=== 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: 523d85c309ba47d334e69c8546862257e289ce38 ################# 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 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (26.12.2022) # # \ 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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/01 job id: 44060912 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/02 job id: 44060915 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/03 job id: 44060918 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/04 job id: 44060919 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/05 job id: 44060920 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/06 job id: 44060921 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/07 job id: 44060922 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/08 job id: 44060923 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/09 job id: 44060924 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/10 job id: 44060925 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/11 job id: 44060926 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/12 job id: 44060927 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/13 job id: 44060929 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/14 job id: 44060930 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/15 job id: 44060932 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/16 job id: 44060933 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/17 job id: 44060934 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/18 job id: 44060935 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/19 job id: 44060936 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/20 job id: 44060937 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/21 job id: 44060938 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/22 job id: 44060939 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/23 job id: 44060940 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/24 job id: 44060941 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/25 job id: 44060942 --- 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/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/26 job id: 44060943 --- 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/523d85c309ba47d334e69c8546862257e289ce38_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 0.035 0.036 135.335 135.336 farming_run 1 2.0 133.900 133.901 135.295 135.296 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.467036E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.057 114.558 114.589 qs_energies 1 2.0 0.000 0.000 114.319 114.321 mp2_main 1 3.0 0.000 0.000 112.477 112.479 mp2_gpw_main 1 4.0 0.020 0.026 111.667 111.669 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 93.159 93.277 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.419 55.537 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.170 41.622 46.582 get_2c_integrals 1 6.0 0.000 0.000 37.097 37.566 integrate_v_rspace 273 8.0 0.437 0.451 25.060 29.677 pw_transfer 6555 10.6 0.377 0.390 27.286 27.531 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 25.951 26.165 grid_integrate_task_list 273 9.0 20.890 25.942 20.890 25.942 fft_wrap_pw1pw2_100 2178 12.4 1.181 1.272 23.512 23.766 compute_2c_integrals 1 7.0 0.038 0.060 19.452 19.453 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.851 18.952 mp2_eri_2c_integrate_gpw 1 9.0 2.391 2.429 18.848 18.948 rpa_ri_compute_en 1 5.0 0.001 0.001 18.400 18.484 cp_fm_cholesky_decompose 12 8.2 17.585 18.060 17.585 18.060 cholesky_decomp 1 7.0 0.000 0.000 16.498 16.962 fft3d_s 5443 13.4 16.095 16.336 16.117 16.358 ao_to_mo_and_store_B_mult_1 272 7.0 10.846 15.541 10.846 15.541 calculate_wavefunction 272 8.0 5.380 5.460 12.477 13.098 rpa_num_int 1 6.0 0.000 0.004 10.502 10.511 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.473 10.500 calc_mat_Q 8 8.0 0.000 0.000 9.333 9.436 contract_S_to_Q 8 9.0 0.000 0.000 8.756 8.861 calc_potential_gpw 544 9.5 0.005 0.006 8.224 8.618 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.193 8.440 parallel_gemm_fm 14 9.1 0.000 0.000 8.330 8.419 parallel_gemm_fm_cosma 14 10.1 8.330 8.419 8.330 8.419 potential_pw2rs 545 10.0 0.107 0.109 7.680 8.368 create_integ_mat 1 6.0 0.014 0.027 7.684 7.693 collocate_single_gaussian 272 10.0 0.040 0.042 7.446 7.692 array2fm 1 7.0 0.000 0.000 6.627 7.142 pw_scatter_s 2720 13.7 4.438 4.616 4.438 4.616 pw_gather_s 2722 13.2 3.867 4.233 3.867 4.233 array2fm_buffer_send 1 8.0 2.923 3.146 2.923 3.146 ao_to_mo_and_store_B_E_Ex_1 272 7.0 1.221 1.367 2.173 2.360 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.669496, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2739.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_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 0.027 0.036 399.280 399.281 farming_run 1 2.0 398.489 398.494 399.245 399.247 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.224733E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 14123339. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.045 209.526 209.526 qs_energies 1 2.0 0.000 0.000 209.266 209.282 scf_env_do_scf 1 3.0 0.000 0.000 106.450 106.450 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.602 105.611 rebuild_ks_matrix 4 6.0 0.000 0.000 105.601 105.610 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.063 105.601 105.610 hfx_ks_matrix 4 8.0 0.001 0.001 105.232 105.235 integrate_four_center 4 9.0 0.143 0.455 105.231 105.234 mp2_main 1 3.0 0.000 0.000 102.536 102.552 mp2_gpw_main 1 4.0 0.032 0.044 101.705 101.721 integrate_four_center_main 4 10.0 0.129 0.605 96.815 99.297 integrate_four_center_bin 269 11.0 96.686 99.229 96.686 99.229 init_scf_loop 1 4.0 0.000 0.000 91.994 91.994 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.981 76.086 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.551 55.655 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.159 42.015 47.311 integrate_v_rspace 95 8.0 0.399 0.565 28.382 33.507 pw_transfer 2240 10.6 0.145 0.181 29.967 30.502 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.943 29.503 grid_integrate_task_list 95 9.0 23.672 28.908 23.672 28.908 ao_to_mo_and_store_B_mult_1 91 7.0 10.847 28.461 10.847 28.461 mp2_ri_gpw_compute_en 1 5.0 0.060 0.073 26.578 28.295 fft_wrap_pw1pw2_100 730 12.4 1.321 1.488 26.682 27.226 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.846 1.902 24.825 24.835 get_2c_integrals 1 6.0 0.000 0.000 20.343 20.365 compute_2c_integrals 1 7.0 0.002 0.003 19.327 19.330 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 19.003 19.200 mp2_eri_2c_integrate_gpw 1 9.0 1.723 1.865 19.002 19.198 fft3d_s 1823 13.4 18.429 18.709 18.442 18.723 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.455 14.455 calculate_wavefunction 91 8.0 2.011 2.040 9.738 9.975 potential_pw2rs 186 10.0 0.034 0.036 8.641 9.236 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.589 8.761 9.186 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.329 8.625 local_gemm 172 8.0 8.204 8.615 8.204 8.615 calc_potential_gpw 182 9.5 0.002 0.003 8.000 8.254 mp2_ri_gpw_compute_en_comm 22 7.0 0.503 0.520 7.827 8.200 collocate_single_gaussian 91 10.0 0.017 0.025 7.934 8.131 mp2_ri_gpw_compute_en_ener 172 7.0 6.346 6.420 6.346 6.420 mp_sendrecv_dm3 2068 8.0 5.852 6.202 5.852 6.202 mp_sync 38 10.4 3.070 5.588 3.070 5.588 pw_gather_s 912 13.2 4.880 5.495 4.880 5.495 pw_scatter_s 910 13.7 3.948 4.412 3.948 4.412 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.696567, 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/523d85c309ba47d334e69c8546862257e289ce38_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 451.129344E+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 1117500. 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 0.015 0.028 58.666 58.667 qs_mol_dyn_low 1 2.0 0.003 0.006 57.125 57.133 qs_forces 11 3.9 0.002 0.003 57.054 57.056 qs_energies 11 4.9 0.002 0.006 55.601 55.613 scf_env_do_scf 11 5.9 0.000 0.001 44.346 44.346 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 41.444 41.444 dbcsr_multiply_generic 2286 12.5 0.095 0.109 33.796 34.258 qs_scf_new_mos 108 7.5 0.000 0.001 31.613 31.925 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.613 31.925 ot_scf_mini 108 9.5 0.002 0.002 29.959 30.181 multiply_cannon 2286 13.5 0.185 0.194 26.536 28.247 multiply_cannon_loop 2286 14.5 1.509 1.601 25.871 27.578 velocity_verlet 10 3.0 0.001 0.001 25.129 25.130 ot_mini 108 10.5 0.001 0.001 18.878 19.145 qs_ot_get_derivative 108 11.5 0.001 0.001 15.913 16.129 mp_waitall_1 245248 16.5 8.406 15.109 8.406 15.109 multiply_cannon_metrocomm3 54864 15.5 0.068 0.074 6.023 13.594 multiply_cannon_multrec 54864 15.5 4.222 6.557 7.787 11.119 init_scf_run 11 5.9 0.000 0.001 9.927 9.927 scf_env_initial_rho_setup 11 6.9 0.001 0.001 9.927 9.927 rebuild_ks_matrix 119 8.3 0.000 0.000 8.101 8.281 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.101 8.281 calculate_first_density_matrix 1 7.0 0.007 0.057 7.836 7.845 qs_ot_get_p 119 10.4 0.001 0.001 7.027 7.375 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.192 7.355 multiply_cannon_sync_h2d 54864 15.5 5.893 7.335 5.893 7.335 mp_sum_l 7207 12.9 5.389 7.137 5.389 7.137 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.337 5.795 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.366 5.499 parallel_gemm_fm 81 9.0 0.000 0.000 5.081 5.085 parallel_gemm_fm_cosma 81 10.0 5.081 5.085 5.081 5.085 dbcsr_mm_accdrv_process 76910 16.1 1.188 1.851 3.487 4.782 make_basis_sm 11 9.8 0.039 0.065 4.717 4.719 sum_up_and_integrate 119 10.3 0.012 0.014 4.457 4.487 integrate_v_rspace 119 11.3 0.002 0.002 4.445 4.476 qs_rho_update_rho_low 119 7.7 0.000 0.000 3.990 4.109 calculate_rho_elec 119 8.7 0.011 0.016 3.989 4.109 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.756 3.839 calculate_dm_sparse 119 9.5 0.000 0.000 3.172 3.320 multiply_cannon_metrocomm1 54864 15.5 0.052 0.059 1.726 3.217 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.877 3.080 apply_single 119 13.6 0.000 0.000 2.877 3.080 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.962 2.964 acc_transpose_blocks 54864 15.5 0.233 0.256 2.234 2.885 init_scf_loop 11 6.9 0.000 0.000 2.885 2.885 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.882 2.883 jit_kernel_multiply 13 15.8 2.237 2.880 2.237 2.880 rs_pw_transfer 974 11.9 0.012 0.013 2.714 2.799 mp_sum_d 4125 12.0 1.707 2.790 1.707 2.790 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.693 2.745 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.686 2.687 cp_fm_redistribute_end 50 14.0 2.447 2.666 2.452 2.668 ot_diis_step 108 11.5 0.006 0.006 2.665 2.666 cp_fm_diag_elpa_base 50 14.0 0.214 2.579 0.215 2.591 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.380 2.434 density_rs2pw 119 9.7 0.004 0.004 2.105 2.204 grid_integrate_task_list 119 12.3 2.030 2.131 2.030 2.131 wfi_extrapolate 11 7.9 0.001 0.001 2.030 2.030 potential_pw2rs 119 12.3 0.004 0.004 1.834 1.865 acc_transpose_blocks_kernels 54864 16.5 0.253 0.386 1.237 1.735 pw_transfer 1439 11.6 0.051 0.056 1.609 1.698 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.676 1.694 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 1.534 1.628 make_m2s 4572 13.5 0.054 0.056 1.562 1.603 make_images 4572 14.5 0.133 0.139 1.479 1.519 mp_alltoall_d11v 2130 13.8 1.252 1.442 1.252 1.442 mp_waitany 12084 13.8 1.230 1.419 1.230 1.419 fft3d_ps 1201 14.6 0.358 0.460 1.313 1.394 grid_collocate_task_list 119 9.7 1.291 1.390 1.291 1.390 jit_kernel_transpose 5 15.5 0.984 1.357 0.984 1.357 create_qs_kind_set 1 2.0 0.001 0.002 1.343 1.355 read_qs_kind 2 3.0 0.012 0.026 1.342 1.355 parser_read_line 2807 4.0 0.001 0.001 1.330 1.346 parser_read_line_low 5 5.0 0.015 1.314 1.329 1.345 broadcast_input_information 5 6.0 0.000 0.000 1.314 1.343 mp_bcast_i 105 3.0 1.306 1.338 1.306 1.338 fft_wrap_pw1pw2_140 487 13.2 0.080 0.094 1.193 1.288 arnoldi_extremal 119 11.4 0.001 0.001 1.143 1.283 arnoldi_normal_ev 119 12.4 0.300 0.503 1.142 1.283 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=58.667000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.363636, yerr=0.979121 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_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 488.124416E+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 1066951. 