845 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.575 79.845 ot_scf_mini 116 9.6 0.004 0.004 75.143 75.482 multiply_cannon 2485 13.5 0.552 0.597 56.664 59.251 multiply_cannon_loop 2485 14.5 1.174 1.200 52.912 54.391 init_scf_loop 11 6.9 0.001 0.003 46.678 46.679 prepare_preconditioner 11 7.9 0.000 0.000 42.506 42.533 make_preconditioner 11 8.9 0.000 0.001 42.506 42.533 ot_mini 116 10.6 0.001 0.001 42.011 42.353 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.046 41.037 multiply_cannon_multrec 29820 15.5 14.070 18.940 25.855 30.803 rebuild_ks_matrix 127 8.3 0.001 0.001 30.120 30.373 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.021 30.119 30.372 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.112 27.329 mp_waitall_1 146592 16.7 17.204 26.814 17.204 26.814 qs_ot_get_derivative 116 11.6 0.001 0.002 22.663 23.012 make_m2s 4970 13.5 0.095 0.098 20.013 20.904 make_images 4970 14.5 1.916 2.239 19.707 20.597 qs_ot_get_p 127 10.4 0.001 0.001 19.539 19.914 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.767 19.234 apply_single 127 13.6 0.001 0.001 18.767 19.234 ot_diis_step 116 11.6 0.017 0.018 19.226 19.228 cp_fm_cholesky_invert 11 10.9 16.314 16.323 16.314 16.323 cp_fm_upper_to_full 104 14.7 10.992 16.270 10.992 16.270 qs_ot_p2m_diag 82 11.4 0.339 0.385 15.311 15.362 sum_up_and_integrate 127 10.3 0.140 0.151 14.957 14.982 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.840 14.873 calculate_rho_elec 127 8.7 0.174 0.189 14.839 14.872 integrate_v_rspace 127 11.3 0.004 0.004 14.817 14.846 multiply_cannon_metrocomm3 29820 15.5 0.047 0.049 6.354 14.796 cp_dbcsr_syevd 82 12.4 0.009 0.010 14.035 14.036 init_scf_run 11 5.9 0.000 0.001 13.197 13.198 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.196 13.198 dbcsr_complete_redistribute 393 12.7 1.500 1.629 9.070 12.813 make_images_data 4970 15.5 0.064 0.067 10.723 12.567 multiply_cannon_sync_h2d 29820 15.5 11.623 12.506 11.623 12.506 dbcsr_mm_accdrv_process 61748 16.2 7.279 8.263 11.364 11.861 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.686 11.416 hybrid_alltoall_any 5155 16.4 0.519 2.187 9.588 11.305 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.926 10.931 cp_fm_redistribute_end 82 14.4 1.897 10.843 1.915 10.849 cp_fm_diag_elpa_base 82 14.4 8.321 10.217 8.896 10.694 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.436 10.053 pw_transfer 1535 11.6 0.088 0.106 9.665 9.734 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.369 9.621 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.437 9.513 mp_alltoall_i22 712 14.1 5.549 9.204 5.549 9.204 density_rs2pw 127 9.7 0.006 0.006 7.322 8.678 fft_wrap_pw1pw2_140 519 13.2 0.480 0.489 8.348 8.440 grid_integrate_task_list 127 12.3 7.452 7.895 7.452 7.895 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.783 7.828 wfi_extrapolate 11 7.9 0.001 0.001 7.772 7.772 cp_fm_cholesky_decompose 22 10.9 7.661 7.757 7.661 7.757 fft3d_ps 1281 14.7 2.754 2.833 7.677 7.727 multiply_cannon_metrocomm4 24850 15.5 0.074 0.085 2.748 7.138 mp_irecv_dv 75445 16.2 2.603 6.874 2.603 6.874 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.718 6.787 rs_pw_transfer 1038 11.9 0.013 0.014 4.903 6.298 calculate_dm_sparse 127 9.5 0.001 0.001 6.093 6.194 grid_collocate_task_list 127 9.7 5.041 5.834 5.041 5.834 mp_alltoall_d11v 2401 14.1 5.020 5.629 5.020 5.629 calculate_first_density_matrix 1 7.0 0.000 0.001 5.214 5.219 potential_pw2rs 127 12.3 0.022 0.023 4.796 4.811 qs_energies_init_hamiltonians 11 5.9 0.007 0.012 4.553 4.579 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.405 4.506 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.141 4.199 acc_transpose_blocks 29820 15.5 0.137 0.140 4.120 4.165 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=191.612000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1076.636364, yerr=19.170397 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.522610E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58209. MP_Allreduce 11002 1175. