=== 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: 444a873b469bbe29fdbc1e637b703a89687f2299 ################# 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 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (03.02.2023) # # \ 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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/01 job id: 44971141 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/02 job id: 44971142 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/03 job id: 44971143 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/04 job id: 44971144 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/05 job id: 44971145 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/06 job id: 44971146 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/07 job id: 44971147 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/08 job id: 44971148 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/09 job id: 44971149 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/10 job id: 44971150 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/11 job id: 44971151 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/12 job id: 44971152 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/13 job id: 44971154 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/14 job id: 44971155 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/15 job id: 44971156 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/16 job id: 44971157 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/17 job id: 44971158 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/18 job id: 44971159 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/19 job id: 44971160 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/20 job id: 44971161 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/21 job id: 44971162 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/22 job id: 44971164 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/23 job id: 44971172 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/24 job id: 44971173 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/25 job id: 44971174 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/26 job id: 44971176 --- 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/444a873b469bbe29fdbc1e637b703a89687f2299_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 0.037 135.945 135.946 farming_run 1 2.0 135.321 135.324 135.915 135.919 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.457832E+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 229 1108280. 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 0.029 115.673 115.674 qs_energies 1 2.0 0.000 0.000 115.365 115.367 mp2_main 1 3.0 0.000 0.000 113.381 113.383 mp2_gpw_main 1 4.0 0.030 0.037 112.545 112.547 mp2_ri_gpw_compute_in 1 5.0 0.189 0.334 93.428 93.839 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.385 55.797 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.169 41.695 47.000 get_2c_integrals 1 6.0 0.000 0.000 36.992 37.715 integrate_v_rspace 273 8.0 0.434 0.448 25.105 30.133 pw_transfer 6555 10.6 0.374 0.388 27.387 27.847 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.067 26.520 grid_integrate_task_list 273 9.0 20.914 26.392 20.914 26.392 fft_wrap_pw1pw2_100 2178 12.4 1.175 1.297 23.580 24.042 compute_2c_integrals 1 7.0 0.003 0.004 19.226 19.227 rpa_ri_compute_en 1 5.0 0.020 0.023 18.998 19.130 compute_2c_integrals_loop_lm 1 8.0 0.009 0.009 18.853 18.949 mp2_eri_2c_integrate_gpw 1 9.0 2.380 2.413 18.844 18.941 cp_fm_cholesky_decompose 12 8.2 17.774 18.517 17.774 18.517 cholesky_decomp 1 7.0 0.000 0.000 16.621 17.360 fft3d_s 5443 13.4 16.163 16.615 16.185 16.635 ao_to_mo_and_store_B_mult_1 272 7.0 10.851 15.571 10.851 15.571 calculate_wavefunction 272 8.0 5.402 5.546 12.519 13.146 rpa_num_int 1 6.0 0.002 0.034 10.645 10.654 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.569 10.611 calc_mat_Q 8 8.0 0.000 0.000 9.370 9.493 contract_S_to_Q 8 9.0 0.000 0.000 8.794 8.912 calc_potential_gpw 544 9.5 0.005 0.005 8.209 8.626 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 0.108 0.111 7.711 8.430 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.191 8.396 create_integ_mat 1 6.0 0.014 0.027 7.824 7.834 collocate_single_gaussian 272 10.0 0.039 0.042 7.451 7.692 array2fm 1 7.0 0.000 0.000 6.698 7.122 pw_scatter_s 2720 13.7 4.442 4.610 4.442 4.610 pw_gather_s 2722 13.2 3.870 4.171 3.870 4.171 array2fm_buffer_send 1 8.0 2.964 3.113 2.964 3.113 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.544471, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2729.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 0.031 0.046 403.350 403.352 farming_run 1 2.0 402.481 402.488 403.295 403.299 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.227010E+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 704 407793. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 2369671. 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 0.014 0.045 210.720 210.721 qs_energies 1 2.0 0.000 0.000 210.402 210.408 scf_env_do_scf 1 3.0 0.000 0.000 106.678 106.678 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.721 105.730 rebuild_ks_matrix 4 6.0 0.000 0.000 105.720 105.729 qs_ks_build_kohn_sham_matrix 4 7.0 0.061 0.070 105.720 105.729 hfx_ks_matrix 4 8.0 0.001 0.001 105.303 105.307 integrate_four_center 4 9.0 0.144 0.457 105.303 105.306 mp2_main 1 3.0 0.000 0.001 103.423 103.429 mp2_gpw_main 1 4.0 0.035 0.050 102.470 102.478 integrate_four_center_main 4 10.0 0.099 0.498 97.240 99.867 integrate_four_center_bin 266 11.0 97.141 99.584 97.141 99.584 init_scf_loop 1 4.0 0.000 0.000 92.361 92.362 mp2_ri_gpw_compute_in 1 5.0 0.064 0.066 75.192 76.394 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.619 55.821 mp2_eri_3c_integrate_gpw 91 7.0 0.146 0.164 42.288 47.456 integrate_v_rspace 95 8.0 0.397 0.567 28.683 33.631 pw_transfer 2240 10.6 0.147 0.197 29.858 30.327 ao_to_mo_and_store_B_mult_1 91 7.0 10.648 29.366 10.648 29.366 fft_wrap_pw1pw2 1868 11.4 0.019 0.025 28.885 29.362 grid_integrate_task_list 95 9.0 23.941 29.092 23.941 29.092 mp2_ri_gpw_compute_en 1 5.0 0.059 0.080 27.118 29.041 fft_wrap_pw1pw2_100 730 12.4 1.248 1.534 26.582 27.097 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.837 1.890 25.253 25.263 get_2c_integrals 1 6.0 0.000 0.000 20.459 20.510 compute_2c_integrals 1 7.0 0.003 0.004 19.436 19.441 compute_2c_integrals_loop_lm 1 8.0 0.002 0.004 18.719 19.308 mp2_eri_2c_integrate_gpw 1 9.0 1.737 1.902 18.718 19.307 fft3d_s 1823 13.4 18.398 18.811 18.411 18.824 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.314 14.314 calculate_wavefunction 91 8.0 2.011 2.046 9.741 9.986 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.586 8.720 9.381 potential_pw2rs 186 10.0 0.033 0.036 8.615 9.259 local_gemm 172 8.0 8.163 8.795 8.163 8.795 mp2_ri_gpw_compute_en_comm 22 7.0 0.495 0.520 8.300 8.761 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.161 8.582 calc_potential_gpw 182 9.5 0.002 0.002 7.890 8.217 collocate_single_gaussian 91 10.0 0.017 0.023 7.817 8.134 mp_sendrecv_dm3 2068 8.0 6.340 6.812 6.340 6.812 mp2_ri_gpw_compute_en_ener 172 7.0 6.352 6.464 6.352 6.464 pw_gather_s 912 13.2 4.904 5.343 4.904 5.343 mp_sync 38 10.4 2.741 4.900 2.741 4.900 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.467408, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1510.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 452.116480E+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 578368. 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 0.014 0.034 54.999 55.002 qs_mol_dyn_low 1 2.0 0.003 0.006 54.682 54.690 qs_forces 11 3.9 0.002 0.002 54.611 54.612 qs_energies 11 4.9 0.001 0.002 53.130 53.146 scf_env_do_scf 11 5.9 0.000 0.001 45.461 45.461 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 43.233 43.233 dbcsr_multiply_generic 2286 12.5 0.093 0.096 35.581 36.046 qs_scf_new_mos 108 7.5 0.000 0.001 32.789 33.089 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.789 33.089 ot_scf_mini 108 9.5 0.002 0.002 31.143 31.346 multiply_cannon 2286 13.5 0.184 0.199 27.831 29.484 multiply_cannon_loop 2286 14.5 1.484 1.563 27.068 28.701 velocity_verlet 10 3.0 0.001 0.002 26.443 26.444 ot_mini 108 10.5 0.001 0.001 19.770 20.025 qs_ot_get_derivative 108 11.5 0.001 0.001 16.734 16.980 mp_waitall_1 245248 16.5 9.254 16.460 9.254 16.460 multiply_cannon_metrocomm3 54864 15.5 0.068 0.072 6.616 14.896 multiply_cannon_multrec 54864 15.5 4.210 6.480 7.778 11.205 rebuild_ks_matrix 119 8.3 0.000 0.000 8.083 8.233 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.083 8.233 multiply_cannon_sync_h2d 54864 15.5 5.891 7.698 5.891 7.698 mp_sum_l 7207 12.9 5.629 7.426 5.629 7.426 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.133 7.266 qs_ot_get_p 119 10.4 0.001 0.001 6.691 7.024 init_scf_run 11 5.9 0.000 0.001 6.424 6.424 scf_env_initial_rho_setup 11 6.9 0.000 0.000 6.424 6.424 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.646 6.122 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.552 5.669 dbcsr_mm_accdrv_process 76910 16.1 1.180 1.826 3.491 4.838 sum_up_and_integrate 119 10.3 0.012 0.014 4.656 4.664 integrate_v_rspace 119 11.3 0.002 0.002 4.644 4.653 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.405 4.501 calculate_rho_elec 119 8.7 0.011 0.017 4.405 4.501 calculate_first_density_matrix 1 7.0 0.001 0.001 4.174 4.179 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.861 3.863 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.752 3.826 acc_transpose_blocks 54864 15.5 0.235 0.253 2.740 3.645 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.566 3.613 multiply_cannon_metrocomm1 54864 15.5 0.051 0.056 1.858 3.344 calculate_dm_sparse 119 9.5 0.000 0.001 2.989 3.124 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.902 3.090 apply_single 119 13.6 0.000 0.000 2.901 3.090 rs_pw_transfer 974 11.9 0.011 0.012 2.906 3.008 jit_kernel_multiply 13 15.8 2.249 2.926 2.249 2.926 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.888 2.889 ot_diis_step 108 11.5 0.006 0.006 2.697 2.698 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.660 2.661 cp_fm_redistribute_end 50 14.0 2.421 2.638 2.426 2.639 cp_fm_diag_elpa_base 50 14.0 0.211 2.548 0.212 2.560 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.496 2.547 density_rs2pw 119 9.7 0.004 0.005 2.414 2.504 acc_transpose_blocks_kernels 54864 16.5 0.254 0.386 1.744 2.494 init_scf_loop 11 6.9 0.000 0.000 2.210 2.210 wfi_extrapolate 11 7.9 0.001 0.001 2.190 2.190 grid_integrate_task_list 119 12.3 2.014 2.117 2.014 2.117 jit_kernel_transpose 5 15.5 1.490 2.117 1.490 2.117 pw_transfer 1439 11.6 0.053 0.058 1.962 2.062 mp_sum_d 4127 12.0 1.394 2.049 1.394 2.049 potential_pw2rs 119 12.3 0.004 0.004 1.999 2.017 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.886 1.990 make_m2s 4572 13.5 0.054 0.056 1.822 1.877 make_images 4572 14.5 0.132 0.137 1.739 1.793 fft3d_ps 1201 14.6 0.369 0.473 1.658 1.759 fft_wrap_pw1pw2_140 487 13.2 0.082 0.097 1.455 1.558 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.516 1.553 mp_alltoall_d11v 2130 13.8 1.345 1.525 1.345 1.525 mp_waitany 12084 13.8 1.281 1.464 1.281 1.464 grid_collocate_task_list 119 9.7 1.292 1.361 1.292 1.361 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.797 1.174 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.002000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 487.284736E+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 1031905. 