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 0.019 0.031 39.486 39.488 qs_mol_dyn_low 1 2.0 0.004 0.005 39.110 39.118 qs_forces 11 3.9 0.002 0.003 39.045 39.050 qs_energies 11 4.9 0.005 0.008 37.369 37.377 scf_env_do_scf 11 5.9 0.001 0.004 31.153 31.154 scf_env_do_scf_inner_loop 108 6.5 0.040 0.063 28.676 28.678 dbcsr_multiply_generic 2286 12.5 0.099 0.103 22.248 22.601 qs_scf_new_mos 108 7.5 0.001 0.001 19.784 20.027 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.784 20.027 ot_scf_mini 108 9.5 0.003 0.003 18.892 19.063 multiply_cannon 2286 13.5 0.209 0.217 17.334 18.880 velocity_verlet 10 3.0 0.001 0.002 17.926 17.927 multiply_cannon_loop 2286 14.5 0.906 0.978 16.232 17.580 ot_mini 108 10.5 0.001 0.001 11.662 11.899 mp_waitall_1 200699 16.5 5.543 10.993 5.543 10.993 multiply_cannon_multrec 27432 15.5 1.974 4.464 7.002 9.917 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.186 9.594 qs_ot_get_derivative 108 11.5 0.001 0.001 9.205 9.380 rebuild_ks_matrix 119 8.3 0.000 0.000 7.086 7.225 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 7.085 7.225 dbcsr_mm_accdrv_process 47894 16.0 3.872 6.088 4.958 6.943 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.254 6.380 init_scf_run 11 5.9 0.001 0.004 4.965 4.965 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.964 4.965 qs_ot_get_p 119 10.4 0.001 0.001 4.358 4.584 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.549 4.393 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.043 4.118 apply_single 119 13.6 0.000 0.000 3.043 4.118 sum_up_and_integrate 119 10.3 0.024 0.027 4.099 4.108 integrate_v_rspace 119 11.3 0.002 0.003 4.075 4.085 mp_sum_l 7207 12.9 2.020 3.976 2.020 3.976 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.565 3.649 calculate_rho_elec 119 8.7 0.021 0.024 3.564 3.649 calculate_first_density_matrix 1 7.0 0.001 0.002 3.461 3.468 calculate_dm_sparse 119 9.5 0.000 0.001 3.137 3.215 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.904 2.923 multiply_cannon_sync_h2d 27432 15.5 2.195 2.853 2.195 2.853 make_m2s 4572 13.5 0.052 0.054 2.450 2.669 make_images 4572 14.5 0.201 0.239 2.361 2.580 rs_pw_transfer 974 11.9 0.010 0.011 2.473 2.554 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.471 2.471 init_scf_loop 11 6.9 0.002 0.004 2.454 2.457 jit_kernel_multiply 9 16.2 1.033 2.421 1.033 2.421 ot_diis_step 108 11.5 0.010 0.011 2.407 2.408 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 2.351 2.361 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.107 2.199 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.149 2.150 cp_fm_redistribute_end 50 14.0 1.780 2.126 1.783 2.127 cp_fm_diag_elpa_base 50 14.0 0.329 2.020 0.342 2.070 density_rs2pw 119 9.7 0.004 0.004 1.912 1.997 grid_integrate_task_list 119 12.3 1.837 1.942 1.837 1.942 potential_pw2rs 119 12.3 0.006 0.006 1.848 1.859 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.855 1.858 pw_transfer 1439 11.6 0.063 0.066 1.749 1.779 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.659 1.698 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.660 1.692 make_images_data 4572 15.5 0.045 0.052 1.131 1.547 prepare_preconditioner 11 7.9 0.000 0.000 1.502 1.528 make_preconditioner 11 8.9 0.001 0.001 1.502 1.528 acc_transpose_blocks 27432 15.5 0.110 0.115 1.185 1.482 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.407 1.464 hybrid_alltoall_any 4725 16.4 0.051 0.112 0.980 1.463 wfi_extrapolate 11 7.9 0.001 0.001 1.447 1.448 fft3d_ps 1201 14.6 0.499 0.553 1.367 1.396 fft_wrap_pw1pw2_140 487 13.2 0.076 0.083 1.303 1.337 grid_collocate_task_list 119 9.7 1.246 1.313 1.246 1.313 mp_alltoall_d11v 2130 13.8 1.181 1.303 1.181 1.303 mp_allgather_i34 2286 14.5 0.544 1.286 0.544 1.286 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.165 1.214 mp_sum_d 4125 12.0 0.603 1.024 0.603 1.024 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.984 0.990 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.888 0.902 acc_transpose_blocks_kernels 27432 16.5 0.182 0.269 0.652 0.859 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.488000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.727273, yerr=1.710444 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 520.347648E+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 0.023 0.051 39.597 39.599 qs_mol_dyn_low 1 2.0 0.003 0.005 37.969 37.976 qs_forces 11 3.9 0.002 0.003 37.910 37.912 qs_energies 11 4.9 0.002 0.010 36.356 36.360 scf_env_do_scf 11 5.9 0.001 0.003 25.916 25.917 scf_env_do_scf_inner_loop 108 6.5 0.004 0.009 22.469 22.470 dbcsr_multiply_generic 2286 12.5 0.093 0.098 17.244 17.313 velocity_verlet 10 3.0 0.001 0.001 14.958 14.959 multiply_cannon 2286 13.5 0.195 0.201 14.004 14.764 qs_scf_new_mos 108 7.5 0.001 0.001 14.630 14.656 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.630 14.655 multiply_cannon_loop 2286 14.5 0.638 0.662 13.214 14.010 ot_scf_mini 108 9.5 0.002 0.003 13.915 13.928 init_scf_run 11 5.9 0.000 0.001 9.252 9.252 scf_env_initial_rho_setup 11 6.9 0.001 0.001 9.252 9.252 ot_mini 108 10.5 0.001 0.001 8.591 8.600 calculate_first_density_matrix 1 7.0 0.014 0.111 7.947 7.948 qs_ot_get_derivative 108 11.5 0.001 0.001 7.109 7.123 rebuild_ks_matrix 119 8.3 0.000 0.000 7.059 7.087 qs_ks_build_kohn_sham_matrix 119 9.3 0.041 0.155 7.058 7.087 multiply_cannon_multrec 18288 15.5 1.945 2.832 6.856 7.058 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.336 6.354 dbcsr_mm_accdrv_process 38222 16.0 4.027 5.473 4.826 5.753 make_basis_sm 11 9.8 0.115 0.237 4.824 4.826 parallel_gemm_fm 81 9.0 0.000 0.000 4.649 4.652 parallel_gemm_fm_cosma 81 10.0 4.649 4.652 4.649 4.652 sum_up_and_integrate 119 10.3 0.030 0.031 3.955 4.015 integrate_v_rspace 119 11.3 0.011 0.072 3.924 3.988 init_scf_loop 11 6.9 0.011 0.079 3.427 3.430 mp_waitall_1 158411 16.6 2.460 3.340 2.460 3.340 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.293 3.307 calculate_rho_elec 119 8.7 0.030 0.031 3.293 3.306 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.655 3.277 qs_ot_get_p 119 10.4 0.001 0.001 3.173 3.188 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.482 2.484 rs_pw_transfer 974 11.9 0.009 0.010 2.249 2.357 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.330 2.337 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.028 2.333 apply_single 119 13.6 0.000 0.000 2.028 2.332 acc_transpose_blocks 18288 15.5 0.077 0.078 2.233 2.250 calculate_dm_sparse 119 9.5 0.001 0.002 2.158 2.185 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.124 2.129 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.336 2.058 make_m2s 4572 13.5 0.044 0.045 1.767 1.922 density_rs2pw 119 9.7 0.004 0.004 1.825 1.921 grid_integrate_task_list 119 12.3 1.800 1.895 1.800 1.895 jit_kernel_multiply 10 16.0 0.747 1.881 0.747 1.881 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.861 1.861 make_images 4572 14.5 0.192 0.202 1.682 1.835 pw_transfer 1439 11.6 0.063 0.066 1.742 1.810 acc_transpose_blocks_kernels 18288 16.5 0.210 0.217 1.792 1.805 potential_pw2rs 119 12.3 0.007 0.008 1.724 1.791 prepare_preconditioner 11 7.9 0.000 0.000 1.767 1.777 make_preconditioner 11 8.9 0.006 0.036 1.767 1.777 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.652 1.721 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.619 1.712 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.629 1.630 cp_fm_redistribute_end 50 14.0 1.216 1.610 1.217 1.611 jit_kernel_transpose 5 15.6 1.583 1.588 1.583 1.588 cp_fm_diag_elpa_base 50 14.0 0.378 1.527 0.392 1.575 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.535 1.551 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.532 1.540 multiply_cannon_sync_h2d 18288 15.5 1.420 1.536 1.420 1.536 ot_diis_step 108 11.5 0.011 0.011 1.464 1.464 create_qs_kind_set 1 2.0 0.001 0.001 1.421 1.437 read_qs_kind 2 3.0 0.012 0.020 1.420 1.437 parser_read_line 2807 4.0 0.001 0.001 1.408 1.429 parser_read_line_low 5 5.0 0.045 1.403 1.408 1.428 broadcast_input_information 5 6.0 0.000 0.000 1.363 1.427 mp_sum_l 7207 12.9 1.089 1.422 1.089 1.422 mp_bcast_i 105 3.0 1.357 1.421 1.357 1.421 fft3d_ps 1201 14.6 0.506 0.522 1.338 1.411 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.320 1.390 grid_collocate_task_list 119 9.7 1.216 1.310 1.216 1.310 wfi_extrapolate 11 7.9 0.001 0.001 1.154 1.154 mp_sum_d 4125 12.0 0.820 1.094 0.820 1.094 qs_vxc_create 119 10.3 0.033 0.118 0.916 0.994 xc_vxc_pw_create 119 11.3 0.109 0.232 0.883 0.990 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.958 0.960 make_images_data 4572 15.5 0.045 0.049 0.786 0.952 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.868 0.891 hybrid_alltoall_any 4725 16.4 0.055 0.111 0.674 0.869 mp_alltoall_d11v 2130 13.8 0.735 0.862 0.735 0.862 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.801 0.803 mp_alltoall_z22v 1201 16.6 0.675 0.792 0.675 0.792 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=39.599000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.727273, yerr=1.482682 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 557.592576E+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 0.052 0.201 38.498 38.502 qs_mol_dyn_low 1 2.0 0.003 0.004 37.982 37.989 qs_forces 11 3.9 0.002 0.003 37.925 37.925 qs_energies 11 4.9 0.001 0.004 36.229 36.234 scf_env_do_scf 11 5.9 0.001 0.005 29.193 29.194 scf_env_do_scf_inner_loop 108 6.5 0.003 0.012 23.932 23.933 dbcsr_multiply_generic 2286 12.5 0.099 0.101 20.027 20.151 velocity_verlet 10 3.0 0.001 0.001 17.361 17.362 multiply_cannon 2286 13.5 0.233 0.267 16.406 16.873 qs_scf_new_mos 108 7.5 0.001 0.001 15.951 16.004 qs_scf_loop_do_ot 108 8.5 0.001 0.005 15.951 16.003 multiply_cannon_loop 2286 14.5 0.944 0.975 15.470 15.834 ot_scf_mini 108 9.5 0.004 0.018 15.025 15.073 multiply_cannon_multrec 27432 15.5 2.335 2.988 9.385 9.719 ot_mini 108 10.5 0.001 0.001 9.239 9.307 rebuild_ks_matrix 119 8.3 0.000 0.000 8.215 8.277 qs_ks_build_kohn_sham_matrix 119 9.3 0.048 0.304 8.215 8.276 dbcsr_mm_accdrv_process 47916 15.9 6.070 7.975 6.955 8.201 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.501 7.556 qs_ot_get_derivative 108 11.5 0.001 0.001 7.437 7.488 init_scf_run 11 5.9 0.000 0.001 5.656 5.657 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.656 5.657 init_scf_loop 11 6.9 0.001 0.005 5.238 5.238 calculate_first_density_matrix 1 7.0 0.000 0.002 4.312 4.312 sum_up_and_integrate 119 10.3 0.035 0.037 3.677 3.686 integrate_v_rspace 119 11.3 0.002 0.003 3.642 3.652 qs_ot_get_p 119 10.4 0.001 0.001 3.262 3.333 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.204 3.236 calculate_rho_elec 119 8.7 0.040 0.046 3.203 3.236 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.673 3.162 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.071 3.073 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.895 2.912 calculate_dm_sparse 119 9.5 0.000 0.001 2.815 2.871 acc_transpose_blocks 27432 15.5 0.114 0.120 2.688 2.723 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.118 2.573 apply_single 119 13.6 0.000 0.000 2.118 2.573 prepare_preconditioner 11 7.9 0.000 0.000 2.500 2.507 make_preconditioner 11 8.9 0.000 0.002 2.500 2.507 mp_sum_d 4123 12.0 2.013 2.478 2.013 2.478 jit_kernel_multiply 9 16.2 0.824 2.434 0.824 2.434 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.117 2.431 make_m2s 4572 13.5 0.054 