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.032 178.527 178.527 qs_mol_dyn_low 1 2.0 0.003 0.004 178.152 178.164 qs_forces 11 3.9 0.003 0.003 178.043 178.050 qs_energies 11 4.9 0.002 0.002 170.510 170.518 scf_env_do_scf 11 5.9 0.001 0.001 150.948 150.955 velocity_verlet 10 3.0 0.002 0.002 115.325 115.329 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 114.753 114.755 dbcsr_multiply_generic 2507 12.6 0.183 0.188 76.292 76.664 qs_scf_new_mos 117 7.6 0.001 0.001 75.596 75.694 qs_scf_loop_do_ot 117 8.6 0.001 0.001 75.596 75.694 ot_scf_mini 117 9.6 0.004 0.004 71.210 71.250 multiply_cannon 2507 13.6 0.587 0.618 56.549 61.009 multiply_cannon_loop 2507 14.6 0.444 0.454 51.922 52.465 ot_mini 117 10.6 0.001 0.001 39.615 39.648 init_scf_loop 11 6.9 0.000 0.000 36.046 36.048 mp_waitall_1 125778 16.7 25.863 32.832 25.863 32.832 prepare_preconditioner 11 7.9 0.000 0.000 32.075 32.110 make_preconditioner 11 8.9 0.000 0.000 32.075 32.110 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.967 30.234 rebuild_ks_matrix 128 8.3 0.001 0.001 30.167 30.230 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 30.166 30.229 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.417 27.474 multiply_cannon_multrec 10028 15.6 10.392 14.541 17.650 20.575 ot_diis_step 117 11.6 0.020 0.020 19.799 19.799 qs_ot_get_derivative 117 11.6 0.002 0.002 19.748 19.784 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.368 19.662 apply_single 128 13.6 0.001 0.001 19.368 19.661 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 12.162 19.106 qs_ot_get_p 128 10.4 0.001 0.001 18.764 18.820 make_m2s 5014 13.6 0.066 0.071 16.178 18.413 cp_fm_cholesky_invert 11 10.9 18.392 18.398 18.392 18.398 make_images 5014 14.6 2.260 2.793 15.870 18.103 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.705 15.737 calculate_rho_elec 128 8.7 0.259 0.269 15.704 15.737 sum_up_and_integrate 128 10.3 0.181 0.191 15.226 15.269 integrate_v_rspace 128 11.3 0.004 0.004 15.044 15.094 qs_ot_p2m_diag 83 11.4 0.495 0.501 14.996 15.014 cp_dbcsr_syevd 83 12.4 0.006 0.006 13.857 13.859 init_scf_run 11 5.9 0.000 0.001 12.743 12.744 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.743 12.743 make_images_data 5014 15.6 0.052 0.060 9.883 12.318 multiply_cannon_sync_h2d 10028 15.6 11.627 12.274 11.627 12.274 hybrid_alltoall_any 5200 16.5 0.801 3.451 9.632 12.149 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.778 10.790 cp_fm_diag_elpa_base 83 14.4 10.530 10.610 10.767 10.778 pw_transfer 1547 11.6 0.086 0.096 10.073 10.112 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.849 9.892 fft_wrap_pw1pw2_140 523 13.2 0.496 0.516 8.648 8.691 grid_integrate_task_list 128 12.3 7.755 8.241 7.755 8.241 cp_fm_cholesky_decompose 22 10.9 8.053 8.179 8.053 8.179 density_rs2pw 128 9.7 0.005 0.006 7.202 8.081 fft3d_ps 1291 14.7 2.707 2.785 8.010 8.037 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.945 7.973 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.525 7.536 dbcsr_mm_accdrv_process 20762 16.1 2.675 3.438 6.886 7.502 multiply_cannon_metrocomm1 10028 15.6 0.029 0.029 4.508 7.473 wfi_extrapolate 11 7.9 0.001 0.001 7.445 7.445 mp_allgather_i34 2507 14.6 2.838 7.274 2.838 7.274 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.426 6.448 calculate_dm_sparse 128 9.5 0.001 0.001 6.033 6.099 grid_collocate_task_list 128 9.7 5.392 6.053 5.392 6.053 mp_alltoall_d11v 2415 14.1 5.143 5.804 5.143 5.804 dbcsr_complete_redistribute 395 12.7 2.117 2.247 5.198 5.599 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 