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 0.031 0.051 41.376 41.384 qs_mol_dyn_low 1 2.0 0.003 0.004 41.101 41.108 qs_forces 11 3.9 0.005 0.031 41.010 41.012 qs_energies 11 4.9 0.008 0.057 39.222 39.228 scf_env_do_scf 11 5.9 0.001 0.001 32.397 32.397 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 29.659 29.660 dbcsr_multiply_generic 2286 12.5 0.100 0.102 22.887 23.237 qs_scf_new_mos 108 7.5 0.001 0.001 20.139 20.372 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.138 20.372 multiply_cannon 2286 13.5 0.206 0.217 17.853 19.557 ot_scf_mini 108 9.5 0.003 0.003 19.253 19.417 velocity_verlet 10 3.0 0.007 0.036 18.971 18.972 multiply_cannon_loop 2286 14.5 0.895 0.975 16.733 18.197 mp_waitall_1 200699 16.5 6.205 12.424 6.205 12.424 ot_mini 108 10.5 0.001 0.001 11.874 12.102 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.707 10.948 qs_ot_get_derivative 108 11.5 0.001 0.005 9.414 9.580 multiply_cannon_multrec 27432 15.5 1.968 4.561 6.113 9.259 rebuild_ks_matrix 119 8.3 0.000 0.000 7.426 7.570 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 7.425 7.569 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.515 6.645 dbcsr_mm_accdrv_process 47894 16.0 3.044 5.207 4.076 6.097 init_scf_run 11 5.9 0.000 0.001 5.395 5.395 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.394 5.395 qs_ot_get_p 119 10.4 0.001 0.002 4.512 4.746 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.685 4.527 sum_up_and_integrate 119 10.3 0.024 0.027 4.375 4.381 integrate_v_rspace 119 11.3 0.002 0.003 4.350 4.360 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.041 4.129 apply_single 119 13.6 0.000 0.000 3.040 4.129 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.029 4.069 calculate_rho_elec 119 8.7 0.021 0.024 4.029 4.069 mp_sum_l 7207 12.9 2.069 3.964 2.069 3.964 calculate_first_density_matrix 1 7.0 0.001 0.002 3.827 3.829 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.307 3.309 rs_pw_transfer 974 11.9 0.010 0.011 2.830 3.270 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.081 3.124 qs_ot_p2m_diag 50 11.0 0.008 0.013 3.011 3.028 acc_transpose_blocks 27432 15.5 0.109 0.113 2.086 2.834 multiply_cannon_sync_h2d 27432 15.5 2.169 2.810 2.169 2.810 density_rs2pw 119 9.7 0.004 0.004 2.349 2.798 make_m2s 4572 13.5 0.052 0.053 2.526 2.760 init_scf_loop 11 6.9 0.001 0.006 2.711 2.714 make_images 4572 14.5 0.199 0.235 2.438 2.668 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.580 2.581 jit_kernel_multiply 11 16.0 0.980 2.465 0.980 2.465 ot_diis_step 108 11.5 0.011 0.011 2.405 2.406 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.241 2.241 cp_fm_redistribute_end 50 14.0 1.855 2.216 1.859 2.217 acc_transpose_blocks_kernels 27432 16.5 0.182 0.272 1.554 2.217 calculate_dm_sparse 119 9.5 0.000 0.001 2.140 2.216 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.126 2.215 cp_fm_diag_elpa_base 50 14.0 0.343 2.102 0.356 2.163 pw_transfer 1439 11.6 0.065 0.072 2.115 2.155 potential_pw2rs 119 12.3 0.006 0.006 2.089 2.106 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.023 2.066 jit_kernel_transpose 5 15.5 1.372 1.950 1.372 1.950 grid_integrate_task_list 119 12.3 1.837 1.930 1.837 1.930 prepare_preconditioner 11 7.9 0.000 0.001 1.746 1.772 make_preconditioner 11 8.9 0.001 0.004 1.745 1.772 fft3d_ps 1201 14.6 0.511 0.565 1.723 1.760 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.635 1.693 fft_wrap_pw1pw2_140 487 13.2 0.079 0.086 1.552 1.599 make_images_data 4572 15.5 0.044 0.050 1.156 1.557 wfi_extrapolate 11 7.9 0.001 0.001 1.513 1.513 hybrid_alltoall_any 4725 16.4 0.050 0.111 1.022 1.451 mp_alltoall_d11v 2130 13.8 1.255 1.426 1.255 1.426 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.364 1.373 grid_collocate_task_list 119 9.7 1.229 1.371 1.229 1.371 mp_allgather_i34 2286 14.5 0.566 1.318 0.566 1.318 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.168 1.214 mp_sum_d 4127 12.0 0.622 1.061 0.622 1.061 mp_waitany 5720 13.7 0.531 0.992 0.531 0.992 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.971 0.989 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.975 0.978 rs_pw_transfer_RS2PW_140 130 11.5 0.138 0.146 0.546 0.974 mp_alltoall_z22v 1201 16.6 0.789 0.892 0.789 0.892 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.607 0.787 0.851 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.384000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.363636, yerr=1.149919 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 522.661888E+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 1607811. 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 0.032 0.060 36.552 36.557 qs_mol_dyn_low 1 2.0 0.003 0.003 36.248 36.256 qs_forces 11 3.9 0.002 0.002 35.911 35.912 qs_energies 11 4.9 0.002 0.002 34.307 34.310 scf_env_do_scf 11 5.9 0.000 0.001 27.482 27.482 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 24.669 24.669 dbcsr_multiply_generic 2286 12.5 0.094 0.096 18.929 19.085 velocity_verlet 10 3.0 0.001 0.002 16.646 16.656 multiply_cannon 2286 13.5 0.196 0.204 15.217 16.192 qs_scf_new_mos 108 7.5 0.001 0.001 15.728 15.748 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.727 15.748 multiply_cannon_loop 2286 14.5 0.633 0.666 14.344 15.421 ot_scf_mini 108 9.5 0.002 0.002 14.976 14.989 ot_mini 108 10.5 0.001 0.001 9.260 9.279 qs_ot_get_derivative 108 11.5 0.001 0.001 7.679 7.695 multiply_cannon_multrec 18288 15.5 1.937 2.879 7.061 7.591 rebuild_ks_matrix 119 8.3 0.000 0.000 6.791 6.819 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.790 6.819 dbcsr_mm_accdrv_process 38222 16.0 4.130 5.613 5.041 6.050 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.022 6.048 init_scf_run 11 5.9 0.000 0.001 5.579 5.579 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.579 5.579 mp_waitall_1 158411 16.6 3.360 4.600 3.360 4.600 sum_up_and_integrate 119 10.3 0.029 0.030 4.318 4.323 integrate_v_rspace 119 11.3 0.002 0.003 4.289 4.297 calculate_first_density_matrix 1 7.0 0.012 0.013 4.283 4.283 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.857 3.869 calculate_rho_elec 119 8.7 0.030 0.031 3.857 3.869 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.576 3.578 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.910 3.546 qs_ot_get_p 119 10.4 0.001 0.001 3.483 3.504 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.346 3.352 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.792 3.090 rs_pw_transfer 974 11.9 0.009 0.010 2.735 3.001 init_scf_loop 11 6.9 0.000 0.000 2.789 2.789 acc_transpose_blocks 18288 15.5 0.076 0.078 2.641 2.670 density_rs2pw 119 9.7 0.004 0.004 2.352 2.634 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.100 2.404 apply_single 119 13.6 0.000 0.000 2.100 2.404 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.247 2.255 acc_transpose_blocks_kernels 18288 16.5 0.210 0.217 2.203 2.226 pw_transfer 1439 11.6 0.065 0.070 2.183 2.216 make_m2s 4572 13.5 0.044 0.045 2.055 2.191 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.090 2.125 make_images 4572 14.5 0.188 0.199 1.970 2.105 potential_pw2rs 119 12.3 0.007 0.008 2.060 2.079 jit_kernel_multiply 10 15.9 0.859 2.063 0.859 2.063 jit_kernel_transpose 5 15.6 1.993 2.010 1.993 2.010 calculate_dm_sparse 119 9.5 0.000 0.000 1.995 2.008 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.960 1.961 prepare_preconditioner 11 7.9 0.000 0.000 1.954 1.956 make_preconditioner 11 8.9 0.000 0.000 1.954 1.956 grid_integrate_task_list 119 12.3 1.803 1.892 1.803 1.892 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.792 1.880 fft3d_ps 1201 14.6 0.523 0.544 1.771 1.809 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.712 1.721 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.654 1.656 mp_sum_l 7207 12.9 1.277 1.635 1.277 1.635 cp_fm_redistribute_end 50 14.0 1.228 1.624 1.230 1.625 fft_wrap_pw1pw2_140 487 13.2 0.089 0.092 1.586 1.615 multiply_cannon_sync_h2d 18288 15.5 1.369 1.580 1.369 1.580 cp_fm_diag_elpa_base 50 14.0 0.377 1.527 0.391 1.574 ot_diis_step 108 11.5 0.011 0.011 1.547 1.547 grid_collocate_task_list 119 9.7 1.210 1.395 1.210 1.395 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.341 1.346 wfi_extrapolate 11 7.9 0.001 0.001 1.247 1.247 make_images_data 4572 15.5 0.044 0.048 0.930 1.106 multiply_cannon_metrocomm1 18288 15.5 0.027 0.029 0.447 1.009 mp_alltoall_z22v 1201 16.6 0.829 0.970 0.829 0.970 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.815 0.967 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.950 0.951 mp_alltoall_d11v 2130 13.8 0.816 0.948 0.816 0.948 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.915 0.944 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.841 0.843 cp_fm_cholesky_invert 11 10.9 0.830 0.834 0.830 0.834 mp_waitany 9880 13.7 0.545 0.818 0.545 0.818 rs_pw_transfer_RS2PW_140 130 11.5 0.120 0.132 0.530 0.795 mp_allgather_i34 2286 14.5 0.338 0.765 0.338 0.765 dbcsr_complete_redistribute 329 12.2 0.137 0.161 0.645 0.752 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=36.557000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.181818, yerr=1.526623 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 544.874496E+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 0.024 0.038 37.594 37.595 qs_mol_dyn_low 1 2.0 0.003 0.004 37.307 37.315 qs_forces 11 3.9 0.002 0.006 37.241 37.242 qs_energies 11 4.9 0.002 0.004 35.493 35.500 scf_env_do_scf 11 5.9 0.001 0.002 28.887 28.889 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.347 25.347 dbcsr_multiply_generic 2286 12.5 0.099 0.101 19.720 19.844 velocity_verlet 10 3.0 0.002 0.002 18.284 18.286 qs_scf_new_mos 108 7.5 0.001 0.001 16.538 16.589 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.537 16.588 multiply_cannon 2286 13.5 0.229 0.271 15.997 16.399 ot_scf_mini 108 9.5 0.002 0.003 15.592 15.640 multiply_cannon_loop 2286 14.5 0.930 0.965 15.014 15.389 ot_mini 108 10.5 0.001 0.001 9.457 9.521 multiply_cannon_multrec 27432 15.5 2.395 3.076 8.821 9.160 qs_ot_get_derivative 108 11.5 0.001 0.002 7.613 7.661 dbcsr_mm_accdrv_process 47916 15.9 5.376 7.373 6.333 7.596 rebuild_ks_matrix 119 8.3 0.000 0.000 6.787 6.842 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 6.787 6.841 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.024 6.072 init_scf_run 11 5.9 0.000 0.001 5.195 5.195 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.194 5.195 sum_up_and_integrate 119 10.3 0.035 0.038 3.982 3.990 integrate_v_rspace 119 11.3 0.003 0.004 3.947 3.955 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.746 3.776 calculate_rho_elec 119 8.7 0.040 0.046 3.746 3.776 calculate_first_density_matrix 1 7.0 0.002 0.009 3.765 3.766 qs_ot_get_p 119 10.4 0.001 0.001 3.493 3.570 init_scf_loop 11 6.9 0.000 0.002 3.518 3.518 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.224 3.227 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.813 3.206 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.991 3.005 acc_transpose_blocks 27432 15.5 0.112 0.114 2.782 2.818 prepare_preconditioner 11 7.9 0.000 0.001 2.655 2.665 make_preconditioner 11 8.9 0.000 0.000 2.655 2.665 mp_waitall_1 137007 16.6 1.932 2.596 1.932 2.596 make_full_inverse_cholesky 11 9.9 0.000 0.001 2.267 2.591 rs_pw_transfer 974 11.9 0.009 0.009 2.236 2.434 make_m2s 4572 13.5 0.054 0.056 2.280 2.416 density_rs2pw 119 9.7 0.004 0.004 2.152 2.348 make_images 4572 14.5 0.269 0.330 2.173 2.308 pw_transfer 1439 11.6 0.065 0.072 2.191 2.230 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.064 2.193 apply_single 119 13.6 0.000 0.000 2.064 2.193 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.177 2.189 acc_transpose_blocks_kernels 27432 16.5 0.268 0.277 2.161 2.184 calculate_dm_sparse 119 9.5 0.000 0.000 2.125 2.183 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.098 2.142 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.968 1.995 jit_kernel_multiply 10 15.9 0.897 1.984 0.897 1.984 grid_integrate_task_list 119 12.3 1.838 1.926 1.838 1.926 jit_kernel_transpose 5 15.6 1.893 1.911 1.893 1.911 cp_dbcsr_syevd 50 12.0 0.004 0.009 1.849 1.850 fft3d_ps 1201 14.6 0.558 0.610 1.773 1.810 ot_diis_step 108 11.5 0.012 0.012 1.793 1.794 fft_wrap_pw1pw2_140 487 13.2 0.088 0.096 1.709 1.757 potential_pw2rs 119 12.3 0.008 0.010 1.708 1.717 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.580 1.584 cp_fm_redistribute_end 50 14.0 1.043 1.551 1.045 1.552 cp_fm_diag_elpa_base 50 14.0 0.484 1.479 0.503 1.520 multiply_cannon_metrocomm3 27432 15.5 0.037 0.038 0.793 1.448 wfi_extrapolate 11 7.9 0.001 0.002 1.379 1.379 grid_collocate_task_list 119 9.7 1.222 1.360 1.222 1.360 mp_sum_l 7207 12.9 1.023 1.336 1.023 1.336 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.298 1.309 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.131 1.148 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.125 1.126 dbcsr_complete_redistribute 329 12.2 0.149 0.198 0.854 1.123 cp_fm_upper_to_full 72 13.5 0.802 1.109 0.802 1.109 multiply_cannon_sync_h2d 27432 15.5 0.997 1.056 0.997 1.056 make_images_data 4572 15.5 0.044 0.048 0.904 1.045 hybrid_alltoall_any 4725 16.4 0.061 0.152 0.787 0.977 mp_alltoall_d11v 2130 13.8 0.789 0.954 0.789 0.954 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.629 0.889 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.871 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.854 0.858 mp_alltoall_z22v 1201 16.6 0.800 0.840 0.800 0.840 cp_fm_cholesky_invert 11 10.9 0.819 0.822 0.819 0.822 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.595000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=517.272727, yerr=3.249921 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 602.865664E+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 0.032 30.842 30.843 qs_mol_dyn_low 1 2.0 0.003 0.004 30.623 30.632 qs_forces 11 3.9 0.002 0.002 30.561 30.562 qs_energies 11 4.9 0.002 0.002 28.822 28.824 scf_env_do_scf 11 5.9 0.000 0.001 22.772 22.772 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 20.090 20.091 velocity_verlet 10 3.0 0.002 0.002 15.147 15.150 dbcsr_multiply_generic 2286 12.5 0.092 0.094 13.796 13.891 multiply_cannon 2286 13.5 0.231 0.242 11.040 11.629 qs_scf_new_mos 108 7.5 0.001 0.001 11.592 11.619 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.591 11.618 ot_scf_mini 108 9.5 0.002 0.002 10.895 10.921 multiply_cannon_loop 2286 14.5 0.327 0.339 10.090 10.303 rebuild_ks_matrix 119 8.3 0.000 0.000 6.273 6.298 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.273 6.298 multiply_cannon_multrec 9144 15.5 1.616 1.861 5.809 6.060 ot_mini 108 10.5 0.001 0.001 5.963 5.997 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.603 5.627 qs_ot_get_derivative 108 11.5 0.001 0.001 4.614 4.638 init_scf_run 11 5.9 0.000 0.001 4.592 4.592 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.592 4.592 dbcsr_mm_accdrv_process 12550 15.8 3.161 3.810 4.091 4.175 sum_up_and_integrate 119 10.3 0.038 0.040 3.900 3.904 integrate_v_rspace 119 11.3 0.003 0.003 3.862 3.867 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.786 3.794 calculate_rho_elec 119 8.7 0.060 0.061 3.786 3.793 