0.056 2.132 2.241 mp_waitall_1 137007 16.6 1.739 2.229 1.739 2.229 make_images 4572 14.5 0.273 0.332 2.025 2.131 acc_transpose_blocks_kernels 27432 16.5 0.267 0.274 2.059 2.086 rs_pw_transfer 974 11.9 0.009 0.009 1.911 2.048 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.007 2.018 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.970 1.979 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.901 1.928 grid_integrate_task_list 119 12.3 1.824 1.904 1.824 1.904 qs_vxc_create 119 10.3 0.145 1.148 1.858 1.901 xc_vxc_pw_create 119 11.3 0.044 0.071 1.713 1.900 jit_kernel_transpose 5 15.6 1.792 1.813 1.792 1.813 density_rs2pw 119 9.7 0.004 0.004 1.690 1.804 mp_sum_l 7207 12.9 1.220 1.774 1.220 1.774 ot_diis_step 108 11.5 0.012 0.012 1.761 1.761 pw_transfer 1439 11.6 0.063 0.066 1.729 1.759 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.639 1.672 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.656 1.656 potential_pw2rs 119 12.3 0.008 0.009 1.472 1.478 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.430 1.431 cp_fm_redistribute_end 50 14.0 0.948 1.409 0.949 1.409 cp_fm_diag_elpa_base 50 14.0 0.438 1.345 0.457 1.383 fft_wrap_pw1pw2_140 487 13.2 0.085 0.094 1.347 1.382 fft3d_ps 1201 14.6 0.534 0.582 1.320 1.343 grid_collocate_task_list 119 9.7 1.221 1.316 1.221 1.316 wfi_extrapolate 11 7.9 0.001 0.001 1.299 1.299 multiply_cannon_metrocomm3 27432 15.5 0.038 0.040 0.772 1.192 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.125 1.125 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.087 1.107 cp_fm_upper_to_full 72 13.5 0.789 1.088 0.789 1.088 dbcsr_complete_redistribute 329 12.2 0.119 0.142 0.778 1.043 multiply_cannon_sync_h2d 27432 15.5 0.985 1.040 0.985 1.040 mp_alltoall_d11v 2130 13.8 0.732 0.938 0.732 0.938 make_images_data 4572 15.5 0.046 0.049 0.818 0.928 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.795 0.874 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.708 0.863 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.557 0.816 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.805 0.810 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=38.502000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.545455, yerr=3.551277 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_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 608.501760E+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 0.043 0.060 28.829 28.829 qs_mol_dyn_low 1 2.0 0.003 0.003 28.598 28.605 qs_forces 11 3.9 0.001 0.002 28.540 28.541 qs_energies 11 4.9 0.001 0.001 26.841 26.844 scf_env_do_scf 11 5.9 0.000 0.001 20.933 20.933 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.455 18.456 velocity_verlet 10 3.0 0.001 0.001 14.077 14.079 dbcsr_multiply_generic 2286 12.5 0.092 0.096 13.366 13.443 multiply_cannon 2286 13.5 0.231 0.239 10.848 11.212 qs_scf_new_mos 108 7.5 0.001 0.001 10.960 10.992 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.960 10.991 ot_scf_mini 108 9.5 0.002 0.002 10.292 10.322 multiply_cannon_loop 2286 14.5 0.331 0.342 9.948 10.153 multiply_cannon_multrec 9144 15.5 1.660 1.913 7.030 7.273 ot_mini 108 10.5 0.001 0.001 5.752 5.785 rebuild_ks_matrix 119 8.3 0.000 0.000 5.752 5.773 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.751 5.772 dbcsr_mm_accdrv_process 12550 15.8 4.152 5.202 5.269 5.352 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.121 5.141 init_scf_run 11 5.9 0.000 0.001 4.485 4.485 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.485 4.485 qs_ot_get_derivative 108 11.5 0.001 0.001 4.426 4.455 sum_up_and_integrate 119 10.3 0.037 0.041 3.517 3.522 integrate_v_rspace 119 11.3 0.002 0.003 3.479 3.485 calculate_first_density_matrix 1 7.0 0.000 0.000 3.378 3.378 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.178 3.186 calculate_rho_elec 119 8.7 0.059 0.061 3.178 3.186 qs_ot_get_p 119 10.4 0.001 0.001 2.753 2.802 calculate_dm_sparse 119 9.5 0.000 0.000 2.783 2.800 init_scf_loop 11 6.9 0.000 0.000 2.455 2.456 jit_kernel_multiply 7 15.7 1.078 2.361 1.078 2.361 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.207 2.212 mp_waitall_1 115863 16.7 1.426 1.943 1.426 1.943 grid_integrate_task_list 119 12.3 1.856 1.930 1.856 1.930 make_m2s 4572 13.5 0.034 0.035 1.726 1.879 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.843 1.845 make_images 4572 14.5 0.269 0.302 1.637 1.787 prepare_preconditioner 11 7.9 0.000 0.000 1.722 1.727 make_preconditioner 11 8.9 0.000 0.000 1.722 1.727 pw_transfer 1439 11.6 0.063 0.065 1.687 1.694 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.680 1.681 rs_pw_transfer 974 11.9 0.008 0.008 1.603 1.674 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.644 1.644 density_rs2pw 119 9.7 0.003 0.004 1.580 1.631 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.602 1.626 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.596 1.604 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.556 1.571 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.525 1.531 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.364 1.365 cp_fm_redistribute_end 50 14.0 0.681 1.345 0.682 1.345 cp_fm_diag_elpa_base 50 14.0 0.618 1.275 0.662 1.329 grid_collocate_task_list 119 9.7 1.271 1.319 1.271 1.319 ot_diis_step 108 11.5 0.012 0.013 1.314 1.315 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.298 1.306 potential_pw2rs 119 12.3 0.010 0.010 1.298 1.302 fft3d_ps 1201 14.6 0.539 0.550 1.264 1.272 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.214 1.214 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.167 1.187 apply_single 119 13.6 0.000 0.000 1.167 1.186 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.112 1.126 wfi_extrapolate 11 7.9 0.001 0.001 1.064 1.064 hybrid_alltoall_any 4725 16.4 0.063 0.176 0.790 1.017 acc_transpose_blocks 9144 15.5 0.039 0.040 0.985 0.991 make_images_data 4572 15.5 0.040 0.043 0.808 0.982 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.870 0.930 mp_alltoall_d11v 2130 13.8 0.823 0.910 0.823 0.910 cp_fm_cholesky_invert 11 10.9 0.868 0.871 0.868 0.871 multiply_cannon_sync_h2d 9144 15.5 0.715 0.788 0.715 0.788 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.752 0.755 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.744 0.747 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.722 0.731 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.667 0.718 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.347 0.665 mp_allgather_i34 2286 14.5 0.235 0.654 0.235 0.654 mp_alltoall_z22v 1201 16.6 0.599 0.633 0.599 0.633 jit_kernel_transpose 5 15.6 0.627 0.629 0.627 0.629 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.829000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=573.636364, yerr=8.967342 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_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 766.980096E+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 0.026 0.047 43.668 43.669 qs_mol_dyn_low 1 2.0 0.003 0.005 43.227 43.234 qs_forces 11 3.9 0.003 0.004 43.167 43.169 qs_energies 11 4.9 0.001 0.001 41.213 41.217 scf_env_do_scf 11 5.9 0.001 0.001 34.193 34.193 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.485 26.486 velocity_verlet 10 3.0 0.001 0.001 22.702 22.708 dbcsr_multiply_generic 2286 12.5 0.100 0.102 20.121 20.356 qs_scf_new_mos 108 7.5 0.001 0.001 17.701 17.796 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.700 17.795 multiply_cannon 2286 13.5 0.303 0.308 16.273 17.230 ot_scf_mini 108 9.5 0.002 0.002 16.623 16.721 multiply_cannon_loop 2286 14.5 0.344 0.351 15.002 15.987 ot_mini 108 10.5 0.001 0.001 10.739 10.853 multiply_cannon_multrec 9144 15.5 3.336 4.684 9.515 10.577 qs_ot_get_derivative 108 11.5 0.001 0.001 8.696 8.796 dbcsr_mm_accdrv_process 12550 15.8 4.680 7.410 6.054 8.328 init_scf_loop 11 6.9 0.000 0.000 7.680 7.681 rebuild_ks_matrix 119 8.3 0.000 0.000 6.779 6.917 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.778 6.917 prepare_preconditioner 11 7.9 0.000 0.000 6.769 6.782 make_preconditioner 11 8.9 0.000 0.000 6.769 6.782 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.336 6.656 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.127 6.253 mp_waitall_1 94719 16.7 3.317 5.338 3.317 5.338 init_scf_run 11 5.9 0.000 0.001 4.873 4.873 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.872 4.873 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.440 4.855 cp_fm_upper_to_full 72 14.2 3.140 4.512 3.140 4.512 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 2.274 4.202 jit_kernel_multiply 10 15.4 1.347 3.771 1.347 3.771 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.615 3.638 calculate_rho_elec 119 8.7 0.118 0.121 3.614 3.637 sum_up_and_integrate 119 10.3 0.064 0.065 3.627 3.633 integrate_v_rspace 119 11.3 0.003 0.003 3.562 3.569 calculate_first_density_matrix 1 7.0 0.000 0.000 3.468 3.473 qs_ot_get_p 119 10.4 0.001 0.001 3.152 3.285 calculate_dm_sparse 119 9.5 0.000 0.000 2.925 2.946 dbcsr_complete_redistribute 329 12.2 0.289 0.296 1.980 2.765 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.196 2.479 apply_single 119 13.6 0.000 0.000 2.195 2.479 make_m2s 4572 13.5 0.038 0.038 2.288 2.472 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.663 2.448 make_images 4572 14.5 0.353 0.383 2.167 2.351 mp_alltoall_i22 627 13.8 1.399 2.226 1.399 2.226 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.428 2.206 grid_integrate_task_list 119 12.3 2.000 2.025 2.000 2.025 ot_diis_step 108 11.5 0.014 0.014 2.017 2.017 pw_transfer 1439 11.6 0.066 0.067 2.004 2.007 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.994 2.005 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.912 1.913 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.908 1.911 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.878 1.878 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.866 1.866 density_rs2pw 119 9.7 0.003 0.003 1.728 1.750 mp_sum_l 7207 12.9 1.034 1.740 1.034 1.740 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.681 1.731 acc_transpose_blocks 9144 15.5 0.038 0.039 1.643 1.670 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.621 1.665 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.639 1.639 fft_wrap_pw1pw2_140 487 13.2 0.089 0.089 1.602 1.605 fft3d_ps 1201 14.6 0.567 0.576 1.540 1.545 grid_collocate_task_list 119 9.7 1.444 1.464 1.444 1.464 cp_fm_cholesky_invert 11 10.9 1.439 1.442 1.439 1.442 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 1.394 1.417 rs_pw_transfer 974 11.9 0.009 0.009 1.364 1.406 hybrid_alltoall_any 4725 16.4 0.087 0.149 1.120 1.380 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.348 1.348 cp_fm_diag_elpa_base 50 14.0 1.207 1.257 1.346 1.346 wfi_extrapolate 11 7.9 0.001 0.001 1.339 1.339 make_images_data 4572 15.5 0.043 0.046 1.073 1.316 jit_kernel_transpose 5 15.6 1.276 1.299 1.276 1.299 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.165 1.188 mp_alltoall_d11v 2130 13.8 1.159 1.187 1.159 1.187 potential_pw2rs 119 12.3 0.014 0.015 1.171 1.172 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.136 1.156 multiply_cannon_sync_h2d 9144 15.5 1.039 1.044 1.039 1.044 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.989 1.040 qs_create_task_list 11 7.9 0.007 0.007 0.950 0.961 generate_qs_task_list 11 8.9 0.372 0.393 0.943 0.955 make_basis_sm 11 9.8 0.364 0.368 0.879 0.879 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.863 0.876 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.669000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=712.363636, yerr=18.227216 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_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 503.537664E+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 1452934. 