5.329 5.330 rs_pw_transfer 1046 11.9 0.012 0.013 4.207 5.115 calculate_first_density_matrix 1 7.0 0.000 0.000 5.081 5.084 potential_pw2rs 128 12.3 0.026 0.027 4.658 4.680 multiply_cannon_metrocomm4 7521 15.6 0.024 0.026 1.846 4.064 mp_irecv_dv 28860 15.9 1.809 3.989 1.809 3.989 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.551 3.834 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.676 3.700 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.382 3.690 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.655 3.663 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.547 3.625 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=178.527000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1428.636364, yerr=53.564358 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 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 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 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 94.243766E+12 0.0% 0.0% 100.0% flops max/rank 11.787405E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547424 0.0% 0.0% 100.0% number of processed stacks 1982128 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.0 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 3.069293E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101360 MPI messages size (bytes): total size 1.145137E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.297725E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45848 35601252352 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58564. MP_Allreduce 11104 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.046 300.807 300.809 qs_mol_dyn_low 1 2.0 0.003 0.004 300.222 300.262 qs_forces 11 3.9 0.003 0.003 300.124 300.127 qs_energies 11 4.9 0.002 0.002 290.747 290.763 scf_env_do_scf 11 5.9 0.001 0.002 266.542 266.557 velocity_verlet 10 3.0 0.002 0.002 215.746 215.754 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 142.529 142.530 init_scf_loop 11 6.9 0.000 0.000 123.757 123.759 prepare_preconditioner 11 7.9 0.000 0.000 118.614 118.644 make_preconditioner 11 8.9 0.000 0.000 118.614 118.644 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.250 115.772 qs_scf_new_mos 118 7.6 0.001 0.001 90.912 91.041 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.911 91.040 dbcsr_multiply_generic 2534 12.6 0.215 0.223 85.605 86.229 ot_scf_mini 118 9.6 0.004 0.004 86.099 86.152 cp_fm_upper_to_full 106 14.8 52.442 74.746 52.442 74.746 multiply_cannon 2534 13.6 0.709 0.768 61.150 61.734 multiply_cannon_loop 2534 14.6 0.477 0.485 57.569 59.001 ot_mini 118 10.6 0.001 0.001 44.660 44.711 dbcsr_complete_redistribute 397 12.7 3.964 3.998 28.983 41.476 rebuild_ks_matrix 129 8.3 0.001 0.001 38.901 38.950 qs_ks_build_kohn_sham_matrix 129 9.3 0.019 0.020 38.900 38.949 copy_fm_to_dbcsr 210 11.7 0.002 0.002 25.602 38.011 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.880 35.935 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.318 35.590 mp_alltoall_i22 720 14.1 21.136 33.697 21.136 33.697 cp_fm_cholesky_invert 11 10.9 33.613 33.619 33.613 33.619 mp_waitall_1 104780 16.8 28.487 32.350 28.487 32.350 qs_ot_get_p 129 10.4 0.001 0.001 25.773 25.820 qs_ot_get_derivative 118 11.6 0.002 0.002 24.369 24.421 qs_ot_p2m_diag 84 11.4 0.889 0.895 21.753 21.784 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.911 20.927 calculate_rho_elec 129 8.7 0.487 0.488 20.910 20.926 make_m2s 5068 13.6 0.077 0.079 19.804 20.650 multiply_cannon_metrocomm3 10136 15.6 0.024 0.025 19.261 20.577 ot_diis_step 118 11.6 0.022 0.023 20.263 20.263 make_images 5068 14.6 3.780 3.914 19.320 20.169 sum_up_and_integrate 129 10.3 0.323 0.326 19.977 20.048 cp_dbcsr_syevd 84 12.4 0.007 0.007 20.000 20.002 