calculate_first_density_matrix 1 7.0 0.001 0.002 3.401 3.403 qs_ot_get_p 119 10.4 0.001 0.001 3.077 3.136 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.835 2.837 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.660 2.669 init_scf_loop 11 6.9 0.000 0.000 2.655 2.655 mp_waitall_1 115863 16.7 1.807 2.331 1.807 2.331 pw_transfer 1439 11.6 0.066 0.069 2.282 2.294 density_rs2pw 119 9.7 0.004 0.004 2.096 2.239 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.188 2.202 acc_transpose_blocks 9144 15.5 0.038 0.039 2.182 2.201 make_m2s 4572 13.5 0.034 0.035 1.891 2.083 rs_pw_transfer 974 11.9 0.008 0.008 1.944 2.081 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.066 2.069 make_images 4572 14.5 0.266 0.299 1.802 1.993 acc_transpose_blocks_kernels 9144 16.5 0.117 0.121 1.943 1.960 grid_integrate_task_list 119 12.3 1.851 1.929 1.851 1.929 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.870 1.871 prepare_preconditioner 11 7.9 0.000 0.000 1.864 1.870 make_preconditioner 11 8.9 0.000 0.000 1.864 1.870 fft3d_ps 1201 14.6 0.562 0.573 1.848 1.860 jit_kernel_transpose 5 15.6 1.826 1.843 1.826 1.843 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.777 1.789 jit_kernel_multiply 10 15.7 0.892 1.786 0.892 1.786 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.750 1.781 calculate_dm_sparse 119 9.5 0.000 0.000 1.747 1.764 potential_pw2rs 119 12.3 0.010 0.011 1.592 1.596 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.564 1.577 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.561 1.562 cp_fm_redistribute_end 50 14.0 0.772 1.517 0.773 1.518 cp_fm_diag_elpa_base 50 14.0 0.699 1.435 0.743 1.493 grid_collocate_task_list 119 9.7 1.274 1.396 1.274 1.396 ot_diis_step 108 11.5 0.012 0.013 1.332 1.333 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.209 1.241 apply_single 119 13.6 0.000 0.000 1.209 1.241 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.234 1.235 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.214 1.226 hybrid_alltoall_any 4725 16.4 0.062 0.174 0.890 1.166 make_images_data 4572 15.5 0.038 0.042 0.910 1.155 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.145 1.150 wfi_extrapolate 11 7.9 0.001 0.001 1.140 1.140 mp_alltoall_d11v 2130 13.8 0.870 0.970 0.870 0.970 cp_fm_cholesky_invert 11 10.9 0.954 0.957 0.954 0.957 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.919 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.795 0.797 multiply_cannon_sync_h2d 9144 15.5 0.716 0.794 0.716 0.794 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.382 0.793 mp_alltoall_z22v 1201 16.6 0.734 0.791 0.734 0.791 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.760 0.769 mp_allgather_i34 2286 14.5 0.282 0.759 0.282 0.759 qs_env_update_s_mstruct 11 6.9 0.001 0.002 0.688 0.737 yz_to_x 606 15.1 0.265 0.276 0.651 0.671 x_to_yz 595 16.2 0.276 0.288 0.625 0.631 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.843000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=570.727273, yerr=6.001377 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 743.596032E+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 0.018 0.037 45.606 45.606 qs_mol_dyn_low 1 2.0 0.003 0.004 45.342 45.349 qs_forces 11 3.9 0.002 0.002 45.278 45.279 qs_energies 11 4.9 0.001 0.001 43.220 43.225 scf_env_do_scf 11 5.9 0.001 0.001 35.955 35.955 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 27.431 27.432 velocity_verlet 10 3.0 0.002 0.002 24.782 24.788 dbcsr_multiply_generic 2286 12.5 0.100 0.101 19.674 19.854 qs_scf_new_mos 108 7.5 0.001 0.001 16.986 17.076 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.985 17.076 multiply_cannon 2286 13.5 0.305 0.314 15.388 16.334 ot_scf_mini 108 9.5 0.002 0.002 15.864 15.959 multiply_cannon_loop 2286 14.5 0.343 0.349 14.010 14.932 ot_mini 108 10.5 0.001 0.001 9.387 9.502 multiply_cannon_multrec 9144 15.5 3.457 4.755 8.738 8.813 init_scf_loop 11 6.9 0.000 0.000 8.494 8.495 rebuild_ks_matrix 119 8.3 0.000 0.000 7.783 7.922 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.782 7.922 prepare_preconditioner 11 7.9 0.000 0.000 7.469 7.483 make_preconditioner 11 8.9 0.000 0.000 7.469 7.483 qs_ot_get_derivative 108 11.5 0.001 0.001 7.285 7.381 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.977 7.355 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.042 7.169 dbcsr_mm_accdrv_process 12550 15.8 4.120 5.699 5.158 6.482 init_scf_run 11 5.9 0.000 0.001 5.176 5.176 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.175 5.175 cp_fm_upper_to_full 72 14.2 3.272 4.701 3.272 4.701 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.513 4.522 calculate_rho_elec 119 8.7 0.118 0.121 4.513 4.522 sum_up_and_integrate 119 10.3 0.065 0.066 4.348 4.354 integrate_v_rspace 119 11.3 0.003 0.003 4.283 4.290 mp_waitall_1 94719 16.7 2.875 3.855 2.875 3.855 qs_ot_get_p 119 10.4 0.001 0.001 3.612 3.752 calculate_first_density_matrix 1 7.0 0.001 0.001 3.610 3.610 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.287 3.288 pw_transfer 1439 11.6 0.068 0.069 3.100 3.111 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.627 3.041 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.001 3.011 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.938 2.967 dbcsr_complete_redistribute 329 12.2 0.286 0.288 2.091 2.935 make_m2s 4572 13.5 0.038 0.038 2.583 2.855 make_images 4572 14.5 0.351 0.385 2.461 2.732 fft3d_ps 1201 14.6 0.600 0.611 2.624 2.633 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.746 2.575 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.253 2.545 apply_single 119 13.6 0.000 0.000 2.252 2.545 fft_wrap_pw1pw2_140 487 13.2 0.095 0.097 2.483 2.496 density_rs2pw 119 9.7 0.004 0.004 2.404 2.429 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.532 2.415 mp_alltoall_i22 627 13.8 1.537 2.396 1.537 2.396 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.486 2.307 calculate_dm_sparse 119 9.5 0.000 0.000 2.284 2.304 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.236 2.238 acc_transpose_blocks 9144 15.5 0.038 0.039 2.197 2.210 ot_diis_step 108 11.5 0.014 0.014 2.074 2.075 grid_integrate_task_list 119 12.3 2.064 2.073 2.064 2.073 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 1.948 1.960 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.934 1.934 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.845 1.893 jit_kernel_multiply 10 15.5 1.011 1.892 1.011 1.892 cp_fm_cholesky_invert 11 10.9 1.855 1.859 1.855 1.859 jit_kernel_transpose 5 15.6 1.830 1.842 1.830 1.842 mp_sum_l 7207 12.9 1.164 1.838 1.164 1.838 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.799 1.800 potential_pw2rs 119 12.3 0.014 0.014 1.694 1.703 rs_pw_transfer 974 11.9 0.009 0.010 1.673 1.697 make_images_data 4572 15.5 0.042 0.045 1.270 1.648 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.594 1.594 cp_fm_diag_elpa_base 50 14.0 1.444 1.500 1.592 1.592 hybrid_alltoall_any 4725 16.4 0.087 0.145 1.270 1.583 grid_collocate_task_list 119 9.7 1.492 1.522 1.492 1.522 wfi_extrapolate 11 7.9 0.001 0.001 1.497 1.497 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.316 1.328 mp_alltoall_d11v 2130 13.8 1.211 1.229 1.211 1.229 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.198 1.223 qs_env_update_s_mstruct 11 6.9 0.002 0.002 1.137 1.157 mp_alltoall_z22v 1201 16.6 1.075 1.125 1.075 1.125 yz_to_x 606 15.1 0.459 0.470 1.080 1.096 multiply_cannon_sync_h2d 9144 15.5 1.044 1.053 1.044 1.053 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.990 1.036 qs_create_task_list 11 7.9 0.019 0.020 0.960 0.973 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.954 0.968 generate_qs_task_list 11 8.9 0.370 0.388 0.941 0.953 x_to_yz 595 16.2 0.478 0.490 0.932 0.936 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=45.606000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=697.545455, yerr=16.411546 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 503.861248E+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 0.027 0.052 86.284 86.286 qs_mol_dyn_low 1 2.0 0.004 0.010 85.843 85.855 qs_forces 11 3.9 0.003 0.009 85.763 85.765 qs_energies 11 4.9 0.002 0.002 82.866 82.881 scf_env_do_scf 11 5.9 0.000 0.001 72.169 72.172 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 66.444 66.444 dbcsr_multiply_generic 2055 12.4 0.107 0.110 54.028 54.288 qs_scf_new_mos 99 7.5 0.000 0.001 48.635 48.781 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.635 48.780 ot_scf_mini 99 9.5 0.002 0.002 46.148 46.309 multiply_cannon 2055 13.4 0.175 0.180 44.397 45.345 multiply_cannon_loop 2055 14.4 1.535 1.578 43.373 44.367 velocity_verlet 10 3.0 0.001 0.002 42.903 42.903 ot_mini 99 10.5 0.001 0.001 27.676 27.831 qs_ot_get_derivative 99 11.5 0.001 0.001 20.869 21.017 multiply_cannon_multrec 49320 15.4 12.143 12.811 17.420 18.044 rebuild_ks_matrix 110 8.3 0.000 0.001 14.789 14.954 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.788 14.953 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.981 13.146 mp_waitall_1 220248 16.4 11.656 12.484 11.656 12.484 multiply_cannon_sync_h2d 49320 15.4 9.939 10.664 9.939 10.664 qs_ot_get_p 110 10.4 0.001 0.001 9.848 9.989 init_scf_run 11 5.9 0.000 0.001 8.535 8.535 scf_env_initial_rho_setup 11 6.9 0.000 0.001 8.534 8.535 multiply_cannon_metrocomm3 49320 15.4 0.078 0.082 6.967 8.268 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.146 7.673 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.185 7.631 apply_single 110 13.6 0.000 0.001 7.185 7.631 sum_up_and_integrate 110 10.3 0.036 0.043 7.246 7.263 integrate_v_rspace 110 11.3 0.003 0.003 7.209 7.235 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.850 6.978 calculate_rho_elec 110 8.6 0.021 0.026 6.849 6.977 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.595 6.671 ot_diis_step 99 11.5 0.006 0.006 6.545 6.545 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.441 6.465 mp_sum_l 6514 12.8 5.147 6.144 5.147 6.144 init_scf_loop 11 6.9 0.000 0.000 5.697 5.697 dbcsr_mm_accdrv_process 87628 16.1 2.072 2.164 5.156 5.449 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.429 5.430 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.156 5.161 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.907 4.908 cp_fm_redistribute_end 48 14.0 4.280 4.879 4.285 4.880 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.784 4.864 cp_fm_diag_elpa_base 48 14.0 0.589 4.735 0.592 4.757 rs_pw_transfer 902 11.9 0.011 0.013 3.873 4.547 calculate_first_density_matrix 1 7.0 0.001 0.002 4.354 4.358 density_rs2pw 110 9.6 0.005 0.005 3.599 4.176 make_m2s 4110 13.4 0.060 0.065 4.032 4.154 wfi_extrapolate 11 7.9 0.001 0.001 4.094 4.094 calculate_dm_sparse 110 9.5 0.000 0.001 3.997 4.088 make_images 4110 14.4 0.176 0.190 3.937 4.064 acc_transpose_blocks 49320 15.4 0.228 0.244 3.725 3.805 multiply_cannon_metrocomm1 49320 15.4 0.058 0.063 2.720 3.794 prepare_preconditioner 11 7.9 0.000 0.000 3.524 3.554 make_preconditioner 11 8.9 0.000 0.000 3.524 3.554 pw_transfer 1331 11.6 0.054 0.063 3.424 3.500 grid_integrate_task_list 110 12.3 3.218 3.429 3.218 3.429 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.337 3.416 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.342 3.413 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.302 3.347 fft3d_ps 1111 14.6 0.777 0.882 2.864 2.934 fft_wrap_pw1pw2_140 451 13.1 0.171 0.192 2.796 2.880 jit_kernel_multiply 13 15.9 2.802 2.856 2.802 2.856 potential_pw2rs 110 12.3 0.006 0.007 2.781 2.813 acc_transpose_blocks_kernels 49320 16.4 0.450 0.475 2.612 2.645 mp_alltoall_d11v 2046 13.8 2.193 2.590 2.193 2.590 mp_waitany 14300 13.8 1.889 2.510 1.889 2.510 grid_collocate_task_list 110 9.6 2.083 2.314 2.083 2.314 jit_kernel_transpose 5 15.6 2.162 2.182 2.162 2.182 mp_sum_d 3881 11.9 1.490 2.022 1.490 2.022 make_images_data 4110 15.4 0.042 0.045 1.831 1.976 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.948 1.966 cp_fm_cholesky_invert 11 10.9 1.946 1.950 1.946 1.950 hybrid_alltoall_any 4261 16.3 0.082 0.482 1.584 1.859 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.793 1.821 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=86.286000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.181818, yerr=2.979267 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 585.318400E+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 0.013 0.031 73.060 73.060 qs_mol_dyn_low 1 2.0 0.003 0.004 72.802 72.812 qs_forces 11 3.9 0.002 0.002 72.725 72.726 qs_energies 11 4.9 0.003 0.004 69.358 69.362 scf_env_do_scf 11 5.9 0.000 0.001 59.365 59.369 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 51.476 51.477 dbcsr_multiply_generic 2055 12.4 0.114 0.118 40.067 40.336 velocity_verlet 10 3.0 0.001 0.002 37.469 37.470 qs_scf_new_mos 99 7.5 0.001 0.001 34.355 34.517 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.354 34.517 multiply_cannon 2055 13.4 0.221 0.245 32.896 34.337 ot_scf_mini 99 9.5 0.003 0.003 32.638 32.788 multiply_cannon_loop 2055 14.4 0.922 0.948 31.507 32.535 ot_mini 99 10.5 0.001 0.001 19.037 19.202 multiply_cannon_multrec 24660 15.4 7.549 8.706 13.859 15.206 rebuild_ks_matrix 110 8.3 0.000 0.000 14.212 14.350 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 14.211 14.350 qs_ot_get_derivative 99 11.5 0.001 0.001 13.088 13.237 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.539 12.661 mp_waitall_1 