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 0.074 0.166 84.438 84.439 qs_mol_dyn_low 1 2.0 0.004 0.006 83.872 83.881 qs_forces 11 3.9 0.002 0.004 83.800 83.801 qs_energies 11 4.9 0.002 0.007 80.934 80.954 scf_env_do_scf 11 5.9 0.001 0.002 71.131 71.133 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 65.482 65.483 dbcsr_multiply_generic 2055 12.4 0.106 0.111 52.877 53.122 qs_scf_new_mos 99 7.5 0.000 0.001 48.546 48.658 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.545 48.657 ot_scf_mini 99 9.5 0.002 0.003 46.149 46.238 multiply_cannon 2055 13.4 0.179 0.188 43.038 43.876 multiply_cannon_loop 2055 14.4 1.554 1.596 42.084 42.965 velocity_verlet 10 3.0 0.001 0.002 42.308 42.309 ot_mini 99 10.5 0.001 0.001 27.825 27.927 qs_ot_get_derivative 99 11.5 0.002 0.019 21.014 21.098 multiply_cannon_multrec 49320 15.4 12.084 12.870 17.454 18.246 rebuild_ks_matrix 110 8.3 0.000 0.001 14.477 14.631 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.477 14.630 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.708 12.842 mp_waitall_1 220248 16.4 11.699 12.680 11.699 12.680 multiply_cannon_sync_h2d 49320 15.4 9.911 10.683 9.911 10.683 qs_ot_get_p 110 10.4 0.001 0.001 9.696 9.807 multiply_cannon_metrocomm3 49320 15.4 0.079 0.085 7.044 8.397 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.166 7.719 init_scf_run 11 5.9 0.000 0.001 7.672 7.673 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.672 7.672 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.138 7.530 apply_single 110 13.6 0.000 0.001 7.138 7.530 sum_up_and_integrate 110 10.3 0.036 0.043 7.082 7.094 integrate_v_rspace 110 11.3 0.002 0.003 7.045 7.066 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.797 6.851 ot_diis_step 99 11.5 0.005 0.006 6.520 6.520 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.421 6.452 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.213 6.338 calculate_rho_elec 110 8.6 0.020 0.024 6.212 6.337 mp_sum_l 6514 12.8 5.435 6.178 5.435 6.178 init_scf_loop 11 6.9 0.005 0.022 5.619 5.621 dbcsr_mm_accdrv_process 87628 16.1 2.086 2.181 5.249 5.539 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.400 5.401 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.927 4.928 cp_fm_redistribute_end 48 14.0 4.300 4.901 4.303 4.902 cp_fm_diag_elpa_base 48 14.0 0.593 4.756 0.596 4.783 calculate_dm_sparse 110 9.5 0.001 0.001 4.088 4.183 make_m2s 4110 13.4 0.061 0.066 3.980 4.103 rs_pw_transfer 902 11.9 0.012 0.013 3.875 4.079 wfi_extrapolate 11 7.9 0.001 0.001 4.016 4.016 make_images 4110 14.4 0.178 0.192 3.885 4.011 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.814 3.819 multiply_cannon_metrocomm1 49320 15.4 0.060 0.064 2.661 3.791 calculate_first_density_matrix 1 7.0 0.001 0.003 3.565 3.573 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.484 3.521 density_rs2pw 110 9.6 0.004 0.004 3.317 3.481 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.403 3.475 prepare_preconditioner 11 7.9 0.000 0.000 3.409 3.438 make_preconditioner 11 8.9 0.000 0.001 3.409 3.438 grid_integrate_task_list 110 12.3 3.272 3.427 3.272 3.427 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.179 3.215 pw_transfer 1331 11.6 0.054 0.066 2.876 2.946 jit_kernel_multiply 13 15.9 2.876 2.909 2.876 2.909 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.789 2.862 potential_pw2rs 110 12.3 0.006 0.007 2.669 2.696 mp_alltoall_d11v 2046 13.8 2.077 2.538 2.077 2.538 acc_transpose_blocks 49320 15.4 0.226 0.235 2.386 2.471 fft_wrap_pw1pw2_140 451 13.1 0.168 0.188 2.332 2.402 fft3d_ps 1111 14.6 0.752 0.840 2.328 2.388 grid_collocate_task_list 110 9.6 2.088 2.201 2.088 2.201 mp_waitany 14300 13.8 1.850 2.154 1.850 2.154 mp_sum_d 3879 11.9 1.504 2.036 1.504 2.036 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.978 2.003 make_images_data 4110 15.4 0.043 0.046 1.814 1.967 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.910 1.933 hybrid_alltoall_any 4261 16.3 0.082 0.481 1.566 1.864 cp_fm_cholesky_invert 11 10.9 1.823 1.826 1.823 1.826 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.439000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.545455, yerr=3.143930 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_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 586.813440E+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 3507907. 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 0.099 0.162 71.076 71.077 qs_mol_dyn_low 1 2.0 0.003 0.004 70.263 70.274 qs_forces 11 3.9 0.002 0.003 70.185 70.187 qs_energies 11 4.9 0.004 0.007 66.889 66.893 scf_env_do_scf 11 5.9 0.001 0.001 56.866 56.869 scf_env_do_scf_inner_loop 99 6.5 0.019 0.037 49.114 49.116 dbcsr_multiply_generic 2055 12.4 0.112 0.119 39.509 39.687 velocity_verlet 10 3.0 0.001 0.001 36.105 36.107 multiply_cannon 2055 13.4 0.225 0.256 32.909 33.977 qs_scf_new_mos 99 7.5 0.001 0.001 33.332 33.481 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.331 33.481 multiply_cannon_loop 2055 14.4 0.929 0.951 31.559 32.362 ot_scf_mini 99 9.5 0.003 0.004 31.673 31.811 ot_mini 99 10.5 0.001 0.001 18.909 19.052 multiply_cannon_multrec 24660 15.4 7.599 9.608 14.332 16.186 rebuild_ks_matrix 110 8.3 0.000 0.001 13.565 13.661 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.564 13.660 qs_ot_get_derivative 99 11.5 0.001 0.001 13.083 13.225 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.965 12.047 mp_waitall_1 176588 16.5 7.971 10.554 7.971 10.554 multiply_cannon_metrocomm3 24660 15.4 0.070 0.073 5.489 8.347 multiply_cannon_sync_h2d 24660 15.4 6.981 8.137 6.981 8.137 init_scf_loop 11 6.9 0.001 0.004 7.709 7.712 init_scf_run 11 5.9 0.000 0.001 7.586 7.587 scf_env_initial_rho_setup 11 6.9 0.009 0.047 7.586 7.587 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.514 7.213 apply_single 110 13.6 0.000 0.001 6.514 7.213 dbcsr_mm_accdrv_process 52282 16.1 5.058 6.003 6.573 7.121 sum_up_and_integrate 110 10.3 0.053 0.059 6.503 6.515 integrate_v_rspace 110 11.3 0.002 0.002 6.450 6.461 qs_ot_get_p 110 10.4 0.001 0.002 6.174 6.317 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.109 5.833 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.751 5.777 ot_diis_step 99 11.5 0.010 0.010 5.777 5.777 calculate_rho_elec 110 8.6 0.039 0.047 5.751 5.777 prepare_preconditioner 11 7.9 0.000 0.000 5.650 5.669 make_preconditioner 11 8.9 0.000 0.001 5.650 5.669 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.248 5.406 make_m2s 4110 13.4 0.057 0.062 4.186 4.690 make_images 4110 14.4 0.401 0.446 4.077 4.579 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.250 4.270 calculate_dm_sparse 110 9.5 0.001 0.001 4.241 4.270 calculate_first_density_matrix 1 7.0 0.011 0.028 3.960 3.970 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.814 3.814 wfi_extrapolate 11 7.9 0.001 0.001 3.529 3.529 pw_transfer 1331 11.6 0.065 0.069 3.312 3.455 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.206 3.348 grid_integrate_task_list 110 12.3 3.161 3.309 3.161 3.309 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.292 3.294 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.267 3.268 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.183 3.259 cp_fm_redistribute_end 48 14.0 2.444 3.242 2.445 3.242 density_rs2pw 110 9.6 0.004 0.004 3.053 3.221 cp_fm_diag_elpa_base 48 14.0 0.763 3.092 0.794 3.186 rs_pw_transfer 902 11.9 0.012 0.013 2.941 3.096 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.938 2.984 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 2.815 2.825 make_images_data 4110 15.4 0.047 0.051 2.319 2.821 hybrid_alltoall_any 4261 16.3 0.102 0.436 2.053 2.817 fft_wrap_pw1pw2_140 451 13.1 0.200 0.218 2.663 2.800 fft3d_ps 1111 14.6 1.074 1.270 2.563 2.690 cp_fm_cholesky_invert 11 10.9 2.622 2.629 2.622 2.629 acc_transpose_blocks 24660 15.4 0.111 0.114 2.566 2.594 jit_kernel_multiply 11 16.3 1.157 2.561 1.157 2.561 mp_sum_l 6514 12.8 1.821 2.416 1.821 2.416 potential_pw2rs 110 12.3 0.008 0.009 2.391 2.405 grid_collocate_task_list 110 9.6 2.122 2.247 2.122 2.247 mp_alltoall_d11v 2046 13.8 1.767 2.036 1.767 2.036 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.950 1.971 acc_transpose_blocks_kernels 24660 16.4 0.309 0.317 1.948 1.969 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.912 1.913 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.765 1.776 mp_allgather_i34 2055 14.4 0.677 1.656 0.677 1.656 jit_kernel_transpose 5 15.6 1.639 1.653 1.639 1.653 multiply_cannon_metrocomm4 22605 15.4 0.073 0.076 0.799 1.609 mp_irecv_dv 57340 16.2 0.675 1.492 0.675 1.492 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.374 1.487 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.077000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.272727, yerr=6.579457 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_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 660.664320E+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 0.039 0.073 60.581 60.582 qs_mol_dyn_low 1 2.0 0.003 0.004 60.216 60.224 qs_forces 11 3.9 0.002 0.003 60.151 60.152 qs_energies 11 4.9 0.011 0.033 56.977 56.981 scf_env_do_scf 11 5.9 0.001 0.001 47.508 47.509 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.069 39.070 velocity_verlet 10 3.0 0.001 0.001 32.034 32.036 dbcsr_multiply_generic 2055 12.4 0.106 0.108 28.598 28.821 qs_scf_new_mos 99 7.5 0.001 0.001 24.734 24.855 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.734 24.855 ot_scf_mini 99 9.5 0.003 0.003 23.522 23.654 multiply_cannon 2055 13.4 0.213 0.222 22.445 23.607 multiply_cannon_loop 2055 14.4 0.618 0.634 21.296 22.397 ot_mini 99 10.5 0.001 0.001 13.387 13.515 rebuild_ks_matrix 110 8.3 0.000 0.001 12.061 12.188 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.061 12.187 multiply_cannon_multrec 16440 15.4 3.954 4.976 9.934 11.152 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.641 10.754 mp_waitall_1 139946 16.5 6.910 10.053 6.910 10.053 qs_ot_get_derivative 99 11.5 0.001 0.001 8.969 9.096 init_scf_loop 11 6.9 0.001 0.003 8.403 8.404 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.228 7.043 prepare_preconditioner 11 7.9 0.000 0.000 6.660 6.678 make_preconditioner 11 8.9 0.000 0.001 6.660 6.678 sum_up_and_integrate 110 10.3 0.061 0.062 6.362 6.377 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.022 6.366 init_scf_run 11 5.9 0.000 0.001 6.317 6.319 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.317 6.319 integrate_v_rspace 110 11.3 0.002 0.003 6.301 6.317 dbcsr_mm_accdrv_process 34862 16.1 4.747 5.532 5.833 6.136 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.760 5.772 calculate_rho_elec 110 8.6 0.058 0.058 5.760 5.771 qs_ot_get_p 110 10.4 0.001 0.001 5.380 5.518 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.936 5.341 apply_single 110 13.6 0.000 0.000 4.936 5.341 make_m2s 4110 13.4 0.050 0.052 4.084 4.427 ot_diis_step 99 11.5 0.010 0.011 4.390 4.390 multiply_cannon_sync_h2d 16440 15.4 3.702 4.346 3.702 4.346 make_images 4110 14.4 0.394 0.511 3.969 4.309 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.808 3.811 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.072 3.705 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.473 3.474 grid_integrate_task_list 110 12.3 3.186 3.408 3.186 3.408 pw_transfer 1331 11.6 0.064 0.071 3.158 3.164 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.052 3.060 calculate_dm_sparse 110 9.5 0.001 0.001 2.983 3.011 calculate_first_density_matrix 1 7.0 0.001 0.003 2.996 2.999 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.960 2.962 wfi_extrapolate 11 7.9 0.001 0.001 2.947 2.947 cp_fm_redistribute_end 48 14.0 1.846 2.932 1.848 2.933 density_rs2pw 110 9.6 0.004 0.004 2.758 2.917 cp_fm_diag_elpa_base 48 14.0 1.022 2.783 1.080 2.889 rs_pw_transfer 902 11.9 0.011 0.011 2.637 2.832 hybrid_alltoall_any 4261 16.3 0.105 0.373 2.083 2.807 make_images_data 4110 15.4 0.044 0.048 2.342 2.797 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.793 2.795 cp_fm_cholesky_invert 11 10.9 2.649 2.656 2.649 2.656 fft_wrap_pw1pw2_140 