integrate_v_rspace 129 11.3 0.004 0.004 19.653 19.725 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.408 19.672 apply_single 129 13.6 0.001 0.001 19.407 19.671 multiply_cannon_multrec 10136 15.6 10.548 12.294 17.986 18.080 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.770 16.771 cp_fm_diag_elpa_base 84 14.4 12.418 13.925 16.767 16.767 multiply_cannon_sync_h2d 10136 15.6 15.779 15.790 15.779 15.790 init_scf_run 11 5.9 0.000 0.001 14.155 14.155 scf_env_initial_rho_setup 11 6.9 0.001 0.001 14.154 14.155 pw_transfer 1559 11.6 0.094 0.094 12.453 12.459 make_images_data 5068 15.6 0.061 0.066 10.473 12.239 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 12.216 12.221 hybrid_alltoall_any 5255 16.5 1.310 3.051 10.566 12.108 fft_wrap_pw1pw2_140 527 13.2 0.547 0.549 10.837 10.848 fft3d_ps 1301 14.7 2.765 2.772 10.237 10.244 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.535 9.572 mp_alltoall_d11v 2429 14.1 8.286 9.215 8.286 9.215 wfi_extrapolate 11 7.9 0.001 0.001 9.097 9.097 dbcsr_mm_accdrv_process 20954 16.1 3.845 5.807 7.198 9.065 cp_fm_cholesky_decompose 22 10.9 8.962 8.992 8.962 8.992 grid_integrate_task_list 129 12.3 8.632 8.792 8.632 8.792 density_rs2pw 129 9.7 0.005 0.005 8.416 8.533 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 8.263 8.330 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.097 8.098 calculate_dm_sparse 129 9.5 0.001 0.001 6.658 6.757 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.456 6.507 grid_collocate_task_list 129 9.7 6.407 6.433 6.407 6.433 rs_scatter_matrices 140 9.7 3.635 4.602 6.056 6.296 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.034 6.111 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=300.809000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2712.454545, yerr=178.575354 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_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 1.262014E+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 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 6805335. 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 0.024 0.036 87.527 87.527 qs_energies 1 2.0 0.000 0.000 87.067 87.081 ls_scf 1 3.0 0.000 0.000 86.154 86.167 dbcsr_multiply_generic 111 6.7 0.014 0.015 75.200 75.389 multiply_cannon 111 7.7 0.018 0.020 58.533 59.756 multiply_cannon_loop 111 8.7 0.211 0.225 55.117 56.517 ls_scf_main 1 4.0 0.000 0.000 52.415 52.415 density_matrix_trs4 2 5.0 0.002 0.003 46.861 46.943 ls_scf_init_scf 1 4.0 0.000 0.000 30.710 30.712 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.597 29.648 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.471 27.492 mp_waitall_1 11031 10.9 22.641 26.807 22.641 26.807 multiply_cannon_multrec 2664 9.7 8.173 8.917 15.494 17.278 multiply_cannon_sync_h2d 2664 9.7 13.760 15.344 13.760 15.344 make_m2s 222 7.7 0.008 0.011 13.068 13.547 make_images 222 8.7 0.099 0.110 13.046 13.527 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.663 12.564 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.676 11.621 make_images_data 222 9.7 0.004 0.005 7.629 8.252 hybrid_alltoall_any 227 10.6 0.215 1.827 6.475 8.056 dbcsr_mm_accdrv_process 4760 10.4 0.515 0.603 6.940 7.931 dbcsr_mm_accdrv_process_sort 4760 11.4 6.225 7.139 6.225 7.139 calculate_norms 4752 9.8 5.515 6.178 5.515 6.178 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.062 5.167 mp_sum_l 807 5.4 3.065 4.483 3.065 4.483 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.063 3.898 mp_irecv_dv 6231 10.9 2.045 3.871 2.045 3.871 compute_matrix_preconditioner 1 6.0 0.000 0.001 3.598 3.603 make_images_sizes 222 9.7 0.000 0.000 0.753 3.375 mp_alltoall_i44 222 10.7 0.753 3.375 0.753 3.375 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.285 3.249 arnoldi_extremal 4 6.8 0.000 