176588 16.5 8.351 11.258 8.351 11.258 multiply_cannon_metrocomm3 24660 15.4 0.070 0.073 5.567 8.847 multiply_cannon_sync_h2d 24660 15.4 7.005 8.481 7.005 8.481 init_scf_loop 11 6.9 0.000 0.000 7.844 7.845 init_scf_run 11 5.9 0.000 0.001 7.594 7.594 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.593 7.594 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.596 7.316 apply_single 110 13.6 0.000 0.001 6.596 7.315 qs_ot_get_p 110 10.4 0.001 0.001 6.839 7.010 sum_up_and_integrate 110 10.3 0.051 0.057 6.946 6.958 integrate_v_rspace 110 11.3 0.002 0.003 6.894 6.908 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.586 6.596 calculate_rho_elec 110 8.6 0.039 0.047 6.585 6.595 dbcsr_mm_accdrv_process 52282 16.1 4.585 5.430 6.153 6.431 ot_diis_step 99 11.5 0.010 0.010 5.880 5.881 prepare_preconditioner 11 7.9 0.000 0.000 5.803 5.829 make_preconditioner 11 8.9 0.000 0.000 5.803 5.828 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.902 5.624 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.370 5.533 make_m2s 4110 13.4 0.057 0.060 4.523 4.997 make_images 4110 14.4 0.395 0.439 4.413 4.884 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.796 4.822 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.688 4.690 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.308 4.345 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.330 4.331 pw_transfer 1331 11.6 0.066 0.074 4.040 4.187 density_rs2pw 110 9.6 0.004 0.005 3.660 4.155 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.933 4.084 calculate_first_density_matrix 1 7.0 0.001 0.001 3.928 3.931 rs_pw_transfer 902 11.9 0.012 0.013 3.262 3.790 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.739 3.741 cp_fm_redistribute_end 48 14.0 2.792 3.707 2.794 3.707 cp_fm_diag_elpa_base 48 14.0 0.875 3.563 0.909 3.649 wfi_extrapolate 11 7.9 0.001 0.001 3.573 3.573 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.449 3.537 fft3d_ps 1111 14.6 1.102 1.316 3.283 3.424 fft_wrap_pw1pw2_140 451 13.1 0.203 0.219 3.252 3.404 grid_integrate_task_list 110 12.3 3.154 3.342 3.154 3.342 calculate_dm_sparse 110 9.5 0.001 0.001 3.039 3.066 make_images_data 4110 15.4 0.046 0.050 2.525 3.001 acc_transpose_blocks 24660 15.4 0.112 0.115 2.872 2.918 mp_sum_l 6514 12.8 2.020 2.848 2.020 2.848 hybrid_alltoall_any 4261 16.3 0.102 0.445 2.172 2.822 cp_fm_cholesky_invert 11 10.9 2.734 2.741 2.734 2.741 potential_pw2rs 110 12.3 0.008 0.009 2.689 2.707 jit_kernel_multiply 13 16.3 1.219 2.598 1.219 2.598 grid_collocate_task_list 110 9.6 2.051 2.477 2.051 2.477 acc_transpose_blocks_kernels 24660 16.4 0.310 0.316 2.257 2.287 mp_alltoall_d11v 2046 13.8 1.892 2.256 1.892 2.256 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.048 2.072 jit_kernel_transpose 5 15.6 1.947 1.972 1.947 1.972 multiply_cannon_metrocomm4 22605 15.4 0.075 0.080 0.791 1.943 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.876 1.876 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.838 1.854 mp_waitany 10164 13.8 1.304 1.817 1.304 1.817 mp_allgather_i34 2055 14.4 0.720 1.806 0.720 1.806 mp_irecv_dv 57340 16.2 0.666 1.730 0.666 1.730 mp_alltoall_z22v 1111 16.6 1.488 1.599 1.488 1.599 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.564 1.575 dbcsr_complete_redistribute 325 12.2 0.246 0.334 1.237 1.504 rs_pw_transfer_RS2PW_140 121 11.5 0.206 0.216 0.977 1.483 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.369 1.476 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=73.060000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.636364, yerr=6.399897 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 660.045824E+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 0.023 0.062 63.467 63.469 qs_mol_dyn_low 1 2.0 0.003 0.005 62.993 63.004 qs_forces 11 3.9 0.002 0.006 62.916 62.919 qs_energies 11 4.9 0.003 0.012 59.662 59.667 scf_env_do_scf 11 5.9 0.001 0.002 50.241 50.241 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 41.284 41.287 velocity_verlet 10 3.0 0.001 0.002 33.669 33.671 dbcsr_multiply_generic 2055 12.4 0.108 0.112 30.419 30.767 qs_scf_new_mos 99 7.5 0.001 0.001 25.682 25.766 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.682 25.765 multiply_cannon 2055 13.4 0.211 0.222 23.796 24.968 ot_scf_mini 99 9.5 0.003 0.004 24.440 24.550 multiply_cannon_loop 2055 14.4 0.612 0.626 22.510 23.782 ot_mini 99 10.5 0.001 0.001 14.050 14.159 rebuild_ks_matrix 110 8.3 0.000 0.000 12.818 12.983 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.018 12.817 12.983 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.287 11.435 mp_waitall_1 139946 16.5 7.495 10.648 7.495 10.648 multiply_cannon_multrec 16440 15.4 3.829 4.775 9.657 10.546 qs_ot_get_derivative 99 11.5 0.001 0.001 9.454 9.564 init_scf_loop 11 6.9 0.001 0.009 8.918 8.924 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.613 7.594 prepare_preconditioner 11 7.9 0.000 0.000 7.060 7.076 make_preconditioner 11 8.9 0.000 0.002 7.060 7.076 init_scf_run 11 5.9 0.001 0.003 6.964 6.965 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.964 6.965 sum_up_and_integrate 110 10.3 0.060 0.061 6.881 6.901 integrate_v_rspace 110 11.3 0.003 0.004 6.822 6.842 make_full_inverse_cholesky 11 9.9 0.001 0.002 6.399 6.750 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.237 6.252 calculate_rho_elec 110 8.6 0.058 0.058 6.236 6.252 dbcsr_mm_accdrv_process 34862 16.1 4.685 5.458 5.684 5.853 qs_ot_get_p 110 10.4 0.001 0.002 5.511 5.635 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.067 5.506 apply_single 110 13.6 0.000 0.000 5.067 5.505 make_m2s 4110 13.4 0.050 0.052 4.345 4.713 make_images 4110 14.4 0.389 0.504 4.230 4.597 ot_diis_step 99 11.5 0.011 0.011 4.544 4.545 density_rs2pw 110 9.6 0.004 0.005 3.316 4.516 multiply_cannon_sync_h2d 16440 15.4 3.652 4.237 3.652 4.237 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 4.149 4.151 rs_pw_transfer 902 11.9 0.010 0.011 2.927 4.126 pw_transfer 1331 11.6 0.065 0.072 3.910 3.925 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.242 3.876 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.799 3.850 calculate_first_density_matrix 1 7.0 0.001 0.003 3.834 3.835 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.803 3.821 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.745 3.751 grid_integrate_task_list 110 12.3 3.155 3.424 3.155 3.424 cp_dbcsr_syevd 48 12.0 0.004 0.008 3.400 3.400 fft_wrap_pw1pw2_140 451 13.1 0.213 0.216 3.228 3.247 fft3d_ps 1111 14.6 1.088 1.097 3.102 3.117 wfi_extrapolate 11 7.9 0.001 0.002 3.041 3.041 make_images_data 4110 15.4 0.043 0.048 2.517 3.036 hybrid_alltoall_any 4261 16.3 0.105 0.372 2.221 2.919 cp_fm_cholesky_invert 11 10.9 2.879 2.885 2.879 2.885 cp_fm_diag_elpa 48 13.0 0.000 0.002 2.831 2.833 cp_fm_redistribute_end 48 14.0 1.763 2.797 1.765 2.798 cp_fm_diag_elpa_base 48 14.0 0.967 2.644 1.026 2.747 calculate_dm_sparse 110 9.5 0.001 0.001 2.584 2.612 acc_transpose_blocks 16440 15.4 0.072 0.073 2.558 2.568 potential_pw2rs 110 12.3 0.010 0.010 2.530 2.550 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.452 2.513 mp_waitany 17072 13.8 1.222 2.483 1.222 2.483 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.901 2.462 grid_collocate_task_list 110 9.6 2.088 2.462 2.088 2.462 mp_irecv_dv 48980 15.7 0.831 2.340 0.831 2.340 mp_sum_l 6514 12.8 1.670 2.318 1.670 2.318 mp_alltoall_d11v 2046 13.8 1.922 2.285 1.922 2.285 acc_transpose_blocks_kernels 16440 16.4 0.208 0.214 2.138 2.145 rs_pw_transfer_RS2PW_140 121 11.5 0.178 0.184 0.911 2.107 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.994 1.997 dbcsr_complete_redistribute 325 12.2 0.327 0.360 1.506 1.961 jit_kernel_transpose 5 15.6 1.930 1.936 1.930 1.936 cp_fm_upper_to_full 70 13.6 1.384 1.847 1.384 1.847 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.735 1.752 mp_allgather_i34 2055 14.4 0.613 1.745 0.613 1.745 cp_fm_cholesky_decompose 22 10.9 1.664 1.684 1.664 1.684 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.529 1.540 jit_kernel_multiply 8 16.5 0.613 1.526 0.613 1.526 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.363 1.462 rs_gather_matrices 110 12.3 0.232 0.260 1.056 1.438 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.984 1.435 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.316 1.324 mp_alltoall_z22v 1111 16.6 1.248 1.283 1.248 1.283 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.469000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.272727, yerr=9.235218 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 731.557888E+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 0.022 0.031 67.598 67.599 qs_mol_dyn_low 1 2.0 0.003 0.004 67.299 67.309 qs_forces 11 3.9 0.003 0.007 67.170 67.171 qs_energies 11 4.9 0.004 0.013 63.738 63.742 scf_env_do_scf 11 5.9 0.001 0.001 53.902 53.905 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 41.932 41.934 velocity_verlet 10 3.0 0.002 0.002 37.382 37.387 dbcsr_multiply_generic 2055 12.4 0.114 0.116 31.016 31.255 qs_scf_new_mos 99 7.5 0.001 0.001 26.640 26.728 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.639 26.728 ot_scf_mini 99 9.5 0.003 0.004 25.003 25.090 multiply_cannon 2055 13.4 0.243 0.260 24.069 25.063 multiply_cannon_loop 2055 14.4 0.878 0.904 22.735 23.361 ot_mini 99 10.5 0.001 0.001 14.315 14.421 multiply_cannon_multrec 24660 15.4 4.257 6.830 12.972 14.124 rebuild_ks_matrix 110 8.3 0.000 0.000 12.366 12.457 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.366 12.456 init_scf_loop 11 6.9 0.001 0.006 11.929 11.933 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.927 11.010 qs_ot_get_derivative 99 11.5 0.001 0.001 10.155 10.242 prepare_preconditioner 11 7.9 0.000 0.001 10.130 10.144 make_preconditioner 11 8.9 0.000 0.002 10.130 10.144 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.354 9.827 dbcsr_mm_accdrv_process 52304 16.0 6.937 8.344 8.571 9.529 init_scf_run 11 5.9 0.000 0.001 7.062 7.062 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.061 7.062 sum_up_and_integrate 110 10.3 0.066 0.069 6.658 6.673 integrate_v_rspace 110 11.3 0.003 0.004 6.592 6.604 mp_waitall_1 121746 16.5 4.395 6.553 4.395 6.553 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.268 6.277 calculate_rho_elec 110 8.6 0.077 0.081 6.268 6.276 qs_ot_get_p 110 10.4 0.001 0.001 5.728 5.848 make_m2s 4110 13.4 0.060 0.062 5.372 5.637 make_images 4110 14.4 0.572 0.695 5.231 5.491 cp_fm_upper_to_full 70 13.8 3.270 4.664 3.270 4.664 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 4.158 4.160 ot_diis_step 99 11.5 0.011 0.011 4.120 4.120 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.026 4.100 apply_single 110 13.6 0.000 0.000 4.025 4.100 pw_transfer 1331 11.6 0.065 0.074 3.979 4.015 calculate_first_density_matrix 1 7.0 0.006 0.011 3.958 3.960 qs_ot_p2m_diag 48 11.0 0.054 0.062 3.934 3.948 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.874 3.913 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.801 3.836 dbcsr_complete_redistribute 325 12.2 0.419 0.464 2.685 3.822 density_rs2pw 110 9.6 0.004 0.004 3.189 3.694 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.459 3.459 grid_integrate_task_list 110 12.3 3.254 3.421 3.254 3.421 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.345 3.392 fft_wrap_pw1pw2_140 451 13.1 0.202 0.210 3.317 3.357 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.509 3.341 multiply_cannon_sync_h2d 24660 15.4 3.176 3.328 3.176 3.328 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.132 3.260 fft3d_ps 1111 14.6 1.097 1.133 3.166 3.193 calculate_dm_sparse 110 9.5 0.001 0.001 3.056 3.081 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.002 3.050 rs_pw_transfer 902 11.9 0.010 0.011 2.506 3.023 wfi_extrapolate 11 7.9 0.001 0.001 3.008 3.009 make_images_data 4110 15.4 0.046 0.050 2.689 2.968 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.910 2.912 hybrid_alltoall_any 4261 16.3 0.119 0.460 2.377 2.906 cp_fm_cholesky_invert 11 10.9 2.887 2.897 2.887 2.897 cp_fm_redistribute_end 48 14.0 1.454 2.879 1.455 2.881 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.767 2.873 acc_transpose_blocks 24660 15.4 0.107 0.109 2.835 2.863 cp_fm_diag_elpa_base 48 14.0 1.344 2.745 1.422 2.853 mp_alltoall_i22 605 13.7 1.649 2.839 1.649 2.839 jit_kernel_multiply 12 15.8 1.306 2.541 1.306 2.541 grid_collocate_task_list 110 9.6 2.182 2.423 2.182 2.423 potential_pw2rs 110 12.3 0.013 0.013 2.309 2.323 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 2.274 2.275 acc_transpose_blocks_kernels 24660 16.4 0.303 0.308 2.194 2.215 mp_alltoall_d11v 2046 13.8 1.767 2.050 1.767 2.050 jit_kernel_transpose 5 15.6 1.891 1.908 1.891 1.908 cp_fm_cholesky_decompose 22 10.9 1.721 1.766 1.721 1.766 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.713 1.746 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.599 1.701 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.685 1.696 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.658 1.673 mp_waitany 13376 13.8 1.094 1.609 1.094 1.609 multiply_cannon_metrocomm4 20550 15.4 0.058 0.061 0.850 1.455 mp_sum_l 6514 12.8 0.901 1.409 0.901 1.409 mp_irecv_dv 62702 16.1 0.751 1.376 0.751 1.376 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.599000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=693.909091, yerr=10.413754 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 