451 13.1 0.210 0.214 2.606 2.615 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.466 2.512 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.317 2.393 fft3d_ps 1111 14.6 1.063 1.076 2.368 2.379 grid_collocate_task_list 110 9.6 2.176 2.370 2.176 2.370 mp_alltoall_d11v 2046 13.8 2.029 2.356 2.029 2.356 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.879 2.289 potential_pw2rs 110 12.3 0.010 0.010 2.198 2.207 mp_irecv_dv 48980 15.7 0.808 2.168 0.808 2.168 mp_sum_l 6514 12.8 1.459 2.032 1.459 2.032 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.018 2.019 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.956 1.969 jit_kernel_multiply 8 16.6 0.689 1.887 0.689 1.887 dbcsr_complete_redistribute 325 12.2 0.317 0.352 1.374 1.834 cp_fm_upper_to_full 70 13.6 1.378 1.832 1.378 1.832 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.620 1.634 cp_fm_cholesky_decompose 22 10.9 1.544 1.563 1.544 1.563 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.370 1.506 mp_allgather_i34 2055 14.4 0.473 1.475 0.473 1.475 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.936 1.378 acc_transpose_blocks 16440 15.4 0.072 0.074 1.341 1.362 mp_waitany 17072 13.8 1.147 1.332 1.147 1.332 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.237 1.249 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.582000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=626.545455, yerr=8.814901 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_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 724.381696E+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 0.048 0.134 68.268 68.270 qs_mol_dyn_low 1 2.0 0.005 0.038 67.439 67.449 qs_forces 11 3.9 0.029 0.037 66.690 66.692 qs_energies 11 4.9 0.005 0.033 63.296 63.325 scf_env_do_scf 11 5.9 0.001 0.001 53.740 53.743 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.090 42.090 velocity_verlet 10 3.0 0.007 0.053 37.313 37.341 dbcsr_multiply_generic 2055 12.4 0.114 0.117 31.137 31.433 qs_scf_new_mos 99 7.5 0.001 0.001 27.820 27.924 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.819 27.923 ot_scf_mini 99 9.5 0.003 0.003 26.202 26.305 multiply_cannon 2055 13.4 0.244 0.264 24.082 25.452 multiply_cannon_loop 2055 14.4 0.889 0.910 22.723 23.416 ot_mini 99 10.5 0.001 0.001 15.661 15.791 multiply_cannon_multrec 24660 15.4 4.223 6.783 13.599 15.279 rebuild_ks_matrix 110 8.3 0.000 0.000 11.882 12.007 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 11.882 12.007 qs_ot_get_derivative 99 11.5 0.001 0.001 11.502 11.612 init_scf_loop 11 6.9 0.004 0.017 11.610 11.611 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.531 10.646 dbcsr_mm_accdrv_process 52304 16.0 7.104 9.563 9.226 10.590 prepare_preconditioner 11 7.9 0.000 0.000 9.822 9.844 make_preconditioner 11 8.9 0.000 0.002 9.822 9.844 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.075 9.515 mp_waitall_1 121746 16.5 4.966 7.399 4.966 7.399 sum_up_and_integrate 110 10.3 0.068 0.076 6.293 6.308 integrate_v_rspace 110 11.3 0.002 0.003 6.224 6.242 init_scf_run 11 5.9 0.000 0.001 5.914 5.915 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.914 5.915 make_m2s 4110 13.4 0.060 0.061 5.446 5.732 qs_ot_get_p 110 10.4 0.001 0.001 5.552 5.710 make_images 4110 14.4 0.578 0.698 5.305 5.587 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.550 5.558 calculate_rho_elec 110 8.6 0.077 0.080 5.550 5.557 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.706 4.757 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 2.058 4.684 cp_fm_upper_to_full 70 13.8 3.237 4.620 3.237 4.620 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.023 4.128 apply_single 110 13.6 0.000 0.000 4.023 4.128 ot_diis_step 99 11.5 0.011 0.011 4.123 4.124 dbcsr_complete_redistribute 325 12.2 0.416 0.472 2.656 3.778 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.737 3.752 grid_integrate_task_list 110 12.3 3.250 3.425 3.250 3.425 multiply_cannon_sync_h2d 24660 15.4 3.177 3.374 3.177 3.374 jit_kernel_multiply 12 15.7 1.786 3.280 1.786 3.280 pw_transfer 1331 11.6 0.064 0.072 3.237 3.262 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.261 3.261 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.115 3.219 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.131 3.160 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.011 3.075 make_images_data 4110 15.4 0.047 0.051 2.749 3.034 calculate_dm_sparse 110 9.5 0.001 0.001 2.962 2.987 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.977 2.978 wfi_extrapolate 11 7.9 0.001 0.001 2.970 2.970 hybrid_alltoall_any 4261 16.3 0.120 0.458 2.289 2.956 density_rs2pw 110 9.6 0.004 0.004 2.728 2.926 calculate_first_density_matrix 1 7.0 0.000 0.003 2.856 2.860 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.738 2.826 mp_alltoall_i22 605 13.7 1.640 2.808 1.640 2.808 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.723 2.725 rs_pw_transfer 902 11.9 0.010 0.011 2.446 2.725 cp_fm_cholesky_invert 11 10.9 2.713 2.721 2.713 2.721 cp_fm_redistribute_end 48 14.0 1.365 2.699 1.366 2.700 fft_wrap_pw1pw2_140 451 13.1 0.202 0.214 2.656 2.689 cp_fm_diag_elpa_base 48 14.0 1.253 2.560 1.331 2.670 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.599 2.637 qs_energies_init_hamiltonians 11 5.9 0.008 0.024 2.516 2.545 fft3d_ps 1111 14.6 1.064 1.097 2.435 2.454 grid_collocate_task_list 110 9.6 2.217 2.356 2.217 2.356 potential_pw2rs 110 12.3 0.012 0.013 2.076 2.089 mp_alltoall_d11v 2046 13.8 1.775 2.063 1.775 2.063 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.692 1.727 cp_fm_cholesky_decompose 22 10.9 1.663 1.723 1.663 1.723 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.604 1.713 acc_transpose_blocks 24660 15.4 0.107 0.110 1.634 1.654 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.607 1.620 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.581 1.595 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.855 1.576 mp_allgather_i34 2055 14.4 0.462 1.561 0.462 1.561 mp_irecv_dv 62702 16.1 0.756 1.496 0.756 1.496 mp_sum_l 6514 12.8 0.969 1.486 0.969 1.486 qs_env_update_s_mstruct 11 6.9 0.017 0.138 1.320 1.415 mp_waitany 13376 13.8 1.115 1.377 1.115 1.377 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=68.270000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=687.000000, yerr=7.942979 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_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 829.599744E+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 0.405 0.413 56.150 56.151 qs_mol_dyn_low 1 2.0 0.003 0.003 55.072 55.081 qs_forces 11 3.9 0.002 0.002 55.003 55.003 qs_energies 11 4.9 0.001 0.001 51.397 51.402 scf_env_do_scf 11 5.9 0.000 0.001 42.687 42.687 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 34.961 34.962 velocity_verlet 10 3.0 0.001 0.001 30.577 30.580 dbcsr_multiply_generic 2055 12.4 0.105 0.107 23.686 23.820 qs_scf_new_mos 99 7.5 0.001 0.001 20.916 20.953 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.915 20.953 ot_scf_mini 99 9.5 0.002 0.002 19.679 19.702 multiply_cannon 2055 13.4 0.246 0.266 18.168 19.565 multiply_cannon_loop 2055 14.4 0.322 0.334 16.805 17.117 rebuild_ks_matrix 110 8.3 0.000 0.000 11.421 11.445 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.420 11.445 ot_mini 99 10.5 0.001 0.001 10.783 10.801 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.181 10.201 multiply_cannon_multrec 8220 15.4 3.237 4.453 8.155 9.052 init_scf_loop 11 6.9 0.000 0.000 7.680 7.681 mp_waitall_1 103326 16.6 6.006 7.535 6.006 7.535 qs_ot_get_derivative 99 11.5 0.001 0.001 6.997 7.020 sum_up_and_integrate 110 10.3 0.080 0.082 6.149 6.160 integrate_v_rspace 110 11.3 0.003 0.003 6.069 6.080 prepare_preconditioner 11 7.9 0.000 0.000 6.034 6.037 make_preconditioner 11 8.9 0.000 0.000 6.034 6.037 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.614 5.686 dbcsr_mm_accdrv_process 17442 15.9 2.996 3.906 4.790 5.643 init_scf_run 11 5.9 0.000 0.001 5.588 5.588 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.587 5.587 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.572 5.585 calculate_rho_elec 110 8.6 0.115 0.115 5.572 5.584 qs_ot_get_p 110 10.4 0.001 0.001 4.978 5.002 make_m2s 4110 13.4 0.038 0.040 4.250 4.542 make_images 4110 14.4 0.637 0.688 4.120 4.412 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.959 4.364 ot_diis_step 99 11.5 0.012 0.012 3.765 3.766 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.679 3.727 apply_single 110 13.6 0.000 0.000 3.679 3.727 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.598 3.602 grid_integrate_task_list 110 12.3 3.370 3.457 3.370 3.457 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.299 3.299 pw_transfer 1331 11.6 0.065 0.069 3.242 3.251 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.136 3.146 multiply_cannon_sync_h2d 8220 15.4 2.919 3.035 2.919 3.035 cp_fm_cholesky_invert 11 10.9 2.906 2.909 2.906 2.909 make_images_data 4110 15.4 0.039 0.044 2.425 2.905 hybrid_alltoall_any 4261 16.3 0.200 0.864 2.356 2.786 calculate_first_density_matrix 1 7.0 0.000 0.000 2.776 2.777 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.771 2.773 density_rs2pw 110 9.6 0.004 0.004 2.629 2.769 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.760 2.762 cp_fm_redistribute_end 48 14.0 0.707 2.745 0.712 2.746 calculate_dm_sparse 110 9.5 0.001 0.001 2.693 2.730 wfi_extrapolate 11 7.9 0.001 0.001 2.725 2.725 cp_fm_diag_elpa_base 48 14.0 1.854 2.556 2.025 2.711 fft_wrap_pw1pw2_140 451 13.1 0.212 0.214 2.674 2.692 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.662 2.663 grid_collocate_task_list 110 9.6 2.328 2.448 2.328 2.448 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.405 2.414 fft3d_ps 1111 14.6 1.114 1.164 2.393 2.406 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.374 2.390 rs_pw_transfer 902 11.9 0.010 0.010 2.133 2.288 jit_kernel_multiply 9 15.6 1.479 2.253 1.479 2.253 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.777 1.991 potential_pw2rs 110 12.3 0.015 0.015 1.900 1.905 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.783 1.794 mp_alltoall_d11v 2046 13.8 1.637 1.747 1.637 1.747 cp_fm_cholesky_decompose 22 10.9 1.676 1.687 1.676 1.687 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.658 1.670 mp_allgather_i34 2055 14.4 0.550 1.652 0.550 1.652 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.504 1.620 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.572 1.577 dbcsr_complete_redistribute 325 12.2 0.561 0.590 1.445 1.532 qs_create_task_list 11 7.9 0.000 0.001 1.226 1.325 generate_qs_task_list 11 8.9 0.379 0.447 1.226 1.325 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.779 1.272 mp_waitany 9240 13.8 1.035 1.207 1.035 1.207 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.148 1.165 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.151000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=780.909091, yerr=11.595097 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_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 1.356812E+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 0.023 0.089 85.835 85.836 qs_mol_dyn_low 1 2.0 0.003 0.003 85.413 85.422 qs_forces 11 3.9 0.002 0.002 85.350 85.350 qs_energies 11 4.9 0.002 0.003 81.327 81.328 scf_env_do_scf 11 5.9 0.001 0.001 70.509 70.510 velocity_verlet 10 3.0 0.001 0.001 54.361 54.367 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.435 42.438 dbcsr_multiply_generic 2055 12.4 0.119 0.121 29.241 29.338 init_scf_loop 11 6.9 0.000 0.000 28.003 28.009 qs_scf_new_mos 99 7.5 0.001 0.001 26.267 26.322 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.266 26.321 prepare_preconditioner 11 7.9 0.000 0.000 26.022 26.030 make_preconditioner 11 8.9 0.000 0.000 26.022 26.030 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.416 25.502 ot_scf_mini 99 9.5 0.002 0.002 24.513 24.551 multiply_cannon 2055 13.4 0.339 