0.000 3.197 3.233 arnoldi_normal_ev 4 7.8 0.001 0.003 3.197 3.233 build_subspace 16 8.4 0.009 0.012 3.101 3.102 ls_scf_post 1 4.0 0.000 0.000 3.029 3.043 acc_transpose_blocks 2664 9.7 0.019 0.021 2.652 2.898 ls_scf_store_result 1 5.0 0.000 0.000 2.853 2.890 dbcsr_special_finalize 555 9.7 0.005 0.006 2.331 2.796 dbcsr_merge_single_wm 555 10.7 0.451 0.579 2.323 2.788 acc_transpose_blocks_kernels 2664 10.7 0.036 0.041 2.544 2.786 jit_kernel_transpose 1 13.0 2.508 2.751 2.508 2.751 make_images_pack 222 9.7 2.208 2.632 2.209 2.634 dbcsr_matrix_vector_mult 304 9.0 0.005 0.013 2.327 2.558 dbcsr_sort_data 658 11.4 2.131 2.552 2.131 2.552 dbcsr_matrix_vector_mult_local 304 10.0 2.070 2.457 2.072 2.459 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.304 2.421 buffer_matrices_ensure_size 222 8.7 1.743 2.085 1.743 2.085 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.787 1.788 rebuild_ks_matrix 3 7.3 0.000 0.000 1.777 1.779 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.777 1.779 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=87.527000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_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 2.111549E+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 266696. MP_Allreduce 3058 10339. 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 0.136 0.195 91.193 91.194 qs_energies 1 2.0 0.000 0.000 90.383 90.389 ls_scf 1 3.0 0.002 0.006 89.015 89.018 dbcsr_multiply_generic 111 6.7 0.015 0.015 74.955 75.283 multiply_cannon 111 7.7 0.028 0.039 53.296 56.846 ls_scf_main 1 4.0 0.000 0.000 54.669 54.676 multiply_cannon_loop 111 8.7 0.115 0.122 50.033 53.303 density_matrix_trs4 2 5.0 0.002 0.003 48.936 49.172 ls_scf_init_scf 1 4.0 0.000 0.002 30.881 30.883 ls_scf_init_matrix_S 1 5.0 0.000 0.002 29.529 29.657 mp_waitall_1 9105 10.9 20.956 29.192 20.956 29.192 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.155 27.172 multiply_cannon_multrec 1332 9.7 13.188 16.562 22.430 26.998 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.765 20.148 make_m2s 222 7.7 0.006 0.007 15.134 15.860 make_images 222 8.7 1.572 1.910 15.104 15.832 dbcsr_mm_accdrv_process 4041 10.4 0.278 0.468 8.841 10.371 dbcsr_mm_accdrv_process_sort 4041 11.4 8.423 9.927 8.423 9.927 make_images_data 222 9.7 0.004 0.004 8.658 9.620 hybrid_alltoall_any 227 10.6 0.522 2.457 8.062 9.083 mp_sum_l 807 5.4 5.235 7.988 5.235 7.988 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.221 7.647 mp_irecv_dv 3311 11.0 3.202 7.594 3.202 7.594 calculate_norms 2376 9.8 6.003 6.760 6.003 6.760 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.034 6.353 multiply_cannon_sync_h2d 1332 9.7 4.809 6.173 4.809 6.173 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.004 5.243 arnoldi_extremal 4 6.8 0.000 0.000 4.629 4.651 arnoldi_normal_ev 4 7.8 0.001 0.005 4.629 4.651 build_subspace 16 8.4 0.014 0.021 4.377 4.381 ls_scf_post 1 4.0 0.001 0.019 3.463 3.470 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.127 3.362 ls_scf_store_result 1 5.0 0.000 0.000 3.190 3.272 dbcsr_matrix_vector_mult_local 304 10.0 2.731 3.217 2.734 3.219 ls_scf_dm_to_ks 2 5.0 0.001 0.028 2.622 2.716 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.165 2.599 make_images_pack 222 9.7 2.024 2.440 2.026 2.442 mp_allgather_i34 111 8.7 0.968 2.374 0.968 2.374 dbcsr_sort_data 436 11.2 1.814 2.018 1.814 2.018 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.011 2.012 rebuild_ks_matrix 3 7.3 0.000 0.000 1.998 2.000 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.998 2.000 dbcsr_data_new 4174 10.1 1.610 1.848 1.610 1.848 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.194000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1768.