834.052096E+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 0.026 0.051 59.454 59.463 qs_mol_dyn_low 1 2.0 0.003 0.004 59.116 59.141 qs_forces 11 3.9 0.002 0.002 59.038 59.039 qs_energies 11 4.9 0.001 0.002 55.269 55.273 scf_env_do_scf 11 5.9 0.000 0.001 45.578 45.579 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 37.208 37.209 velocity_verlet 10 3.0 0.002 0.002 32.581 32.585 dbcsr_multiply_generic 2055 12.4 0.105 0.106 24.571 24.719 qs_scf_new_mos 99 7.5 0.001 0.001 21.414 21.489 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.414 21.488 ot_scf_mini 99 9.5 0.002 0.002 20.147 20.208 multiply_cannon 2055 13.4 0.247 0.265 18.801 20.120 multiply_cannon_loop 2055 14.4 0.318 0.331 17.404 17.851 rebuild_ks_matrix 110 8.3 0.000 0.000 12.495 12.546 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.495 12.545 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.102 11.151 ot_mini 99 10.5 0.001 0.001 10.676 10.735 multiply_cannon_multrec 8220 15.4 3.178 4.814 7.452 8.671 init_scf_loop 11 6.9 0.000 0.000 8.321 8.323 mp_waitall_1 103326 16.6 6.394 8.023 6.394 8.023 sum_up_and_integrate 110 10.3 0.079 0.081 6.884 6.895 integrate_v_rspace 110 11.3 0.003 0.003 6.805 6.817 qs_ot_get_derivative 99 11.5 0.001 0.001 6.752 6.811 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.644 6.662 calculate_rho_elec 110 8.6 0.115 0.116 6.643 6.662 prepare_preconditioner 11 7.9 0.000 0.000 6.542 6.550 make_preconditioner 11 8.9 0.000 0.000 6.542 6.550 init_scf_run 11 5.9 0.000 0.001 6.512 6.512 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.512 6.512 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.113 6.194 qs_ot_get_p 110 10.4 0.001 0.001 5.515 5.600 dbcsr_mm_accdrv_process 17442 15.9 2.845 3.633 4.144 5.060 make_m2s 4110 13.4 0.038 0.040 4.358 4.600 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.136 4.557 make_images 4110 14.4 0.632 0.691 4.228 4.470 pw_transfer 1331 11.6 0.065 0.070 4.425 4.436 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.318 4.334 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.995 4.000 ot_diis_step 99 11.5 0.012 0.012 3.897 3.898 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.794 3.850 apply_single 110 13.6 0.000 0.000 3.794 3.850 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.833 3.835 density_rs2pw 110 9.6 0.004 0.004 3.345 3.723 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.688 3.689 fft_wrap_pw1pw2_140 451 13.1 0.215 0.219 3.665 3.686 calculate_first_density_matrix 1 7.0 0.001 0.001 3.594 3.595 grid_integrate_task_list 110 12.3 3.349 3.576 3.349 3.576 fft3d_ps 1111 14.6 1.147 1.194 3.563 3.571 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.429 3.453 cp_fm_cholesky_invert 11 10.9 3.199 3.203 3.199 3.203 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.138 3.139 cp_fm_redistribute_end 48 14.0 0.805 3.110 0.810 3.111 cp_fm_diag_elpa_base 48 14.0 2.115 2.899 2.292 3.065 multiply_cannon_sync_h2d 8220 15.4 2.899 3.022 2.899 3.022 make_images_data 4110 15.4 0.038 0.043 2.481 2.903 wfi_extrapolate 11 7.9 0.001 0.001 2.822 2.822 hybrid_alltoall_any 4261 16.3 0.198 0.856 2.429 2.815 rs_pw_transfer 902 11.9 0.010 0.010 2.357 2.754 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.709 2.710 calculate_dm_sparse 110 9.5 0.001 0.001 2.546 2.585 grid_collocate_task_list 110 9.6 2.279 2.538 2.279 2.538 potential_pw2rs 110 12.3 0.015 0.016 2.323 2.327 mp_alltoall_d11v 2046 13.8 1.881 2.279 1.881 2.279 acc_transpose_blocks 8220 15.4 0.035 0.036 2.123 2.135 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.032 2.060 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.772 1.996 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.939 1.970 acc_transpose_blocks_kernels 8220 16.4 0.109 0.113 1.893 1.900 cp_fm_cholesky_decompose 22 10.9 1.834 1.849 1.834 1.849 jit_kernel_transpose 5 15.6 1.784 1.787 1.784 1.787 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.729 1.734 mp_allgather_i34 2055 14.4 0.582 1.700 0.582 1.700 qs_env_update_s_mstruct 11 6.9 0.003 0.005 1.545 1.668 dbcsr_complete_redistribute 325 12.2 0.566 0.598 1.508 1.597 mp_waitany 9240 13.8 1.128 1.543 1.128 1.543 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.489 1.503 jit_kernel_multiply 8 15.8 0.992 1.403 0.992 1.403 rs_gather_matrices 110 12.3 0.324 0.363 1.040 1.400 qs_create_task_list 11 7.9 0.009 0.023 1.251 1.349 generate_qs_task_list 11 8.9 0.379 0.447 1.242 1.331 mp_alltoall_z22v 1111 16.6 1.306 1.328 1.306 1.328 multiply_cannon_metrocomm1 8220 15.4 0.020 0.021 0.897 1.313 multiply_cannon_metrocomm4 6165 15.4 0.018 0.020 0.482 1.296 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.224 1.260 mp_irecv_dv 24056 15.7 0.458 1.256 0.458 1.256 x_to_yz 550 16.1 0.573 0.585 1.212 1.223 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.166 0.815 1.214 yz_to_x 561 15.1 0.525 0.538 1.192 1.202 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=59.463000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=784.272727, yerr=11.802087 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 1.350492E+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 0.103 0.131 92.685 92.693 qs_mol_dyn_low 1 2.0 0.003 0.004 92.198 92.208 qs_forces 11 3.9 0.002 0.003 92.128 92.129 qs_energies 11 4.9 0.002 0.002 87.894 87.897 scf_env_do_scf 11 5.9 0.001 0.001 76.331 76.331 velocity_verlet 10 3.0 0.002 0.002 57.865 57.871 scf_env_do_scf_inner_loop 99 6.5 0.003 0.011 46.660 46.662 dbcsr_multiply_generic 2055 12.4 0.119 0.120 30.662 30.807 init_scf_loop 11 6.9 0.000 0.000 29.597 29.601 qs_scf_new_mos 99 7.5 0.001 0.001 27.533 27.618 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.532 27.617 prepare_preconditioner 11 7.9 0.000 0.000 27.441 27.454 make_preconditioner 11 8.9 0.000 0.000 27.441 27.454 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.453 26.891 ot_scf_mini 99 9.5 0.002 0.002 25.763 25.829 multiply_cannon 2055 13.4 0.352 0.378 23.280 24.097 multiply_cannon_loop 2055 14.4 0.341 0.344 21.390 21.817 cp_fm_upper_to_full 70 14.2 13.152 18.841 13.152 18.841 rebuild_ks_matrix 110 8.3 0.000 0.001 14.735 14.820 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.734 14.820 ot_mini 99 10.5 0.001 0.001 14.250 14.319 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.367 13.445 dbcsr_complete_redistribute 325 12.2 1.016 1.065 7.607 11.016 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.550 9.949 multiply_cannon_multrec 8220 15.4 4.513 4.692 9.780 9.873 qs_ot_get_derivative 99 11.5 0.001 0.001 9.503 9.570 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.973 9.327 mp_waitall_1 84994 16.7 8.002 9.249 8.002 9.249 mp_alltoall_i22 605 13.7 5.590 9.019 5.590 9.019 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.117 8.156 calculate_rho_elec 110 8.6 0.227 0.228 8.116 8.155 sum_up_and_integrate 110 10.3 0.150 0.151 7.694 7.708 integrate_v_rspace 110 11.3 0.004 0.004 7.544 7.558 init_scf_run 11 5.9 0.000 0.001 7.179 7.179 scf_env_initial_rho_setup 11 6.9 0.000 0.000 7.179 7.179 qs_ot_get_p 110 10.4 0.001 0.001 6.235 6.319 make_m2s 4110 13.4 0.043 0.044 5.469 6.039 make_images 4110 14.4 0.875 0.930 5.280 5.852 cp_fm_cholesky_invert 11 10.9 5.704 5.709 5.704 5.709 pw_transfer 1331 11.6 0.075 0.075 5.614 5.618 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.497 5.502 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.826 5.375 apply_single 110 13.6 0.000 0.000 4.826 5.375 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 5.016 5.362 dbcsr_mm_accdrv_process 11614 15.7 3.285 3.616 5.123 5.347 fft_wrap_pw1pw2_140 451 13.1 0.221 0.223 4.748 4.753 ot_diis_step 99 11.5 0.016 0.016 4.720 4.721 fft3d_ps 1111 14.6 1.297 1.301 4.693 4.701 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.531 4.537 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 4.149 4.155 density_rs2pw 110 9.6 0.004 0.004 4.090 4.111 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.073 4.073 multiply_cannon_sync_h2d 8220 15.4 3.954 3.961 3.954 3.961 grid_integrate_task_list 110 12.3 3.693 3.808 3.693 3.808 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.781 3.782 hybrid_alltoall_any 4261 16.3 0.256 0.550 3.002 3.739 make_images_data 4110 15.4 0.041 0.044 3.030 3.723 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.632 3.660 calculate_first_density_matrix 1 7.0 0.001 0.001 3.648 3.648 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.973 3.438 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.420 3.421 cp_fm_diag_elpa_base 48 14.0 2.864 3.075 3.418 3.419 wfi_extrapolate 11 7.9 0.001 0.001 3.405 3.405 calculate_dm_sparse 110 9.5 0.001 0.001 3.192 3.211 grid_collocate_task_list 110 9.6 2.654 2.683 2.654 2.683 potential_pw2rs 110 12.3 0.021 0.022 2.654 2.664 rs_pw_transfer 902 11.9 0.010 0.011 2.545 2.570 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.527 2.565 qs_env_update_s_mstruct 11 6.9 0.027 0.036 2.312 2.378 mp_alltoall_d11v 2046 13.8 2.191 2.272 2.191 2.272 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.084 2.204 acc_transpose_blocks 8220 15.4 0.034 0.035 2.119 2.143 cp_fm_cholesky_decompose 22 10.9 2.087 2.106 2.087 2.106 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.970 1.978 qs_create_task_list 11 7.9 0.001 0.001 1.900 1.946 generate_qs_task_list 11 8.9 0.735 0.788 1.899 1.946 acc_transpose_blocks_kernels 8220 16.4 0.109 0.111 1.885 1.911 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=92.693000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1214.727273, yerr=59.775475 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 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 628.256768E+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 3175955098256 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 0.035 0.105 208.051 208.055 qs_mol_dyn_low 1 2.0 0.004 0.009 207.474 207.490 qs_forces 11 3.9 0.004 0.010 207.362 207.365 qs_energies 11 4.9 0.002 0.005 201.657 201.673 scf_env_do_scf 11 5.9 0.001 0.001 183.282 183.287 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 162.097 162.099 dbcsr_multiply_generic 2507 12.6 0.177 0.181 125.649 126.391 velocity_verlet 10 3.0 0.001 0.002 124.220 124.221 qs_scf_new_mos 117 7.6 0.001 0.001 121.512 121.835 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.511 121.834 ot_scf_mini 117 9.6 0.003 0.003 114.974 115.247 multiply_cannon 2507 13.6 0.237 0.244 102.491 104.215 multiply_cannon_loop 2507 14.6 2.098 2.163 100.282 101.824 ot_mini 117 10.6 0.001 0.001 65.538 65.879 multiply_cannon_multrec 60168 15.6 33.084 35.301 41.758 43.860 qs_ot_get_derivative 117 11.6 0.001 0.002 40.724 41.016 rebuild_ks_matrix 128 8.3 0.001 0.001 34.068 34.393 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.018 34.068 34.392 mp_waitall_1 267128 16.5 28.182 31.675 28.182 31.675 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.618 30.904 multiply_cannon_sync_h2d 60168 15.6 27.366 29.576 27.366 29.576 qs_ot_get_p 128 10.4 0.001 0.001 27.635 28.032 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.276 25.014 apply_single 128 13.6 0.001 0.001 24.276 25.014 ot_diis_step 117 11.6 0.008 0.008 24.479 24.480 init_scf_loop 11 6.9 0.000 0.001 21.103 21.104 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.877 20.944 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.914 19.155 cp_dbcsr_syevd 83 12.4 0.004 0.005 18.271 18.272 multiply_cannon_metrocomm3 60168 15.6 0.114 0.122 15.462 18.143 prepare_preconditioner 11 7.9 0.000 0.000 16.407 16.441 make_preconditioner 11 8.9 0.000 0.000 16.407 16.441 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.638 15.836 cp_fm_diag_elpa 83 13.4 0.000 0.000 15.354 15.360 cp_fm_redistribute_end 83 14.4 12.134 15.276 12.148 15.281 cp_fm_diag_elpa_base 83 14.4 3.086 14.912 3.120 15.028 sum_up_and_integrate 128 10.3 0.089 0.106 14.517 14.530 integrate_v_rspace 128 11.3 0.003 0.004 14.427 14.443 make_m2s 5014 13.6 0.104 0.112 13.908 14.294 init_scf_run 11 5.9 0.000 0.001 14.201 14.201 scf_env_initial_rho_setup 11 6.9 0.000 0.001 14.201 14.201 make_images 5014 14.6 0.398 0.417 13.726 14.123 qs_rho_update_rho_low 128 7.7 0.001 0.002 14.017 14.121 calculate_rho_elec 128 8.7 0.045 0.064 14.016 14.120 density_rs2pw 128 9.7 0.007 0.007 7.365 10.602 cp_fm_cholesky_invert 11 10.9 9.444 9.452 9.444 9.452 mp_sum_l 7870 13.0 8.049 9.387 8.049 9.387 rs_pw_transfer 1046 11.9 0.016 0.018 5.943 9.174 wfi_extrapolate 11 7.9 0.001 0.001 9.160 9.160 dbcsr_mm_accdrv_process 124484 16.2 3.308 3.466 8.236 8.773 calculate_dm_sparse 128 9.5 0.001 0.001 8.610 8.746 pw_transfer 1547 11.6 0.075 0.098 8.198 8.442 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 7.993 8.246 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.004 8.189 8.198 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.000 8.161 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.805 7.911 multiply_cannon_metrocomm1 60168 15.6 0.086 0.091 6.049 7.770 make_images_data 5014 15.6 0.065 0.071 6.832 7.763 grid_integrate_task_list 128 12.3 7.062 7.507 7.062 7.507 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 7.326 7.451 hybrid_alltoall_any 5200 16.5 0.290 2.257 5.981 7.242 fft_wrap_pw1pw2_140 523 13.2 0.445 0.515 6.918 7.208 fft3d_ps 1291 14.7 2.103 2.887 6.779 6.970 mp_waitany 16020 13.9 2.720 6.042 2.720 6.042 grid_collocate_task_list 128 9.7 4.556 5.821 4.556 5.821 mp_alltoall_d11v 2415 14.1 4.272 5.505 4.272 5.505 rs_pw_transfer_RS2PW_140 139 11.5 0.281 0.299 2.131 5.384 potential_pw2rs 128 12.3 0.009 0.011 4.883 4.912 calculate_first_density_matrix 1 7.0 0.001 0.004 4.878 4.883 cp_fm_cholesky_decompose 22 10.9 4.772 4.786 4.772 4.786 acc_transpose_blocks 60168 15.6 0.316 0.336 4.608 4.733 mp_sum_d 4457 12.1 3.694 4.469 3.694 4.469 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=208.055000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.636364, yerr=6.865894 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 833.310720E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57341. MP_Allreduce 11227 947. MP_Sync 170 MP_Alltoall 1969 5619023. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.127 198.181 198.183 qs_mol_dyn_low 1 2.0 0.003 0.006 197.592 197.607 qs_forces 11 3.9 0.005 0.013 197.490 197.492 qs_energies 11 4.9 0.006 0.016 190.562 190.577 scf_env_do_scf 11 5.9 0.001 0.002 171.331 171.342 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 137.309 137.312 velocity_verlet 10 3.0 0.002 0.004 123.031 123.033 dbcsr_multiply_generic 2507 12.6 0.186 0.191 100.513 101.816 qs_scf_new_mos 117 7.6 0.001 0.001 96.347 96.919 qs_scf_loop_do_ot 117 8.6 0.001 0.001 96.346 96.918 ot_scf_mini 117 9.6 0.004 0.005 91.462 92.069 multiply_cannon 2507 13.6 0.480 0.528 79.824 84.784 multiply_cannon_loop 2507 14.6 1.247 1.297 76.689 79.402 ot_mini 117 10.6 0.001 0.002 50.448 50.980 mp_waitall_1 214728 16.6 25.357 38.783 25.357 38.783 multiply_cannon_multrec 30084 15.6 21.791 26.042 31.428 36.057 rebuild_ks_matrix 128 8.3 0.001 0.001 33.899 34.538 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.023 33.898 34.538 init_scf_loop 11 6.9 0.002 0.008 33.908 33.911 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.468 31.043 prepare_preconditioner 11 7.9 0.000 0.001 29.368 29.423 make_preconditioner 11 8.9 0.000 0.002 29.368 29.423 qs_ot_get_derivative 117 11.6 0.001 0.003 28.452 29.053 multiply_cannon_metrocomm3 30084 15.6 0.095 0.101 16.282 28.688 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.056 28.584 qs_ot_get_p 128 10.4 0.001 0.002 22.820 23.428 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 22.093 23.226 apply_single 128 13.6 0.001 0.001 22.093 23.226 ot_diis_step 117 11.6 0.015 0.022 21.799 21.804 multiply_cannon_sync_h2d 30084 15.6 19.051 21.355 19.051 21.355 qs_ot_p2m_diag 83 11.4 0.187 0.215 17.841 17.884 cp_fm_cholesky_invert 11 10.9 17.168 17.181 17.168 17.181 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.632 16.634 make_m2s 5014 13.6 0.090 0.097 14.554 15.902 make_images 5014 14.6 1.141 1.323 14.343 15.693 sum_up_and_integrate 128 10.3 0.115 0.137 15.253 15.283 integrate_v_rspace 128 11.3 0.004 0.005 15.137 15.173 qs_rho_update_rho_low 128 7.7 0.001 0.006 14.616 14.651 calculate_rho_elec 128 8.7 0.088 0.106 14.614 14.650 init_scf_run 11 5.9 0.001 0.003 14.277 14.279 scf_env_initial_rho_setup 11 6.9 0.000 0.002 14.277 14.279 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.458 13.469 cp_fm_redistribute_end 83 14.4 7.857 13.383 7.874 13.390 cp_fm_diag_elpa_base 83 14.4 5.265 12.832 5.493 13.222 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.288 11.718 multiply_cannon_metrocomm4 27577 15.6 0.097 0.111 3.815 10.611 make_images_data 5014 15.6 0.063 0.071 8.817 10.465 density_rs2pw 128 9.7 0.006 0.008 7.915 10.398 mp_irecv_dv 69486 16.3 3.619 10.214 3.619 10.214 dbcsr_mm_accdrv_process 62242 16.2 4.500 5.275 9.103 9.970 hybrid_alltoall_any 5200 16.5 0.341 1.523 7.479 9.618 pw_transfer 1547 11.6 0.087 0.104 9.483 9.548 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.257 9.329 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 8.626 8.634 wfi_extrapolate 11 7.9 0.003 0.018 8.619 8.620 rs_pw_transfer 1046 11.9 0.014 0.016 6.085 8.594 fft_wrap_pw1pw2_140 523 13.2 0.473 0.526 8.086 8.190 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 7.687 7.863 fft3d_ps 1291 14.7 2.758 2.911 7.622 7.668 grid_integrate_task_list 128 12.3 7.166 7.618 7.166 7.618 cp_fm_cholesky_decompose 22 10.9 7.332 7.415 7.332 7.415 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.262 7.070 calculate_dm_sparse 128 9.5 0.001 0.001 6.581 6.721 mp_sum_l 7870 13.0 4.422 6.445 4.422 6.445 grid_collocate_task_list 128 9.7 4.702 5.879 4.702 5.879 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.451 5.544 calculate_first_density_matrix 1 7.0 0.002 0.008 5.439 5.445 potential_pw2rs 128 12.3 0.015 0.017 5.414 5.435 mp_waitany 11748 13.9 2.581 5.152 2.581 5.152 mp_alltoall_d11v 2415 14.1 4.378 4.877 4.378 4.877 mp_allgather_i34 2507 14.6 1.706 4.841 1.706 4.841 rs_pw_transfer_RS2PW_140 139 11.5 0.352 0.385 2.140 4.636 acc_transpose_blocks 30084 15.6 0.142 0.146 4.166 4.320 mp_sum_d 4457 12.1 2.748 4.145 2.748 4.145 dbcsr_complete_redistribute 395 12.7 0.782 0.858 3.207 4.055 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=198.183000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=793.363636, yerr=3.699453 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 942.829568E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172905552 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4079 57277. MP_Allreduce 11236 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_comm_split 82 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.059 182.631 182.636 qs_mol_dyn_low 1 2.0 0.003 0.004 182.074 182.088 qs_forces 11 3.9 0.010 0.060 181.866 181.869 qs_energies 11 4.9 0.012 0.050 175.059 175.071 scf_env_do_scf 11 5.9 0.001 0.016 157.733 157.735 scf_env_do_scf_inner_loop 116 6.6 0.004 0.035 121.757 121.758 velocity_verlet 10 3.0 0.013 0.017 115.529 115.534 dbcsr_multiply_generic 2485 12.5 0.180 0.185 83.904 84.906 qs_scf_new_mos 116 7.6 0.001 0.001 82.716 83.079 qs_scf_loop_do_ot 116 8.6 0.001 0.001 82.715 83.078 ot_scf_mini 116 9.6 0.004 0.004 78.543 78.958 multiply_cannon 2485 13.5 0.495 0.519 63.409 67.857 multiply_cannon_loop 2485 14.5 0.842 0.872 60.164 62.827 ot_mini 116 10.6 0.001 0.001 43.259 43.679 init_scf_loop 11 6.9 0.001 0.007 35.860 35.863 mp_waitall_1 169034 16.6 26.194 35.707 26.194 35.707 rebuild_ks_matrix 127 8.3 0.001 0.001 31.373 31.848 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.023 31.373 31.847 prepare_preconditioner 11 7.9 0.000 0.000 31.714 31.768 make_preconditioner 11 8.9 0.000 0.002 31.714 31.768 make_full_inverse_cholesky 11 9.9 0.001 0.003 29.287 30.695 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.226 28.668 multiply_cannon_multrec 19880 15.5 13.202 16.444 22.056 25.182 multiply_cannon_metrocomm3 19880 15.5 0.059 0.063 15.869 25.151 qs_ot_get_derivative 116 11.6 0.001 0.002 23.384 23.803 qs_ot_get_p 127 10.4 0.001 0.001 20.547 21.089 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.992 21.070 apply_single 127 13.6 0.001 0.001 19.991 21.070 ot_diis_step 116 11.6 0.018 0.019 19.766 19.767 make_m2s 4970 13.5 0.081 0.085 15.342 16.294 qs_ot_p2m_diag 82 11.4 0.262 0.269 16.151 16.161 make_images 4970 14.5 1.163 1.259 15.110 16.062 multiply_cannon_sync_h2d 19880 15.5 14.041 15.851 14.041 15.851 sum_up_and_integrate 127 10.3 0.130 0.141 15.067 15.096 cp_dbcsr_syevd 82 12.4 0.008 0.029 15.057 15.060 integrate_v_rspace 127 11.3 0.004 0.005 14.937 14.963 cp_fm_cholesky_invert 11 10.9 14.897 14.907 14.897 14.907 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.705 14.729 calculate_rho_elec 127 8.7 0.131 0.146 14.704 14.729 init_scf_run 11 5.9 0.000 0.001 12.140 12.140 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.140 12.140 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.696 11.699 cp_fm_redistribute_end 82 14.4 4.420 11.608 4.442 11.611 cp_fm_diag_elpa_base 82 14.4 6.740 10.995 7.144 11.451 make_images_data 4970 15.5 0.060 0.067 9.599 11.048 hybrid_alltoall_any 5155 16.4 0.430 1.978 8.312 9.856 density_rs2pw 127 9.7 0.006 0.007 7.583 9.613 pw_transfer 1535 11.6 0.086 0.108 9.481 9.604 multiply_cannon_metrocomm4 17395 15.5 0.061 0.072 3.521 9.452 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.099 9.396 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.258 9.385 mp_irecv_dv 49801 16.2 3.400 9.205 3.400 9.205 dbcsr_mm_accdrv_process 41158 16.2 4.496 5.462 8.314 8.401 fft_wrap_pw1pw2_140 519 13.2 0.474 0.518 8.084 8.222 grid_integrate_task_list 127 12.3 7.222 7.755 7.222 7.755 wfi_extrapolate 11 7.9 0.002 0.011 7.696 7.696 fft3d_ps 1281 14.7 2.647 2.876 7.535 7.616 cp_fm_cholesky_decompose 22 10.9 7.553 7.582 7.553 7.582 cp_fm_upper_to_full 104 14.5 5.761 7.498 5.761 7.498 rs_pw_transfer 1038 11.9 0.014 0.016 5.465 7.469 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.861 6.868 dbcsr_complete_redistribute 393 12.7 1.171 1.200 4.801 6.595 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.906 6.043 calculate_dm_sparse 127 9.5 0.001 0.001 5.860 5.957 grid_collocate_task_list 127 9.7 4.861 5.732 4.861 5.732 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.751 5.438 mp_alltoall_d11v 2401 14.1 4.649 5.391 4.649 5.391 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.534 5.326 potential_pw2rs 127 12.3 0.020 0.023 5.145 5.177 mp_allgather_i34 2485 14.5 1.779 5.045 1.779 5.045 mp_sum_l 7804 13.0 3.358 4.830 3.358 4.830 mp_waitany 11660 13.9 2.400 4.488 2.400 4.488 calculate_first_density_matrix 1 7.0 0.001 0.003 4.241 4.243 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.407 4.180 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.055 4.088 mp_alltoall_i22 712 14.1 2.000 3.980 2.000 3.980 rs_pw_transfer_RS2PW_140 138 11.5 0.328 0.352 1.924 3.927 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 3.781 3.784 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.661 3.700 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=182.636000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=896.181818, yerr=12.074151 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 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 1.135219E+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 660837789680 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4079 57274. MP_Allreduce 11236 1068. MP_Sync 168 MP_Alltoall 1700 12496381. 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 0.027 0.053 191.240 191.242 qs_mol_dyn_low 1 2.0 0.003 0.004 190.756 190.770 qs_forces 11 3.9 0.003 0.009 190.559 190.563 qs_energies 11 4.9 0.013 0.060 183.332 183.349 scf_env_do_scf 11 5.9 0.010 0.075 164.769 164.777 velocity_verlet 10 3.0 0.002 0.002 125.507 125.513 scf_env_do_scf_inner_loop 116 6.6 0.005 0.011 117.658 117.659 dbcsr_multiply_generic 2485 12.5 0.188 0.194 80.588 81.307 qs_scf_new_mos 116 7.6 0.001 0.001 79.844 80.170 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.843 80.170 ot_scf_mini 116 9.6 0.004 0.004 75.395 75.686 multiply_cannon 2485 13.5 0.547 0.594 55.845 59.622 multiply_cannon_loop 2485 14.5 1.162 1.196 52.250 53.986 init_scf_loop 11 6.9 0.002 0.008 46.914 46.916 prepare_preconditioner 11 7.9 0.000 0.000 42.712 42.742 make_preconditioner 11 8.9 0.000 0.002 42.712 42.742 ot_mini 116 10.6 0.001 0.001 42.123 42.410 make_full_inverse_cholesky 11 9.9 0.000 0.001 36.316 41.258 multiply_cannon_multrec 29820 15.5 13.917 19.008 26.020 30.860 rebuild_ks_matrix 127 8.3 0.001 0.001 30.154 30.420 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.027 30.154 30.419 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.160 27.392 mp_waitall_1 146592 16.7 17.666 27.305 17.666 27.305 qs_ot_get_derivative 116 11.6 0.001 0.002 22.662 22.954 make_m2s 4970 13.5 0.096 0.100 20.424 21.338 make_images 4970 14.5 1.933 2.207 20.121 21.034 qs_ot_get_p 127 10.4 0.001 0.001 19.561 19.908 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.929 19.450 apply_single 127 13.6 0.001 0.001 18.929 19.449 ot_diis_step 116 11.6 0.017 0.018 19.335 19.337 cp_fm_cholesky_invert 11 10.9 16.483 16.492 16.483 16.492 cp_fm_upper_to_full 104 14.7 11.037 16.244 11.037 16.244 qs_ot_p2m_diag 82 11.4 0.338 0.385 15.459 15.512 multiply_cannon_metrocomm3 29820 15.5 0.046 0.048 6.527 15.326 sum_up_and_integrate 127 10.3 0.139 0.151 14.901 14.924 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.768 14.797 calculate_rho_elec 127 8.7 0.174 0.189 14.767 14.796 integrate_v_rspace 127 11.3 0.004 0.005 14.761 14.790 cp_dbcsr_syevd 82 12.4 0.005 0.008 14.185 14.186 make_images_data 4970 15.5 0.062 0.067 11.151 12.936 dbcsr_complete_redistribute 393 12.7 1.513 1.628 9.149 12.866 init_scf_run 11 5.9 0.000 0.001 12.524 12.525 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.523 12.525 multiply_cannon_sync_h2d 29820 15.5 11.640 12.456 11.640 12.456 dbcsr_mm_accdrv_process 61748 16.2 7.285 8.413 11.682 12.288 hybrid_alltoall_any 5155 16.4 0.522 2.204 9.944 12.101 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.742 11.451 cp_fm_diag_elpa 82 13.4 0.000 0.001 10.935 10.938 cp_fm_redistribute_end 82 14.4 1.893 10.857 1.909 10.862 cp_fm_diag_elpa_base 82 14.4 8.346 10.235 8.919 10.723 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.373 9.969 pw_transfer 1535 11.6 0.088 0.107 9.545 9.603 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.291 9.514 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 9.319 9.383 mp_alltoall_i22 712 14.1 5.577 9.256 5.577 9.256 density_rs2pw 127 9.7 0.006 0.006 7.222 8.538 fft_wrap_pw1pw2_140 519 13.2 0.480 0.488 8.245 8.322 grid_integrate_task_list 127 12.3 7.466 7.946 7.466 7.946 cp_fm_cholesky_decompose 22 10.9 7.669 7.850 7.669 7.850 wfi_extrapolate 11 7.9 0.001 0.002 7.749 7.749 fft3d_ps 1281 14.7 2.751 2.846 7.555 7.615 multiply_cannon_metrocomm4 24850 15.5 0.076 0.087 2.794 7.312 mp_irecv_dv 75445 16.2 2.649 7.049 2.649 7.049 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.865 6.916 calculate_dm_sparse 127 9.5 0.001 0.001 6.309 6.414 rs_pw_transfer 1038 11.9 0.013 0.014 4.774 6.144 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.798 5.863 mp_alltoall_d11v 2401 14.1 5.138 5.824 5.138 5.824 grid_collocate_task_list 127 9.7 5.039 5.748 5.039 5.748 potential_pw2rs 127 12.3 0.022 0.023 4.736 4.754 calculate_first_density_matrix 1 7.0 0.013 0.047 4.563 4.568 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.428 4.521 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.477 4.478 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.161 4.213 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=191.242000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1074.272727, yerr=17.990356 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 1.511404E+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 4020 57953. MP_Allreduce 11047 1173. 