0.362 22.280 23.042 multiply_cannon_loop 2055 14.4 0.342 0.346 20.493 20.838 cp_fm_upper_to_full 70 14.2 12.574 17.893 12.574 17.893 ot_mini 99 10.5 0.001 0.001 13.768 13.802 rebuild_ks_matrix 110 8.3 0.001 0.001 13.008 13.054 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.008 13.053 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.825 11.867 dbcsr_complete_redistribute 325 12.2 1.029 1.044 7.182 10.337 multiply_cannon_multrec 8220 15.4 4.335 4.529 9.993 10.264 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.161 9.309 qs_ot_get_derivative 99 11.5 0.001 0.001 9.213 9.252 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.590 8.717 mp_waitall_1 84994 16.7 7.598 8.609 7.598 8.609 mp_alltoall_i22 605 13.7 5.213 8.347 5.213 8.347 init_scf_run 11 5.9 0.000 0.001 6.576 6.576 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.575 6.576 sum_up_and_integrate 110 10.3 0.150 0.152 6.552 6.568 integrate_v_rspace 110 11.3 0.003 0.004 6.401 6.416 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.361 6.396 calculate_rho_elec 110 8.6 0.227 0.227 6.361 6.396 dbcsr_mm_accdrv_process 11614 15.7 3.247 3.533 5.517 5.917 make_m2s 4110 13.4 0.043 0.044 5.208 5.708 qs_ot_get_p 110 10.4 0.001 0.001 5.602 5.651 make_images 4110 14.4 0.880 0.929 5.020 5.520 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.933 5.444 cp_fm_cholesky_invert 11 10.9 5.292 5.296 5.292 5.296 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.667 5.156 apply_single 110 13.6 0.000 0.000 4.667 5.155 ot_diis_step 99 11.5 0.015 0.015 4.531 4.531 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.999 4.006 multiply_cannon_sync_h2d 8220 15.4 3.950 3.960 3.950 3.960 grid_integrate_task_list 110 12.3 3.652 3.705 3.652 3.705 pw_transfer 1331 11.6 0.073 0.074 3.637 3.643 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.641 3.642 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.157 3.609 hybrid_alltoall_any 4261 16.3 0.256 0.550 2.872 3.578 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.561 3.562 calculate_dm_sparse 110 9.5 0.001 0.001 3.507 3.539 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.521 3.528 make_images_data 4110 15.4 0.043 0.045 2.799 3.458 calculate_first_density_matrix 1 7.0 0.000 0.000 3.231 3.236 wfi_extrapolate 11 7.9 0.001 0.001 3.231 3.231 fft_wrap_pw1pw2_140 451 13.1 0.215 0.215 3.059 3.064 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.050 3.054 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.975 2.975 cp_fm_diag_elpa_base 48 14.0 2.451 2.634 2.973 2.973 density_rs2pw 110 9.6 0.004 0.004 2.873 2.887 fft3d_ps 1111 14.6 1.263 1.271 2.733 2.738 grid_collocate_task_list 110 9.6 2.623 2.645 2.623 2.645 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.554 2.567 jit_kernel_multiply 10 15.4 2.066 2.433 2.066 2.433 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.368 2.392 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.176 2.233 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.084 2.175 rs_pw_transfer 902 11.9 0.011 0.011 2.099 2.143 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.048 2.068 cp_fm_cholesky_decompose 22 10.9 2.030 2.049 2.030 2.049 mp_alltoall_d11v 2046 13.8 1.981 2.033 1.981 2.033 potential_pw2rs 110 12.3 0.021 0.022 2.022 2.026 qs_create_task_list 11 7.9 0.000 0.001 1.901 1.949 generate_qs_task_list 11 8.9 0.739 0.793 1.901 1.949 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.787 1.791 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.742 1.776 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=85.836000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1211.909091, yerr=66.147843 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 627.937280E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56513. MP_Allreduce 11298 784. MP_Sync 170 MP_Alltoall 2226 2699530. 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 0.013 0.029 202.499 202.501 qs_mol_dyn_low 1 2.0 0.003 0.004 202.129 202.143 qs_forces 11 3.9 0.003 0.004 202.044 202.045 qs_energies 11 4.9 0.001 0.002 196.547 196.562 scf_env_do_scf 11 5.9 0.001 0.001 179.873 179.877 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 159.217 159.218 dbcsr_multiply_generic 2507 12.6 0.180 0.195 124.207 125.073 velocity_verlet 10 3.0 0.001 0.001 121.943 121.944 qs_scf_new_mos 117 7.6 0.001 0.001 120.702 120.995 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.702 120.995 ot_scf_mini 117 9.6 0.003 0.003 114.096 114.360 multiply_cannon 2507 13.6 0.236 0.246 101.084 103.064 multiply_cannon_loop 2507 14.6 2.112 2.182 98.877 100.491 ot_mini 117 10.6 0.001 0.001 65.452 65.713 multiply_cannon_multrec 60168 15.6 33.203 35.681 41.606 43.592 qs_ot_get_derivative 117 11.6 0.001 0.001 40.442 40.689 rebuild_ks_matrix 128 8.3 0.001 0.001 33.183 33.647 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.183 33.647 mp_waitall_1 267128 16.5 28.214 31.014 28.214 31.014 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.872 30.295 multiply_cannon_sync_h2d 60168 15.6 27.324 29.393 27.324 29.393 qs_ot_get_p 128 10.4 0.001 0.001 26.739 27.049 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.417 25.339 apply_single 128 13.6 0.001 0.001 24.416 25.339 ot_diis_step 117 11.6 0.007 0.008 24.642 24.643 init_scf_loop 11 6.9 0.000 0.000 20.580 20.581 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.043 20.137 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.741 18.932 multiply_cannon_metrocomm3 60168 15.6 0.115 0.119 15.668 17.694 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.467 17.469 prepare_preconditioner 11 7.9 0.000 0.000 15.975 16.024 make_preconditioner 11 8.9 0.000 0.000 15.975 16.024 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.203 15.369 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.646 14.650 cp_fm_redistribute_end 83 14.4 11.581 14.573 11.592 14.576 cp_fm_diag_elpa_base 83 14.4 2.942 14.195 2.973 14.320 make_m2s 5014 13.6 0.104 0.113 13.818 14.208 make_images 5014 14.6 0.401 0.418 13.636 14.037 sum_up_and_integrate 128 10.3 0.089 0.105 13.897 13.925 integrate_v_rspace 128 11.3 0.003 0.004 13.808 13.840 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.543 12.681 calculate_rho_elec 128 8.7 0.045 0.063 12.543 12.680 init_scf_run 11 5.9 0.000 0.001 12.514 12.514 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.514 12.514 mp_sum_l 7870 13.0 8.035 9.185 8.035 9.185 cp_fm_cholesky_invert 11 10.9 9.106 9.113 9.106 9.113 wfi_extrapolate 11 7.9 0.001 0.001 8.998 8.999 calculate_dm_sparse 128 9.5 0.001 0.001 8.716 8.804 dbcsr_mm_accdrv_process 124484 16.2 3.289 3.440 7.965 8.516 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.904 8.053 multiply_cannon_metrocomm1 60168 15.6 0.089 0.094 6.034 7.948 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.852 7.936 make_images_data 5014 15.6 0.068 0.073 6.773 7.715 grid_integrate_task_list 128 12.3 7.069 7.505 7.069 7.505 hybrid_alltoall_any 5200 16.5 0.289 2.242 5.940 7.323 pw_transfer 1547 11.6 0.074 0.108 6.807 7.080 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.604 6.849 density_rs2pw 128 9.7 0.006 0.007 6.362 6.839 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.636 6.644 rs_pw_transfer 1046 11.9 0.017 0.018 5.553 5.958 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.795 5.870 fft_wrap_pw1pw2_140 523 13.2 0.445 0.511 5.692 5.865 fft3d_ps 1291 14.7 2.094 2.568 5.398 5.587 mp_alltoall_d11v 2415 14.1 4.233 5.431 4.233 5.431 grid_collocate_task_list 128 9.7 4.722 5.114 4.722 5.114 cp_fm_cholesky_decompose 22 10.9 4.661 4.675 4.661 4.675 mp_sum_d 4469 12.2 3.641 4.478 3.641 4.478 potential_pw2rs 128 12.3 0.009 0.010 4.396 4.433 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=202.501000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=593.818182, yerr=6.886326 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_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 831.369216E+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 4913833. 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 0.018 0.050 194.781 194.782 qs_mol_dyn_low 1 2.0 0.004 0.007 194.375 194.389 qs_forces 11 3.9 0.003 0.006 191.330 191.330 qs_energies 11 4.9 0.001 0.002 183.795 184.654 scf_env_do_scf 11 5.9 0.001 0.001 167.195 167.205 scf_env_do_scf_inner_loop 117 6.6 0.011 0.066 134.360 134.363 velocity_verlet 10 3.0 0.001 0.002 121.697 121.760 dbcsr_multiply_generic 2507 12.6 0.185 0.190 98.351 99.577 qs_scf_new_mos 117 7.6 0.001 0.001 95.959 96.520 qs_scf_loop_do_ot 117 8.6 0.001 0.001 95.958 96.519 ot_scf_mini 117 9.6 0.095 0.732 91.152 91.806 multiply_cannon 2507 13.6 0.483 0.534 77.756 81.884 multiply_cannon_loop 2507 14.6 1.255 1.299 74.681 77.106 ot_mini 117 10.6 0.001 0.002 50.537 51.205 multiply_cannon_multrec 30084 15.6 22.007 27.265 32.127 38.030 mp_waitall_1 214728 16.6 24.499 37.201 24.499 37.201 rebuild_ks_matrix 128 8.3 0.001 0.001 32.544 33.150 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.021 32.543 33.149 init_scf_loop 11 6.9 0.000 0.000 32.748 32.749 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.346 29.905 qs_ot_get_derivative 117 11.6 0.006 0.101 28.693 29.415 prepare_preconditioner 11 7.9 0.000 0.000 28.240 28.325 make_preconditioner 11 8.9 0.000 0.000 28.240 28.325 multiply_cannon_metrocomm3 30084 15.6 0.095 0.100 15.764 28.284 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.977 27.545 qs_ot_get_p 128 10.4 0.001 0.001 22.422 23.106 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.890 22.901 apply_single 128 13.6 0.001 0.001 21.890 22.901 multiply_cannon_sync_h2d 30084 15.6 19.346 22.134 19.346 22.134 ot_diis_step 117 11.6 0.017 0.034 21.663 21.665 qs_ot_p2m_diag 83 11.4 0.188 0.216 17.390 17.424 cp_fm_cholesky_invert 11 10.9 16.491 16.504 16.491 16.504 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.235 16.237 make_m2s 5014 13.6 0.089 0.097 14.231 15.565 make_images 5014 14.6 1.164 1.348 14.021 15.355 sum_up_and_integrate 128 10.3 0.116 0.134 14.242 14.268 integrate_v_rspace 128 11.3 0.003 0.003 14.125 14.157 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.092 13.099 cp_fm_redistribute_end 83 14.4 7.637 13.011 7.649 13.013 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.888 12.928 calculate_rho_elec 128 8.7 0.088 0.105 12.887 12.928 cp_fm_diag_elpa_base 83 14.4 5.129 12.495 5.347 12.874 init_scf_run 11 5.9 0.000 0.001 11.713 11.714 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.713 11.714 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.939 11.390 multiply_cannon_metrocomm4 27577 15.6 0.095 0.109 3.749 10.802 mp_irecv_dv 69486 16.3 3.556 10.411 3.556 10.411 dbcsr_mm_accdrv_process 62242 16.2 4.931 5.809 9.577 10.299 make_images_data 5014 15.6 0.066 0.073 8.614 10.176 hybrid_alltoall_any 5200 16.5 0.343 1.519 7.324 8.931 wfi_extrapolate 11 7.9 0.001 0.002 8.345 8.345 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.988 7.821 pw_transfer 1547 11.6 0.085 0.102 7.657 7.735 grid_integrate_task_list 128 12.3 7.181 7.524 7.181 7.524 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.432 7.496 density_rs2pw 128 9.7 0.006 0.006 6.652 7.046 cp_fm_cholesky_decompose 22 10.9 6.918 7.001 6.918 7.001 calculate_dm_sparse 128 9.5 0.001 0.002 6.698 6.863 mp_sum_l 7870 13.0 4.607 6.857 4.607 6.857 fft_wrap_pw1pw2_140 523 13.2 0.469 0.515 6.498 6.557 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.175 6.185 rs_pw_transfer 1046 11.9 0.014 0.017 5.543 5.968 fft3d_ps 1291 14.7 2.782 2.960 5.804 5.838 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.300 5.467 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.333 5.397 grid_collocate_task_list 128 9.7 4.911 5.256 4.911 5.256 mp_sum_d 4455 12.2 3.538 5.091 3.538 5.091 mp_allgather_i34 2507 14.6 1.642 4.816 1.642 4.816 potential_pw2rs 128 12.3 0.014 0.016 4.759 4.777 mp_alltoall_d11v 2415 14.1 4.043 4.672 4.043 4.672 dbcsr_complete_redistribute 