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_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 2.719850E+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 265470. MP_Allreduce 3058 11181. 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 0.041 0.097 93.872 93.873 qs_energies 1 2.0 0.000 0.000 93.178 93.195 ls_scf 1 3.0 0.001 0.006 91.761 91.764 dbcsr_multiply_generic 111 6.7 0.015 0.016 76.372 76.645 ls_scf_main 1 4.0 0.000 0.000 57.239 57.250 multiply_cannon 111 7.7 0.044 0.145 52.719 56.691 multiply_cannon_loop 111 8.7 0.100 0.107 49.130 52.669 density_matrix_trs4 2 5.0 0.002 0.003 51.164 51.363 mp_waitall_1 7281 11.0 23.664 33.036 23.664 33.036 ls_scf_init_scf 1 4.0 0.001 0.005 30.916 30.927 ls_scf_init_matrix_S 1 5.0 0.000 0.006 29.666 29.748 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.260 27.275 multiply_cannon_multrec 888 9.7 12.563 15.078 21.093 24.438 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.960 22.123 make_m2s 222 7.7 0.006 0.006 16.975 18.275 make_images 222 8.7 1.971 2.290 16.937 18.236 hybrid_alltoall_any 227 10.6 0.619 2.864 9.416 10.667 make_images_data 222 9.7 0.004 0.008 9.678 10.601 dbcsr_mm_accdrv_process 3754 10.4 0.261 0.442 8.053 9.266 dbcsr_mm_accdrv_process_sort 3754 11.4 7.681 8.824 7.681 8.824 mp_sum_l 807 5.4 5.137 8.177 5.137 8.177 multiply_cannon_sync_h2d 888 9.7 6.036 7.319 6.036 7.319 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.473 7.097 mp_irecv_dv 2335 11.1 2.457 7.050 2.457 7.050 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.645 6.695 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.844 6.366 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.987 5.203 arnoldi_extremal 4 6.8 0.000 0.000 5.127 5.147 arnoldi_normal_ev 4 7.8 0.001 0.005 5.127 5.147 build_subspace 16 8.4 0.014 0.020 4.821 4.827 calculate_norms 1584 9.8 4.266 4.567 4.266 4.567 mp_allgather_i34 111 8.7 1.431 3.804 1.431 3.804 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.449 3.787 ls_scf_post 1 4.0 0.000 0.002 3.606 3.611 dbcsr_matrix_vector_mult_local 304 10.0 3.003 3.577 3.005 3.579 ls_scf_store_result 1 5.0 0.000 0.000 3.347 3.435 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.888 2.980 make_images_pack 222 9.7 1.817 2.145 1.820 2.148 dbcsr_sort_data 325 11.1 1.900 2.081 1.900 2.081 make_images_sizes 222 9.7 0.000 0.000 0.966 1.986 mp_alltoall_i44 222 10.7 0.966 1.986 0.966 1.986 dbcsr_data_release 9322 10.9 1.327 1.971 1.327 1.971 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.948 1.958 rebuild_ks_matrix 3 7.3 0.000 0.000 1.929 1.940 qs_ks_build_kohn_sham_matrix 3 8.3 0.009 0.024 1.929 1.940 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.873000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2211.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_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 3.332039E+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 266696. MP_Allreduce 3058 13371. 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 0.038 0.059 97.475 97.477 qs_energies 1 2.0 0.000 0.000 96.608 96.614 ls_scf 1 3.0 0.000 0.001 94.897 94.902 dbcsr_multiply_generic 111 6.7 0.017 0.022 78.581 78.841 ls_scf_main 1 4.0 0.000 0.000 59.036 59.037 multiply_cannon 111 7.7 0.059 0.119 51.445 55.478 density_matrix_trs4 2 5.0 0.002 0.004 52.918 53.065 multiply_cannon_loop 111 8.7 0.115 0.126 46.330 50.081 ls_scf_init_scf 1 4.0 0.000 0.002 32.625 32.628 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.331 31.420 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.866 28.882 mp_waitall_1 6369 11.0 22.440 28.442 22.440 28.442 multiply_cannon_multrec 1332 9.7 14.165 17.428 22.014 24.608 make_m2s 222 7.7 0.007 0.008 21.086 22.539 make_images 222 8.7 3.130 3.589 21.035 22.491 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 