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 0.017 0.036 177.443 177.445 qs_mol_dyn_low 1 2.0 0.003 0.003 177.050 177.064 qs_forces 11 3.9 0.003 0.003 176.885 176.891 qs_energies 11 4.9 0.002 0.002 169.253 169.261 scf_env_do_scf 11 5.9 0.001 0.001 150.644 150.647 velocity_verlet 10 3.0 0.035 0.047 115.460 115.465 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 114.615 114.616 dbcsr_multiply_generic 2507 12.6 0.182 0.186 75.707 76.196 qs_scf_new_mos 117 7.6 0.001 0.001 75.616 75.708 qs_scf_loop_do_ot 117 8.6 0.001 0.001 75.616 75.707 ot_scf_mini 117 9.6 0.003 0.004 71.203 71.298 multiply_cannon 2507 13.6 0.584 0.607 56.124 60.490 multiply_cannon_loop 2507 14.6 0.440 0.455 51.402 52.490 ot_mini 117 10.6 0.001 0.001 39.941 40.059 init_scf_loop 11 6.9 0.000 0.000 35.881 35.883 mp_waitall_1 125778 16.7 26.270 34.072 26.270 34.072 prepare_preconditioner 11 7.9 0.000 0.000 31.930 31.958 make_preconditioner 11 8.9 0.000 0.000 31.930 31.958 rebuild_ks_matrix 128 8.3 0.001 0.001 30.139 30.282 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 30.139 30.281 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.823 30.069 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.361 27.495 multiply_cannon_multrec 10028 15.6 10.393 16.192 17.658 22.062 multiply_cannon_metrocomm3 10028 15.6 0.023 0.024 12.855 20.124 ot_diis_step 117 11.6 0.020 0.021 20.084 20.084 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.659 20.017 apply_single 128 13.6 0.001 0.001 19.658 20.017 qs_ot_get_derivative 117 11.6 0.001 0.002 19.790 19.892 qs_ot_get_p 128 10.4 0.001 0.001 18.356 18.518 cp_fm_cholesky_invert 11 10.9 18.417 18.422 18.417 18.422 make_m2s 5014 13.6 0.066 0.069 15.943 18.350 make_images 5014 14.6 2.248 2.574 15.637 18.046 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.549 15.598 calculate_rho_elec 128 8.7 0.259 0.268 15.548 15.597 sum_up_and_integrate 128 10.3 0.180 0.190 15.195 15.241 integrate_v_rspace 128 11.3 0.004 0.004 15.015 15.071 qs_ot_p2m_diag 83 11.4 0.495 0.501 14.568 14.584 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.438 13.439 multiply_cannon_sync_h2d 10028 15.6 11.551 12.307 11.551 12.307 make_images_data 5014 15.6 0.051 0.060 9.711 12.242 hybrid_alltoall_any 5200 16.5 0.838 3.807 9.574 12.099 init_scf_run 11 5.9 0.000 0.001 11.762 11.762 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.762 11.762 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.363 10.372 cp_fm_diag_elpa_base 83 14.4 10.115 10.196 10.356 10.365 pw_transfer 1547 11.6 0.085 0.094 9.975 10.015 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.751 9.799 fft_wrap_pw1pw2_140 523 13.2 0.495 0.516 8.595 8.647 grid_integrate_task_list 128 12.3 7.747 8.257 7.747 8.257 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 7.955 8.033 density_rs2pw 128 9.7 0.005 0.006 7.122 8.017 cp_fm_cholesky_decompose 22 10.9 7.859 7.955 7.859 7.955 fft3d_ps 1291 14.7 2.706 2.781 7.914 7.932 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 4.445 7.601 dbcsr_mm_accdrv_process 20762 16.1 2.630 3.427 6.898 7.492 wfi_extrapolate 11 7.9 0.001 0.001 7.435 7.435 mp_allgather_i34 2507 14.6 2.927 7.381 2.927 7.381 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.502 6.511 calculate_dm_sparse 128 9.5 0.001 0.001 6.084 6.168 grid_collocate_task_list 128 9.7 5.392 6.064 5.392 6.064 mp_alltoall_d11v 2415 14.1 5.114 6.046 5.114 6.046 multiply_cannon_metrocomm4 7521 15.6 0.024 0.029 1.889 5.829 mp_irecv_dv 28860 15.9 1.851 5.741 1.851 5.741 dbcsr_complete_redistribute 395 12.7 2.118 2.163 5.185 5.579 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.399 5.443 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.340 5.340 rs_pw_transfer 1046 11.9 0.012 0.013 4.167 5.099 potential_pw2rs 128 12.3 0.027 0.028 4.621 4.636 calculate_first_density_matrix 1 7.0 0.001 0.001 4.111 4.111 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.560 3.897 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.740 3.792 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.619 3.720 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.682 3.700 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.366 3.680 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=177.445000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1414.000000, yerr=47.918491 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_performance_tests/20/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 11.696234E+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 1964048 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 3.027812E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 100280 MPI messages size (bytes): total size 1.136195E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.330227E+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 45208 35089547264 4194304 < size <= 16777216 44352 379752284160 16777216 < size 10104 721350232272 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 58886. MP_Allreduce 11047 1513. MP_Sync 87 MP_Alltoall 1712 36974159. MP_SendRecv 1792 218624. MP_ISendRecv 1792 218624. MP_Wait 9802 MP_ISend 6408 1080322. MP_IRecv 6408 1080322. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.068 308.647 308.649 qs_mol_dyn_low 1 2.0 0.015 0.104 308.072 308.095 qs_forces 11 3.9 0.005 0.024 307.632 307.636 qs_energies 11 4.9 0.014 0.103 297.902 297.918 scf_env_do_scf 11 5.9 0.001 0.005 273.966 273.975 velocity_verlet 10 3.0 0.002 0.003 221.048 221.082 scf_env_do_scf_inner_loop 117 6.6 0.005 0.025 144.516 144.519 init_scf_loop 11 6.9 0.001 0.007 129.183 129.196 prepare_preconditioner 11 7.9 0.000 0.001 123.965 124.019 make_preconditioner 11 8.9 0.001 0.006 123.965 124.018 make_full_inverse_cholesky 11 9.9 0.000 0.000 98.565 121.040 qs_scf_new_mos 117 7.6 0.001 0.002 93.044 93.274 qs_scf_loop_do_ot 117 8.6 0.001 0.006 93.043 93.273 ot_scf_mini 117 9.6 0.004 0.005 88.269 88.406 dbcsr_multiply_generic 2507 12.6 0.214 0.228 84.279 85.165 cp_fm_upper_to_full 105 14.8 54.786 79.177 54.786 79.177 multiply_cannon 2507 13.6 0.705 0.780 59.249 60.427 multiply_cannon_loop 2507 14.6 0.469 0.480 55.613 57.632 ot_mini 117 10.6 0.001 0.004 44.825 44.979 dbcsr_complete_redistribute 395 12.7 3.973 4.038 31.174 44.689 copy_fm_to_dbcsr 209 11.7 0.002 0.004 27.737 41.235 rebuild_ks_matrix 128 8.3 0.001 0.001 38.806 38.966 qs_ks_build_kohn_sham_matrix 128 9.3 0.020 0.040 38.805 38.965 transfer_fm_to_dbcsr 11 9.9 0.001 0.004 25.354 38.711 mp_alltoall_i22 716 14.1 23.220 36.865 23.220 36.865 qs_ks_update_qs_env 128 7.6 0.001 0.004 35.631 35.786 cp_fm_cholesky_invert 11 10.9 34.357 34.365 34.357 34.365 mp_waitall_1 103674 16.8 27.637 32.175 27.637 32.175 qs_ot_get_p 128 10.4 0.001 0.004 28.031 28.172 qs_ot_get_derivative 117 11.6 0.002 0.004 24.571 24.709 qs_ot_p2m_diag 83 11.4 0.878 0.884 23.748 23.776 cp_dbcsr_syevd 83 12.4 0.006 0.010 21.980 21.982 qs_rho_update_rho_low 128 7.7 0.001 0.002 21.008 21.027 calculate_rho_elec 128 8.7 0.483 0.490 21.007 21.027 make_m2s 5014 13.6 0.075 0.078 19.900 20.759 make_images 5014 14.6 3.744 3.914 19.421 20.285 ot_diis_step 117 11.6 0.024 0.033 20.210 20.211 sum_up_and_integrate 128 10.3 0.321 0.324 20.072 20.153 multiply_cannon_metrocomm3 10028 15.6 0.023 0.024 18.309 20.045 integrate_v_rspace 128 11.3 0.008 0.039 19.751 19.833 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.195 19.603 apply_single 128 13.6 0.001 0.001 19.195 19.603 cp_fm_diag_elpa 83 13.4 0.001 0.002 18.624 18.625 cp_fm_diag_elpa_base 83 14.4 14.110 15.793 18.615 18.617 multiply_cannon_multrec 10028 15.6 10.478 12.185 17.872 17.949 multiply_cannon_sync_h2d 10028 15.6 15.661 15.709 15.661 15.709 init_scf_run 11 5.9 0.000 0.002 13.510 13.511 scf_env_initial_rho_setup 11 6.9 0.000 0.000 13.510 13.511 pw_transfer 1547 11.6 0.092 0.093 12.563 12.579 hybrid_alltoall_any 5200 16.5 1.301 3.045 10.694 12.412 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 12.329 12.346 make_images_data 5014 15.6 0.059 0.065 10.544 12.198 fft_wrap_pw1pw2_140 523 13.2 0.542 0.548 10.902 10.924 fft3d_ps 1291 14.7 2.740 2.743 10.366 10.390 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.550 9.658 mp_alltoall_d11v 2415 14.1 8.454 9.524 8.454 9.524 cp_fm_cholesky_decompose 22 10.9 9.315 9.390 9.315 9.390 wfi_extrapolate 11 7.9 0.001 0.001 9.204 9.205 dbcsr_mm_accdrv_process 20762 16.1 3.818 5.744 7.155 8.980 grid_integrate_task_list 128 12.3 8.591 8.775 8.591 8.775 density_rs2pw 128 9.7 0.006 0.012 8.565 8.707 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 8.053 8.054 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.545 7.598 calculate_dm_sparse 128 9.5 0.001 0.001 6.619 6.741 grid_collocate_task_list 128 9.7 6.360 6.392 6.360 6.392 rs_scatter_matrices 139 9.7 3.613 4.558 6.055 6.290 copy_dbcsr_to_fm 186 11.8 0.004 0.005 6.142 6.222 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=308.649000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2701.454545, yerr=160.093361 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 1.261724E+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 0.023 0.038 87.099 87.101 qs_energies 1 2.0 0.000 0.000 86.590 86.605 ls_scf 1 3.0 0.005 0.041 85.683 85.699 dbcsr_multiply_generic 111 6.7 0.015 0.018 74.169 74.344 multiply_cannon 111 7.7 0.018 0.020 57.373 58.474 multiply_cannon_loop 111 8.7 0.210 0.228 53.998 55.266 ls_scf_main 1 4.0 0.000 0.000 52.391 52.396 density_matrix_trs4 2 5.0 0.002 0.003 46.796 46.864 ls_scf_init_scf 1 4.0 0.000 0.001 30.254 30.255 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.624 28.668 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.504 26.526 mp_waitall_1 11031 10.9 22.418 25.255 22.418 25.255 multiply_cannon_multrec 2664 9.7 8.190 8.896 15.476 17.165 multiply_cannon_sync_h2d 2664 9.7 13.807 15.125 13.807 15.125 make_m2s 222 7.7 0.008 0.011 13.123 13.586 make_images 222 8.7 0.099 0.112 13.101 13.567 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.540 11.764 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.534 9.405 make_images_data 222 9.7 0.004 0.005 7.668 8.228 hybrid_alltoall_any 227 10.6 0.215 1.845 6.566 7.973 dbcsr_mm_accdrv_process 4760 10.4 0.508 0.619 6.903 7.848 dbcsr_mm_accdrv_process_sort 4760 11.4 6.194 7.094 6.194 7.094 calculate_norms 4752 9.8 5.534 6.266 5.534 6.266 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.037 5.171 mp_sum_l 807 5.4 3.140 4.505 3.140 4.505 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.056 3.708 mp_irecv_dv 6231 10.9 2.039 3.682 2.039 3.682 make_images_sizes 222 9.7 0.000 0.000 0.751 3.498 mp_alltoall_i44 222 10.7 0.751 3.498 0.751 3.498 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.317 3.428 arnoldi_extremal 4 6.8 0.000 0.000 3.216 3.252 arnoldi_normal_ev 4 7.8 0.001 0.002 3.216 3.252 build_subspace 16 8.4 0.009 0.012 3.107 3.111 ls_scf_post 1 4.0 0.001 0.011 3.032 3.049 ls_scf_store_result 1 5.0 0.000 0.000 2.837 2.883 dbcsr_special_finalize 555 9.7 0.005 0.006 2.351 2.845 dbcsr_merge_single_wm 555 10.7 0.456 0.586 2.343 2.837 make_images_pack 222 9.7 2.206 2.627 2.208 2.629 compute_matrix_preconditioner 1 6.0 0.000 0.001 2.604 2.610 dbcsr_matrix_vector_mult 304 9.0 0.005 0.013 2.311 2.552 dbcsr_sort_data 658 11.4 2.146 2.546 2.146 2.546 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.459 2.069 2.461 ls_scf_dm_to_ks 2 5.0 0.000 0.025 2.304 2.375 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.316 2.318 rebuild_ks_matrix 3 7.3 0.000 0.000 2.307 2.308 qs_ks_build_kohn_sham_matrix 3 8.3 0.050 0.167 2.307 2.308 buffer_matrices_ensure_size 222 8.7 1.748 2.089 1.748 2.089 acc_transpose_blocks 2664 9.7 0.019 0.021 1.756 1.907 acc_transpose_blocks_kernels 2664 10.7 0.036 0.041 1.648 1.795 jit_kernel_transpose 1 13.0 1.611 1.758 1.611 1.758 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=87.101000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1133.