395 12.7 0.767 0.856 3.096 3.958 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=194.782000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=791.090909, yerr=2.644189 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 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 930.648064E+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 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57194. MP_Allreduce 11251 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 0.081 0.184 176.059 176.060 qs_mol_dyn_low 1 2.0 0.004 0.005 175.455 175.493 qs_forces 11 3.9 0.003 0.005 175.340 175.351 qs_energies 11 4.9 0.006 0.021 168.854 168.867 scf_env_do_scf 11 5.9 0.008 0.017 153.245 153.246 scf_env_do_scf_inner_loop 116 6.6 0.020 0.044 117.604 117.611 velocity_verlet 10 3.0 0.004 0.022 111.317 111.319 dbcsr_multiply_generic 2485 12.5 0.181 0.187 81.126 82.448 qs_scf_new_mos 116 7.6 0.001 0.001 81.462 81.827 qs_scf_loop_do_ot 116 8.6 0.001 0.002 81.462 81.826 ot_scf_mini 116 9.6 0.004 0.006 77.380 77.815 multiply_cannon 2485 13.5 0.498 0.523 61.491 65.143 multiply_cannon_loop 2485 14.5 0.851 0.880 58.141 61.455 ot_mini 116 10.6 0.001 0.001 42.233 42.675 init_scf_loop 11 6.9 0.233 0.469 35.534 35.535 mp_waitall_1 169034 16.6 24.249 34.099 24.249 34.099 prepare_preconditioner 11 7.9 0.000 0.000 30.919 30.979 make_preconditioner 11 8.9 0.000 0.001 30.919 30.979 rebuild_ks_matrix 127 8.3 0.001 0.001 29.761 30.359 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.022 29.761 30.359 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.561 29.970 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.857 27.405 multiply_cannon_multrec 19880 15.5 13.476 16.468 22.114 25.053 multiply_cannon_metrocomm3 19880 15.5 0.060 0.065 14.721 24.830 qs_ot_get_derivative 116 11.6 0.001 0.002 22.718 23.163 qs_ot_get_p 127 10.4 0.001 0.001 20.792 21.310 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.653 20.499 apply_single 127 13.6 0.001 0.001 19.652 20.498 ot_diis_step 116 11.6 0.018 0.021 19.415 19.415 qs_ot_p2m_diag 82 11.4 0.262 0.269 16.330 16.336 multiply_cannon_sync_h2d 19880 15.5 14.238 15.984 14.238 15.984 make_m2s 4970 13.5 0.080 0.085 14.418 15.477 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.334 15.335 make_images 4970 14.5 1.176 1.254 14.186 15.244 cp_fm_cholesky_invert 11 10.9 14.441 14.451 14.441 14.451 sum_up_and_integrate 127 10.3 0.131 0.145 14.023 14.047 integrate_v_rspace 127 11.3 0.003 0.004 13.891 13.919 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.911 12.949 calculate_rho_elec 127 8.7 0.131 0.147 12.911 12.948 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.203 12.205 cp_fm_redistribute_end 82 14.4 4.609 12.132 4.622 12.134 cp_fm_diag_elpa_base 82 14.4 7.088 11.550 7.491 12.002 init_scf_run 11 5.9 0.000 0.001 10.559 10.560 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.559 10.560 make_images_data 4970 15.5 0.061 0.070 8.820 10.292 hybrid_alltoall_any 5155 16.4 0.425 1.915 7.858 9.525 multiply_cannon_metrocomm4 17395 15.5 0.061 0.071 3.379 9.196 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.873 9.169 mp_irecv_dv 49801 16.2 3.258 8.952 3.258 8.952 dbcsr_mm_accdrv_process 41158 16.2 4.365 5.185 8.093 8.266 pw_transfer 1535 11.6 0.084 0.104 7.700 7.814 grid_integrate_task_list 127 12.3 7.274 7.738 7.274 7.738 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.479 7.598 cp_fm_upper_to_full 104 14.5 5.750 7.518 5.750 7.518 wfi_extrapolate 11 7.9 0.001 0.001 7.362 7.362 cp_fm_cholesky_decompose 22 10.9 7.307 7.327 7.307 7.327 density_rs2pw 127 9.7 0.006 0.006 6.399 6.873 dbcsr_complete_redistribute 393 12.7 1.176 1.211 4.812 6.808 fft_wrap_pw1pw2_140 519 13.2 0.474 0.527 6.551 6.685 calculate_dm_sparse 127 9.5 0.001 0.002 5.892 6.011 fft3d_ps 1281 14.7 2.677 2.908 5.760 5.817 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.627 5.625 grid_collocate_task_list 127 9.7 5.035 5.604 5.035 5.604 rs_pw_transfer 1038 11.9 0.014 0.015 4.970 5.441 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.421 5.427 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.604 5.343 mp_sum_l 7804 13.0 3.470 4.929 3.470 4.929 mp_allgather_i34 2485 14.5 1.879 4.862 1.879 4.862 mp_alltoall_d11v 2401 14.1 4.210 4.795 4.210 4.795 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.521 4.686 potential_pw2rs 127 12.3 0.020 0.022 4.468 4.490 mp_alltoall_i22 712 14.1 2.205 4.385 2.205 4.385 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.338 4.088 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.965 3.994 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.742 3.745 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.060000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=883.818182, yerr=8.663593 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.148461E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57332. MP_Allreduce 11226 1068. MP_Sync 170 MP_Alltoall 1712 12503084. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.056 188.308 188.309 qs_mol_dyn_low 1 2.0 0.004 0.005 187.899 187.911 qs_forces 11 3.9 0.009 0.055 187.769 187.789 qs_energies 11 4.9 0.003 0.008 180.727 180.738 scf_env_do_scf 11 5.9 0.011 0.029 163.736 163.746 velocity_verlet 10 3.0 0.001 0.002 123.498 123.502 scf_env_do_scf_inner_loop 117 6.6 0.098 0.262 116.850 116.861 dbcsr_multiply_generic 2507 12.6 0.190 0.197 80.899 81.575 qs_scf_new_mos 117 7.6 0.001 0.001 81.010 81.255 qs_scf_loop_do_ot 117 8.6 0.017 0.052 81.009 81.255 ot_scf_mini 117 9.6 0.021 0.053 76.494 76.816 multiply_cannon 2507 13.6 0.553 0.600 54.773 57.294 multiply_cannon_loop 2507 14.6 1.187 1.206 51.120 52.819 init_scf_loop 11 6.9 0.026 0.077 46.748 46.749 ot_mini 117 10.6 0.077 0.609 42.954 43.289 prepare_preconditioner 11 7.9 0.000 0.000 42.066 42.092 make_preconditioner 11 8.9 0.000 0.001 42.066 42.092 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.760 40.664 multiply_cannon_multrec 30084 15.6 14.237 19.069 26.157 30.692 rebuild_ks_matrix 128 8.3 0.001 0.001 29.068 29.279 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.024 29.067 29.278 mp_waitall_1 147882 16.7 17.491 27.971 17.491 27.971 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.292 26.485 qs_ot_get_derivative 117 11.6 0.002 0.006 22.735 23.076 make_m2s 5014 13.6 0.095 0.100 20.800 21.849 make_images 5014 14.6 1.993 2.280 20.493 21.539 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.623 20.104 apply_single 128 13.6 0.001 0.001 19.623 20.103 ot_diis_step 117 11.6 0.019 0.028 20.016 20.093 qs_ot_get_p 128 10.4 0.001 0.001 19.776 20.072 cp_fm_cholesky_invert 11 10.9 16.161 16.170 16.161 16.170 cp_fm_upper_to_full 105 14.7 10.898 16.115 10.898 16.115 qs_ot_p2m_diag 83 11.4 0.344 0.391 15.529 15.582 multiply_cannon_metrocomm3 30084 15.6 0.048 0.050 6.230 15.304 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.248 14.250 sum_up_and_integrate 128 10.3 0.141 0.153 14.041 14.069 integrate_v_rspace 128 11.3 0.003 0.004 13.900 13.933 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.114 13.158 calculate_rho_elec 128 8.7 0.175 0.190 13.113 13.157 make_images_data 5014 15.6 0.065 0.069 11.310 12.776 dbcsr_complete_redistribute 395 12.7 1.521 1.634 8.987 12.691 multiply_cannon_sync_h2d 30084 15.6 11.698 12.662 11.698 12.662 dbcsr_mm_accdrv_process 62264 16.2 7.320 8.363 11.496 12.045 hybrid_alltoall_any 5200 16.5 0.524 2.188 10.009 11.929 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.583 11.257 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.109 11.112 init_scf_run 11 5.9 0.000 0.001 11.038 11.039 scf_env_initial_rho_setup 11 6.9 0.002 0.005 11.037 11.039 cp_fm_redistribute_end 83 14.4 1.920 11.023 1.936 11.029 cp_fm_diag_elpa_base 83 14.4 8.485 10.409 9.058 10.884 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.284 9.916 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.404 9.608 mp_alltoall_i22 716 14.1 5.558 9.231 5.558 9.231 pw_transfer 1547 11.6 0.085 0.100 7.872 7.938 grid_integrate_task_list 128 12.3 7.520 7.843 7.520 7.843 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.648 7.716 cp_fm_cholesky_decompose 22 10.9 7.541 7.648 7.541 7.648 wfi_extrapolate 11 7.9 0.001 0.003 7.636 7.636 multiply_cannon_metrocomm4 25070 15.6 0.076 0.087 2.730 7.200 mp_irecv_dv 76098 16.2 2.585 6.934 2.585 6.934 fft_wrap_pw1pw2_140 523 13.2 0.481 0.494 6.751 6.836 calculate_dm_sparse 128 9.5 0.001 0.001 6.538 6.631 density_rs2pw 128 9.7 0.006 0.006 6.260 6.601 fft3d_ps 1291 14.7 2.793 2.863 5.883 5.938 grid_collocate_task_list 128 9.7 5.204 5.561 5.204 5.561 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.475 5.532 mp_alltoall_d11v 2415 14.1 4.839 5.412 4.839 5.412 rs_pw_transfer 1046 11.9 0.014 0.015 4.573 4.899 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.419 4.555 mp_sum_l 7870 13.0 3.412 4.525 3.412 4.525 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.377 4.452 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.427 4.428 potential_pw2rs 128 12.3 0.022 0.023 4.240 4.258 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.168 4.230 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.309000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1078.545455, yerr=22.431162 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 5.910120E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.487864E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526748E+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 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592466352 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57630. MP_Allreduce 11104 1171. MP_Sync 88 MP_Alltoall 1724 18848034. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.038 172.026 172.027 qs_mol_dyn_low 1 2.0 0.003 0.003 171.653 171.665 qs_forces 11 3.9 0.002 0.003 171.547 171.553 qs_energies 11 4.9 0.001 0.002 164.257 164.264 scf_env_do_scf 11 5.9 0.001 0.001 147.007 147.018 velocity_verlet 10 3.0 0.001 0.001 112.418 112.422 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 111.295 111.296 qs_scf_new_mos 118 7.6 0.001 0.001 75.534 75.686 qs_scf_loop_do_ot 118 8.6 0.001 0.001 75.533 75.686 dbcsr_multiply_generic 2529 12.6 0.184 0.190 75.210 75.648 ot_scf_mini 118 9.6 0.004 0.004 71.043 71.123 multiply_cannon 2529 13.6 0.593 0.635 55.414 59.897 multiply_cannon_loop 2529 14.6 0.449 0.460 50.683 51.496 ot_mini 118 10.6 0.001 0.001 40.004 40.085 init_scf_loop 11 6.9 0.000 0.000 35.562 35.563 mp_waitall_1 126876 16.7 25.952 32.886 25.952 32.886 prepare_preconditioner 11 7.9 0.000 0.000 31.500 31.527 make_preconditioner 11 8.9 0.000 0.000 31.500 31.527 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.311 29.558 rebuild_ks_matrix 129 8.3 0.001 0.001 28.670 28.749 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 28.669 28.748 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.218 26.295 multiply_cannon_multrec 10116 15.6 10.507 14.368 18.164 21.049 ot_diis_step 118 11.6 0.020 0.021 20.014 20.014 qs_ot_get_derivative 118 11.6 0.002 0.002 19.919 19.997 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.548 19.853 apply_single 129 13.6 0.001 0.001 19.548 19.853 multiply_cannon_metrocomm3 10116 15.6 0.023 0.025 12.359 19.562 make_m2s 5058 13.6 0.066 0.071 16.246 18.773 make_images 5058 14.6 2.331 2.708 15.936 18.467 qs_ot_get_p 129 10.4 0.001 0.001 18.211 18.303 cp_fm_cholesky_invert 11 10.9 17.955 17.961 17.955 17.961 qs_ot_p2m_diag 84 11.4 0.502 0.508 14.267 14.283 sum_up_and_integrate 129 10.3 0.183 0.192 14.033 14.082 integrate_v_rspace 129 11.3 0.004 0.004 13.850 13.908 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.118 13.156 calculate_rho_elec 129 8.7 0.260 0.271 13.117 13.155 cp_dbcsr_syevd 84 12.4 0.005 0.005 13.129 13.130 make_images_data 5058 15.6 0.054 0.062 9.896 12.401 multiply_cannon_sync_h2d 10116 15.6 11.774 12.262 11.774 12.262 hybrid_alltoall_any 