8.761 17.532 make_images_data 222 9.7 0.004 0.006 11.717 13.330 hybrid_alltoall_any 227 10.6 0.795 3.742 11.026 12.812 dbcsr_mm_accdrv_process 3641 10.4 0.204 0.402 7.498 9.036 dbcsr_mm_accdrv_process_sort 3641 11.4 7.122 8.620 7.122 8.620 mp_sum_l 807 5.4 4.474 8.258 4.474 8.258 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.503 6.706 multiply_cannon_sync_h2d 1332 9.7 5.502 6.018 5.502 6.018 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.100 5.976 mp_irecv_dv 3229 10.9 2.077 5.932 2.077 5.932 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.789 5.535 arnoldi_extremal 4 6.8 0.000 0.000 5.195 5.210 arnoldi_normal_ev 4 7.8 0.001 0.005 5.195 5.210 build_subspace 16 8.4 0.014 0.021 4.866 4.875 mp_allgather_i34 111 8.7 2.236 4.751 2.236 4.751 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.521 4.705 calculate_norms 2376 9.8 4.186 4.532 4.186 4.532 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.577 3.864 dbcsr_matrix_vector_mult_local 304 10.0 3.163 3.663 3.165 3.664 dbcsr_sort_data 658 11.4 3.054 3.440 3.054 3.440 ls_scf_post 1 4.0 0.000 0.001 3.236 3.241 dbcsr_special_finalize 555 9.7 0.007 0.026 2.799 3.218 dbcsr_merge_single_wm 555 10.7 0.539 0.662 2.790 3.210 ls_scf_dm_to_ks 2 5.0 0.000 0.004 3.029 3.080 ls_scf_store_result 1 5.0 0.000 0.000 2.983 3.040 dbcsr_data_release 10477 10.7 1.574 2.412 1.574 2.412 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.019 2.020 rebuild_ks_matrix 3 7.3 0.000 0.000 1.996 1.997 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.004 1.996 1.997 dbcsr_finalize 304 7.8 0.049 0.062 1.804 1.983 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.477000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2748.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_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 4.624470E+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 265558. MP_Allreduce 3049 15663. 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 0.040 0.056 99.082 99.082 qs_energies 1 2.0 0.000 0.000 98.329 98.336 ls_scf 1 3.0 0.000 0.000 96.370 96.381 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.583 77.801 ls_scf_main 1 4.0 0.000 0.000 62.156 62.156 multiply_cannon 111 7.7 0.121 0.238 55.281 60.839 density_matrix_trs4 2 5.0 0.002 0.003 54.810 54.901 multiply_cannon_loop 111 8.7 0.069 0.080 50.696 52.509 mp_waitall_1 5436 11.0 26.097 32.109 26.097 32.109 ls_scf_init_scf 1 4.0 0.000 0.000 30.603 30.609 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.229 29.279 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.087 27.098 multiply_cannon_multrec 444 9.7 14.008 16.292 21.045 24.103 make_m2s 222 7.7 0.004 0.005 17.366 19.915 make_images 222 8.7 3.724 4.432 17.304 19.855 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.908 16.144 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.307 14.138 make_images_data 222 9.7 0.003 0.004 9.687 12.089 hybrid_alltoall_any 227 10.6 0.792 3.789 9.408 11.878 multiply_cannon_sync_h2d 444 9.7 6.542 8.894 6.542 8.894 dbcsr_mm_accdrv_process 3003 10.4 0.175 0.335 6.739 7.867 dbcsr_mm_accdrv_process_sort 3003 11.4 6.425 7.516 6.425 7.516 mp_allgather_i34 111 8.7 2.821 6.959 2.821 6.959 arnoldi_extremal 4 6.8 0.000 0.000 5.824 5.835 arnoldi_normal_ev 4 7.8 0.002 0.005 5.824 5.835 build_subspace 16 8.4 0.015 0.020 5.441 5.449 mp_sum_l 807 5.4 2.982 5.278 2.982 5.278 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.558 4.733 multiply_cannon_metrocomm4 333 9.7 0.001 0.001 1.647 4.535 mp_irecv_dv 1241 11.2 1.627 4.508 1.627 4.508 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.124 4.325 dbcsr_matrix_vector_mult_local 304 10.0 3.659 4.116 3.661 4.118 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.045 3.941 