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 2.102534E+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 0.034 0.047 92.280 92.282 qs_energies 1 2.0 0.000 0.001 91.754 91.760 ls_scf 1 3.0 0.000 0.001 90.418 90.423 dbcsr_multiply_generic 111 6.7 0.015 0.015 76.347 76.663 multiply_cannon 111 7.7 0.028 0.043 54.626 58.158 ls_scf_main 1 4.0 0.000 0.002 54.709 54.714 multiply_cannon_loop 111 8.7 0.116 0.122 51.354 54.210 density_matrix_trs4 2 5.0 0.002 0.004 48.919 49.126 ls_scf_init_scf 1 4.0 0.000 0.002 32.170 32.171 ls_scf_init_matrix_S 1 5.0 0.000 0.002 30.980 31.075 mp_waitall_1 9105 10.9 21.312 30.370 21.312 30.370 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.586 28.597 multiply_cannon_multrec 1332 9.7 13.030 16.580 22.212 26.949 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.997 21.307 make_m2s 222 7.7 0.006 0.008 15.075 15.720 make_images 222 8.7 1.569 1.933 15.045 15.690 dbcsr_mm_accdrv_process 4041 10.4 0.290 0.472 8.779 10.343 dbcsr_mm_accdrv_process_sort 4041 11.4 8.366 9.894 8.366 9.894 make_images_data 222 9.7 0.004 0.004 8.593 9.478 hybrid_alltoall_any 227 10.6 0.520 2.452 8.046 9.079 mp_sum_l 807 5.4 5.358 8.574 5.358 8.574 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.228 7.586 mp_irecv_dv 3311 11.0 3.209 7.534 3.209 7.534 calculate_norms 2376 9.8 5.997 6.825 5.997 6.825 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.109 6.775 multiply_cannon_sync_h2d 1332 9.7 4.794 5.920 4.794 5.920 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.117 5.349 arnoldi_extremal 4 6.8 0.000 0.000 4.735 4.755 arnoldi_normal_ev 4 7.8 0.001 0.005 4.735 4.755 build_subspace 16 8.4 0.014 0.021 4.476 4.478 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.350 3.625 ls_scf_post 1 4.0 0.000 0.001 3.538 3.545 ls_scf_store_result 1 5.0 0.000 0.000 3.251 3.363 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.142 3.362 dbcsr_matrix_vector_mult_local 304 10.0 2.735 3.214 2.737 3.216 ls_scf_dm_to_ks 2 5.0 0.000 0.010 2.616 2.707 make_images_pack 222 9.7 2.023 2.430 2.026 2.432 mp_allgather_i34 111 8.7 0.969 2.259 0.969 2.259 compute_matrix_preconditioner 1 6.0 0.000 0.001 2.170 2.176 dbcsr_sort_data 436 11.2 1.758 1.991 1.758 1.991 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.887 1.891 rebuild_ks_matrix 3 7.3 0.000 0.000 1.875 1.878 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.004 1.875 1.878 dbcsr_data_new 4174 10.1 1.618 1.854 1.618 1.854 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.282000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1748.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 2.687980E+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 0.050 0.097 95.494 95.495 qs_energies 1 2.0 0.001 0.011 94.759 94.781 ls_scf 1 3.0 0.003 0.014 93.327 93.341 dbcsr_multiply_generic 111 6.7 0.015 0.016 77.536 77.767 multiply_cannon 111 7.7 0.038 0.131 53.869 58.029 ls_scf_main 1 4.0 0.000 0.002 57.488 57.497 multiply_cannon_loop 111 8.7 0.099 0.106 50.303 53.693 density_matrix_trs4 2 5.0 0.003 0.016 51.469 51.621 mp_waitall_1 7281 11.0 23.955 33.806 23.955 33.806 ls_scf_init_scf 1 4.0 0.001 0.003 32.228 32.230 ls_scf_init_matrix_S 1 5.0 0.000 0.001 30.702 30.774 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.289 28.303 multiply_cannon_multrec 888 9.7 12.547 15.213 21.057 24.283 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.098 22.494 make_m2s 222 7.7 0.006 0.007 17.227 18.381 make_images 222 8.7 1.961 2.276 17.189 18.343 make_images_data 222 9.7 0.003 0.004 9.932 10.912 hybrid_alltoall_any 227 10.6 0.622 2.873 9.557 10.836 dbcsr_mm_accdrv_process 3754 10.4 0.256 0.431 8.027 9.240 dbcsr_mm_accdrv_process_sort 3754 11.4 7.652 8.810 7.652 8.810 mp_sum_l 807 5.4 4.897 8.208 4.897 8.208 multiply_cannon_sync_h2d 888 9.7 6.072 7.600 6.072 7.600 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.465 7.020 mp_irecv_dv 2335 11.1 2.450 6.979 2.450 6.979 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.701 6.453 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.550 6.157 arnoldi_extremal 4 6.8 0.000 0.000 5.219 5.238 arnoldi_normal_ev 4 7.8 0.001 0.005 5.219 5.238 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.979 5.219 build_subspace 16 8.4 0.014 0.020 4.913 4.918 calculate_norms 1584 9.8 4.280 4.647 4.280 4.647 mp_allgather_i34 111 8.7 1.406 3.916 1.406 3.916 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.501 3.859 ls_scf_post 1 4.0 0.004 0.016 3.608 3.623 dbcsr_matrix_vector_mult_local 304 10.0 3.004 3.571 3.006 3.573 ls_scf_store_result 1 5.0 0.000 0.000 3.340 3.417 ls_scf_dm_to_ks 2 5.0 0.001 0.018 2.860 2.954 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.138 2.140 dbcsr_sort_data 325 11.1 1.912 2.138 1.912 2.138 rebuild_ks_matrix 3 7.3 0.000 0.000 2.119 2.122 qs_ks_build_kohn_sham_matrix 3 8.3 0.015 0.051 2.119 2.122 make_images_pack 222 9.7 1.806 2.119 1.809 2.121 make_images_sizes 222 9.7 0.000 0.000 0.971 2.086 mp_alltoall_i44 222 10.7 0.971 2.085 0.971 2.085 compute_matrix_preconditioner 1 6.0 0.000 0.001 2.069 2.076 dbcsr_data_release 9322 10.9 1.314 1.957 1.314 1.957 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.495000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2193.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 3.346956E+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 0.043 0.067 98.214 98.215 qs_energies 1 2.0 0.001 0.009 97.464 97.478 ls_scf 1 3.0 0.000 0.001 95.793 95.802 dbcsr_multiply_generic 111 6.7 0.017 0.021 79.388 79.616 ls_scf_main 1 4.0 0.000 0.005 58.679 58.680 multiply_cannon 111 7.7 0.062 0.153 52.500 57.225 density_matrix_trs4 2 5.0 0.002 0.004 52.516 52.615 multiply_cannon_loop 111 8.7 0.115 0.128 47.401 50.469 ls_scf_init_scf 1 4.0 0.000 0.002 33.897 33.899 ls_scf_init_matrix_S 1 5.0 0.000 0.001 32.663 32.726 mp_waitall_1 6369 11.0 22.600 30.180 22.600 30.180 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 30.164 30.178 multiply_cannon_multrec 1332 9.7 14.185 17.335 21.994 24.481 make_m2s 222 7.7 0.006 0.008 21.131 22.586 make_images 222 8.7 3.136 3.599 21.081 22.538 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.109 17.028 make_images_data 222 9.7 0.004 0.004 11.799 13.509 hybrid_alltoall_any 227 10.6 0.798 3.844 11.157 12.991 dbcsr_mm_accdrv_process 3641 10.4 0.211 0.408 7.455 8.953 dbcsr_mm_accdrv_process_sort 3641 11.4 7.082 8.532 7.082 8.532 mp_sum_l 807 5.4 4.170 7.348 4.170 7.348 multiply_cannon_sync_h2d 1332 9.7 5.482 6.106 5.482 6.106 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.077 6.041 mp_irecv_dv 3229 10.9 2.053 5.964 2.053 5.964 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.181 5.764 arnoldi_extremal 4 6.8 0.000 0.000 5.287 5.302 arnoldi_normal_ev 4 7.8 0.001 0.005 5.287 5.302 build_subspace 16 8.4 0.014 0.021 4.946 4.954 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.519 4.945 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.537 4.735 mp_allgather_i34 111 8.7 2.228 4.714 2.228 4.714 calculate_norms 2376 9.8 4.193 4.522 4.193 4.522 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.604 3.893 dbcsr_matrix_vector_mult_local 304 10.0 3.166 3.659 3.168 3.661 dbcsr_sort_data 658 11.4 3.118 3.584 3.118 3.584 dbcsr_special_finalize 555 9.7 0.006 0.007 2.860 3.249 dbcsr_merge_single_wm 555 10.7 0.539 0.667 2.852 3.241 ls_scf_post 1 4.0 0.001 0.004 3.217 3.228 ls_scf_dm_to_ks 2 5.0 0.000 0.001 3.071 3.110 ls_scf_store_result 1 5.0 0.000 0.000 2.955 3.011 dbcsr_data_release 10477 10.7 1.572 2.440 1.572 2.440 dbcsr_finalize 304 7.8 0.049 0.061 1.810 2.085 compute_matrix_preconditioner 1 6.0 0.001 0.002 2.050 2.059 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.997 1.999 rebuild_ks_matrix 3 7.3 0.000 0.000 1.974 1.976 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.005 1.974 1.976 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.215000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2696.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 4.638097E+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 0.061 0.077 99.697 99.698 qs_energies 1 2.0 0.000 0.000 98.904 98.909 ls_scf 1 3.0 0.000 0.000 96.954 96.965 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.031 78.240 ls_scf_main 1 4.0 0.000 0.000 61.647 61.648 multiply_cannon 111 7.7 0.088 0.143 55.778 60.813 density_matrix_trs4 2 5.0 0.002 0.002 54.265 54.345 multiply_cannon_loop 111 8.7 0.069 0.082 51.230 53.068 ls_scf_init_scf 1 4.0 0.000 0.000 31.748 31.753 mp_waitall_1 5436 11.0 25.470 30.932 25.470 30.932 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.382 30.421 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.224 28.234 multiply_cannon_multrec 444 9.7 13.960 16.136 20.960 22.872 make_m2s 222 7.7 0.004 0.005 17.448 19.975 make_images 222 8.7 3.722 4.405 17.387 19.914 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.534 15.249 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.025 14.567 make_images_data 222 9.7 0.003 0.004 9.710 12.159 hybrid_alltoall_any 227 10.6 0.787 3.767 9.443 12.010 dbcsr_mm_accdrv_process 3003 10.4 0.177 0.347 6.706 7.829 multiply_cannon_sync_h2d 444 9.7 6.511 7.698 6.511 7.698 dbcsr_mm_accdrv_process_sort 3003 11.4 6.389 7.481 6.389 7.481 mp_allgather_i34 111 8.7 2.777 6.974 2.777 6.974 arnoldi_extremal 4 6.8 0.000 0.000 5.918 5.931 arnoldi_normal_ev 4 7.8 0.001 0.005 5.918 5.931 build_subspace 16 8.4 0.015 0.019 5.515 5.528 mp_sum_l 807 5.4 2.854 5.053 2.854 5.053 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.579 4.761 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.196 4.396 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.707 4.188 mp_irecv_dv 1241 11.2 1.689 4.167 1.689 4.167 dbcsr_matrix_vector_mult_local 304 10.0 3.660 4.134 3.662 4.136 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.773 3.864 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.973 3.799 calculate_norms 792 9.8 3.585 3.731 3.585 3.731 ls_scf_post 1 4.0 0.000 0.000 3.559 3.565 ls_scf_store_result 1 5.0 0.000 0.000 3.333 3.372 make_images_sizes 222 9.7 0.000 0.000 1.095 3.255 mp_alltoall_i44 222 10.7 1.095 3.255 1.095 3.255 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.266 2.267 dbcsr_finalize 304 7.8 0.062 0.078 2.200 2.244 rebuild_ks_matrix 3 7.3 0.000 0.000 2.234 2.235 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.234 2.235 dbcsr_merge_all 275 8.9 0.472 0.519 2.049 2.087 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.698000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3634.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/444a873b469bbe29fdbc1e637b703a89687f2299_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 8.722797E+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 0.081 0.097 110.898 110.898 qs_energies 1 2.0 0.000 0.000 109.444 109.465 ls_scf 1 3.0 0.000 0.000 106.501 106.521 dbcsr_multiply_generic 111 6.7 0.024 0.027 79.688 79.793 ls_scf_main 1 4.0 0.000 0.000 65.979 65.980 density_matrix_trs4 2 5.0 0.002 0.003 56.609 56.660 multiply_cannon 111 7.7 0.145 0.265 50.981 52.063 multiply_cannon_loop 111 8.7 0.067 0.070 47.807 49.540 ls_scf_init_scf 1 4.0 0.000 0.000 36.816 36.816 ls_scf_init_matrix_S 1 5.0 0.000 0.000 35.152 35.167 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 32.302 32.312 mp_waitall_1 4527 11.1 22.464 25.825 22.464 25.825 make_m2s 222 7.7 0.005 0.005 23.994 24.698 make_images 222 8.7 4.581 5.009 23.888 24.589 multiply_cannon_multrec 444 9.7 17.873 18.600 22.446 22.978 hybrid_alltoall_any 227 10.6 1.660 3.631 13.055 15.556 make_images_data 222 9.7 0.003 0.003 13.211 15.404 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.649 12.250 multiply_cannon_sync_h2d 444 9.7 8.847 8.898 8.847 8.898 arnoldi_extremal 4 6.8 0.000 0.000 7.493 7.504 arnoldi_normal_ev 4 7.8 0.002 0.009 7.493 7.504 build_subspace 16 8.4 0.026 0.036 6.929 6.939 dbcsr_matrix_vector_mult 304 9.0 0.016 0.032 5.506 5.691 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.317 5.410 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.084 5.327 dbcsr_matrix_vector_mult_local 304 10.0 4.955 5.256 4.958 5.258 dbcsr_mm_accdrv_process 1814 10.4 0.208 0.320 4.411 4.531 dbcsr_mm_accdrv_process_sort 1814 11.4 4.110 4.239 4.110 4.239 ls_scf_post 1 4.0 0.000 0.000 3.706 3.727 make_images_sizes 222 9.7 0.000 0.000 1.474 3.559 mp_alltoall_i44 222 10.7 1.474 3.559 1.474 3.559 ls_scf_store_result 1 5.0 0.000 0.000 3.456 3.466 calculate_norms 792 9.8 3.242 3.279 3.242 3.279 dbcsr_finalize 304 7.8 0.082 0.089 3.088 3.182 mp_sum_l 807 5.4 2.384 3.153 2.384 3.153 compute_matrix_preconditioner 1 6.0 0.002 0.002 3.138 3.144 dbcsr_merge_all 275 8.9 0.891 0.920 2.871 2.964 qs_energies_init_hamiltonians 1 3.0 0.002 0.004 2.913 2.913 dbcsr_complete_redistribute 5 7.6 1.433 1.472 2.776 2.885 dbcsr_data_release 12724 10.6 2.326 2.834 2.326 2.834 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.599 2.601 matrix_ls_to_qs 2 6.0 0.000 0.000 2.449 2.586 rebuild_ks_matrix 3 7.3 0.000 0.000 2.533 2.535 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.533 2.535 dbcsr_sort_data 325 11.1 2.440 2.516 2.440 2.516 mp_allgather_i34 111 8.7 0.822 2.480 0.822 2.480 dbcsr_new_transposed 4 7.5 0.243 0.254 2.337 2.364 dbcsr_frobenius_norm 74 6.6 2.057 2.137 2.200 2.242 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=110.898000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6827.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 444a873b469bbe29fdbc1e637b703a89687f2299 Summary: empty Status: OK