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.763 3.858 calculate_norms 792 9.8 3.560 3.705 3.560 3.705 ls_scf_post 1 4.0 0.000 0.000 3.610 3.617 ls_scf_store_result 1 5.0 0.000 0.000 3.387 3.440 make_images_sizes 222 9.7 0.000 0.000 1.026 3.242 mp_alltoall_i44 222 10.7 1.025 3.242 1.025 3.242 dbcsr_finalize 304 7.8 0.062 0.078 2.201 2.276 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.256 2.257 rebuild_ks_matrix 3 7.3 0.000 0.000 2.224 2.225 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.224 2.225 dbcsr_merge_all 275 8.9 0.473 0.536 2.051 2.113 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.082000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3590.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/ba7248420c2af75c068fa8700be06de812d89520_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 8.715940E+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 284111. MP_Allreduce 3043 21950. 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 0.079 0.099 108.184 108.184 qs_energies 1 2.0 0.000 0.000 106.783 106.798 ls_scf 1 3.0 0.000 0.000 103.852 103.866 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.191 77.313 ls_scf_main 1 4.0 0.000 0.000 65.947 65.974 density_matrix_trs4 2 5.0 0.002 0.004 56.599 56.649 multiply_cannon 111 7.7 0.147 0.268 49.673 51.623 multiply_cannon_loop 111 8.7 0.068 0.070 46.210 47.090 ls_scf_init_scf 1 4.0 0.000 0.000 34.263 34.268 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.573 32.592 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.764 29.775 mp_waitall_1 4527 11.1 22.029 25.525 22.029 25.525 make_m2s 222 7.7 0.005 0.005 23.976 24.974 make_images 222 8.7 4.589 4.988 23.869 24.866 multiply_cannon_multrec 444 9.7 17.850 18.674 22.452 23.159 hybrid_alltoall_any 227 10.6 1.659 3.618 12.942 15.937 make_images_data 222 9.7 0.003 0.003 13.183 15.605 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.269 10.829 multiply_cannon_sync_h2d 444 9.7 8.844 8.900 8.844 8.900 arnoldi_extremal 4 6.8 0.000 0.000 7.342 7.350 arnoldi_normal_ev 4 7.8 0.003 0.009 7.342 7.350 build_subspace 16 8.4 0.026 0.036 6.794 6.807 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.413 5.579 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.318 5.411 dbcsr_matrix_vector_mult_local 304 10.0 4.981 5.300 4.983 5.302 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.994 5.240 dbcsr_mm_accdrv_process 1814 10.4 0.252 0.314 4.426 4.545 dbcsr_mm_accdrv_process_sort 1814 11.4 4.128 4.251 4.128 4.251 ls_scf_post 1 4.0 0.000 0.000 3.641 3.655 make_images_sizes 222 9.7 0.000 0.000 1.454 3.546 mp_alltoall_i44 222 10.7 1.454 3.546 1.454 3.546 mp_allgather_i34 111 8.7 1.107 3.491 1.107 3.491 ls_scf_store_result 1 5.0 0.000 0.000 3.393 3.404 calculate_norms 792 9.8 3.243 3.279 3.243 3.279 dbcsr_finalize 304 7.8 0.082 0.090 3.089 3.168 dbcsr_merge_all 275 8.9 0.891 0.920 2.874 2.949 dbcsr_complete_redistribute 5 7.6 1.447 1.481 2.784 2.908 qs_energies_init_hamiltonians 1 3.0 0.002 0.002 2.901 2.901 dbcsr_data_release 12724 10.6 2.338 2.856 2.338 2.856 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.630 2.635 rebuild_ks_matrix 3 7.3 0.000 0.000 2.564 2.569 qs_ks_build_kohn_sham_matrix 3 8.3 0.027 0.031 2.564 2.569 matrix_ls_to_qs 2 6.0 0.000 0.000 2.422 2.556 dbcsr_sort_data 325 11.1 2.444 2.504 2.444 2.504 dbcsr_new_transposed 4 7.5 0.244 0.254 2.296 2.316 dbcsr_frobenius_norm 74 6.6 2.062 2.139 2.208 2.247 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.205 dbcsr_add_anytype 103 7.2 0.860 0.893 2.132 2.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.184000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6891.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: ba7248420c2af75c068fa8700be06de812d89520 Summary: empty Status: OK