5245 16.5 0.817 3.820 9.719 12.005 init_scf_run 11 5.9 0.000 0.001 10.479 10.479 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.479 10.479 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.061 10.074 cp_fm_diag_elpa_base 84 14.4 9.800 9.892 10.049 10.062 grid_integrate_task_list 129 12.3 7.880 8.145 7.880 8.145 dbcsr_mm_accdrv_process 20934 16.1 2.745 3.807 7.279 7.982 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 7.925 7.980 cp_fm_cholesky_decompose 22 10.9 7.848 7.961 7.848 7.961 pw_transfer 1559 11.6 0.084 0.090 7.740 7.762 multiply_cannon_metrocomm1 10116 15.6 0.029 0.030 4.437 7.727 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.517 7.544 mp_allgather_i34 2529 14.6 2.922 7.514 2.922 7.514 wfi_extrapolate 11 7.9 0.001 0.001 7.207 7.207 fft_wrap_pw1pw2_140 527 13.2 0.505 0.524 6.631 6.663 calculate_dm_sparse 129 9.5 0.001 0.001 6.536 6.617 density_rs2pw 129 9.7 0.005 0.006 5.917 6.195 grid_collocate_task_list 129 9.7 5.522 5.766 5.522 5.766 fft3d_ps 1301 14.7 2.742 2.824 5.667 5.709 dbcsr_complete_redistribute 397 12.7 2.135 2.222 5.285 5.655 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.200 5.201 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.106 5.115 mp_alltoall_d11v 2429 14.1 4.681 5.023 4.681 5.023 rs_pw_transfer 1054 12.0 0.013 0.013 4.016 4.310 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.035 4.055 potential_pw2rs 129 12.3 0.027 0.028 4.011 4.027 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.572 3.878 multiply_cannon_metrocomm4 7587 15.6 0.024 0.027 1.856 3.822 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.473 3.790 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.741 3.774 mp_irecv_dv 29102 15.9 1.818 3.753 1.818 3.753 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.709 3.742 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.522 3.603 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=172.027000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1389.727273, yerr=40.400475 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 3.062534E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+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 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4049 58480. MP_Allreduce 11119 1510. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.064 287.095 287.096 qs_mol_dyn_low 1 2.0 0.003 0.003 286.567 286.578 qs_forces 11 3.9 0.003 0.003 286.476 286.478 qs_energies 11 4.9 0.002 0.002 277.957 277.974 scf_env_do_scf 11 5.9 0.001 0.001 256.477 256.493 velocity_verlet 10 3.0 0.001 0.002 208.009 208.017 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 133.144 133.146 init_scf_loop 11 6.9 0.000 0.000 123.081 123.086 prepare_preconditioner 11 7.9 0.000 0.000 118.419 118.450 make_preconditioner 11 8.9 0.000 0.000 118.419 118.450 make_full_inverse_cholesky 11 9.9 0.000 0.000 94.966 115.545 qs_scf_new_mos 118 7.6 0.001 0.001 90.987 91.100 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.987 91.099 ot_scf_mini 118 9.6 0.004 0.004 86.155 86.190 dbcsr_multiply_generic 2535 12.6 0.215 0.222 83.811 84.431 cp_fm_upper_to_full 106 14.8 52.728 75.113 52.728 75.113 multiply_cannon 2535 13.6 0.717 0.751 59.437 59.961 multiply_cannon_loop 2535 14.6 0.477 0.487 55.695 57.116 ot_mini 118 10.6 0.001 0.001 44.842 44.877 dbcsr_complete_redistribute 397 12.7 4.010 4.056 29.054 41.569 copy_fm_to_dbcsr 210 11.7 0.002 0.002 25.704 38.240 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.408 35.866 rebuild_ks_matrix 129 8.3 0.001 0.001 33.630 33.676 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 33.629 33.675 mp_alltoall_i22 720 14.1 21.219 33.621 21.219 33.621 cp_fm_cholesky_invert 11 10.9 33.199 33.205 33.199 33.205 mp_waitall_1 104820 16.8 28.378 32.284 28.378 32.284 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.424 31.476 qs_ot_get_p 129 10.4 0.001 0.001 25.651 25.673 qs_ot_get_derivative 118 11.6 0.002 0.002 24.323 24.359 qs_ot_p2m_diag 84 11.4 0.890 0.895 21.659 21.689 make_m2s 5070 13.6 0.077 0.079 19.708 20.738 multiply_cannon_metrocomm3 10140 15.6 0.025 0.025 19.247 20.568 ot_diis_step 118 11.6 0.022 0.022 20.487 20.487 make_images 5070 14.6 3.800 3.864 19.226 20.259 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.912 19.913 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.654 19.902 apply_single 129 13.6 0.001 0.001 19.654 19.901 multiply_cannon_multrec 10140 15.6 10.619 12.330 18.125 18.259 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.709 16.710 cp_fm_diag_elpa_base 84 14.4 12.321 13.888 16.705 16.705 multiply_cannon_sync_h2d 10140 15.6 15.784 15.820 15.784 15.820 sum_up_and_integrate 129 10.3 0.325 0.326 15.720 15.814 integrate_v_rspace 129 11.3 0.004 0.004 15.395 15.489 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.955 14.974 calculate_rho_elec 129 8.7 0.486 0.487 14.955 14.974 hybrid_alltoall_any 5257 16.5 1.314 3.061 10.545 12.219 make_images_data 5070 15.6 0.062 0.066 10.424 12.193 init_scf_run 11 5.9 0.000 0.001 11.646 11.647 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.646 11.647 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.554 9.587 dbcsr_mm_accdrv_process 20958 16.1 3.946 5.778 7.264 9.124 cp_fm_cholesky_decompose 22 10.9 8.799 8.845 8.799 8.845 grid_integrate_task_list 129 12.3 8.602 8.776 8.602 8.776 wfi_extrapolate 11 7.9 0.001 0.001 8.591 8.591 pw_transfer 1559 11.6 0.091 0.092 8.295 8.302 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.060 8.067 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.869 7.870 fft_wrap_pw1pw2_140 527 13.2 0.545 0.549 7.161 7.167 mp_alltoall_d11v 2429 14.1 6.864 7.004 6.864 7.004 calculate_dm_sparse 129 9.5 0.001 0.001 6.666 6.776 grid_collocate_task_list 129 9.7 6.386 6.426 6.386 6.426 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.284 6.379 fft3d_ps 1301 14.7 2.794 2.821 6.087 6.096 copy_dbcsr_to_fm 187 11.8 0.004 0.004 5.993 6.073 density_rs2pw 129 9.7 0.005 0.005 5.995 6.030 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=287.096000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2722.454545, yerr=191.300983 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_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 1.261855E+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 44549627. 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 0.022 0.045 84.591 84.591 qs_energies 1 2.0 0.000 0.000 84.135 84.149 ls_scf 1 3.0 0.000 0.000 83.221 83.234 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.463 72.651 multiply_cannon 111 7.7 0.017 0.020 55.813 57.127 multiply_cannon_loop 111 8.7 0.208 0.222 52.425 53.740 ls_scf_main 1 4.0 0.000 0.000 52.085 52.086 density_matrix_trs4 2 5.0 0.002 0.003 46.658 46.730 ls_scf_init_scf 1 4.0 0.000 0.002 28.122 28.123 ls_scf_init_matrix_S 1 5.0 0.000 0.002 27.061 27.116 mp_waitall_1 11031 10.9 22.338 25.050 22.338 25.050 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.947 24.968 multiply_cannon_multrec 2664 9.7 8.147 8.942 15.578 17.359 multiply_cannon_sync_h2d 2664 9.7 13.559 15.034 13.559 15.034 make_m2s 222 7.7 0.008 0.011 13.004 13.407 make_images 222 8.7 0.099 0.109 12.982 13.387 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.652 11.799 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.458 8.621 make_images_data 222 9.7 0.004 0.005 7.556 8.160 dbcsr_mm_accdrv_process 4760 10.4 0.516 0.622 7.051 8.019 hybrid_alltoall_any 227 10.6 0.215 1.839 6.482 8.011 dbcsr_mm_accdrv_process_sort 4760 11.4 6.336 7.249 6.336 7.249 calculate_norms 4752 9.8 5.521 6.164 5.521 6.164 apply_matrix_preconditioner 6 5.3 0.000 0.001 5.022 5.149 mp_sum_l 807 5.4 3.120 4.576 3.120 4.576 make_images_sizes 222 9.7 0.000 0.000 0.757 3.603 mp_alltoall_i44 222 10.7 0.756 3.602 0.756 3.602 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.324 3.459 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.063 3.254 mp_irecv_dv 6231 10.9 2.045 3.225 2.045 3.225 arnoldi_extremal 4 6.8 0.000 0.000 3.192 3.216 arnoldi_normal_ev 4 7.8 0.001 0.004 3.192 3.216 build_subspace 16 8.4 0.009 0.012 3.098 3.100 ls_scf_post 1 4.0 0.000 0.000 3.014 3.028 ls_scf_store_result 1 5.0 0.000 0.000 2.834 2.876 dbcsr_special_finalize 555 9.7 0.005 0.006 2.340 2.767 dbcsr_merge_single_wm 555 10.7 0.457 0.571 2.332 2.759 make_images_pack 222 9.7 2.207 2.645 2.208 2.647 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.305 2.547 dbcsr_sort_data 658 11.4 2.132 2.504 2.132 2.504 dbcsr_matrix_vector_mult_local 304 10.0 2.062 2.453 2.064 2.455 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.186 2.246 buffer_matrices_ensure_size 222 8.7 1.762 2.067 1.762 2.067 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.732 1.733 rebuild_ks_matrix 3 7.3 0.000 0.000 1.722 1.724 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.722 1.724 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.591000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_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 2.119414E+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 0.025 0.041 91.643 91.644 qs_energies 1 2.0 0.000 0.000 91.216 91.220 ls_scf 1 3.0 0.000 0.000 89.904 89.908 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.831 76.216 multiply_cannon 111 7.7 0.028 0.039 53.424 57.685 ls_scf_main 1 4.0 0.000 0.000 55.278 55.283 multiply_cannon_loop 111 8.7 0.115 0.123 50.136 53.521 density_matrix_trs4 2 5.0 0.002 0.003 49.649 49.855 ls_scf_init_scf 1 4.0 0.000 0.000 30.868 30.869 mp_waitall_1 9105 10.9 21.503 30.000 21.503 30.000 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.713 29.830 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.299 27.311 multiply_cannon_multrec 1332 9.7 12.976 16.265 22.324 26.667 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.948 20.398 make_m2s 222 7.7 0.006 0.007 15.564 16.301 make_images 222 8.7 1.568 1.916 15.534 16.272 dbcsr_mm_accdrv_process 4041 10.4 0.252 0.448 8.950 10.593 dbcsr_mm_accdrv_process_sort 4041 11.4 8.546 10.144 8.546 10.144 make_images_data 222 9.7 0.004 0.004 9.000 10.045 hybrid_alltoall_any 227 10.6 0.522 2.447 8.364 9.550 mp_sum_l 807 5.4 5.536 8.635 5.536 8.635 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.240 7.601 mp_irecv_dv 3311 11.0 3.220 7.539 3.220 7.539 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.332 6.978 calculate_norms 2376 9.8 5.981 6.608 5.981 6.608 multiply_cannon_sync_h2d 1332 9.7 4.769 5.825 4.769 5.825 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.018 5.214 arnoldi_extremal 4 6.8 0.000 0.000 4.594 4.612 arnoldi_normal_ev 4 7.8 0.001 0.005 4.594 4.612 build_subspace 16 8.4 0.014 0.021 4.342 4.345 ls_scf_post 1 4.0 0.000 0.000 3.758 3.764 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.120 3.352 ls_scf_store_result 1 5.0 0.000 0.000 3.185 3.277 dbcsr_matrix_vector_mult_local 304 10.0 2.738 3.220 2.740 3.221 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.507 2.591 mp_allgather_i34 111 8.7 0.982 2.489 0.982 2.489 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.185 2.431 make_images_pack 222 9.7 2.024 2.421 2.026 2.423 dbcsr_sort_data 436 11.2 1.847 2.097 1.847 2.097 dbcsr_data_new 4174 10.1 1.612 1.853 1.612 1.853 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.644000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1762.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_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 2.746638E+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 0.085 0.459 94.886 94.887 qs_energies 1 2.0 0.000 0.000 93.870 93.877 ls_scf 1 3.0 0.000 0.001 92.469 92.476 dbcsr_multiply_generic 111 6.7 0.017 0.023 77.228 77.464 ls_scf_main 1 4.0 0.000 0.000 57.818 57.823 multiply_cannon 111 7.7 0.089 0.402 52.927 56.976 multiply_cannon_loop 111 8.7 0.100 0.113 49.332 53.048 density_matrix_trs4 2 5.0 0.002 0.003 51.926 52.137 mp_waitall_1 7281 11.0 23.978 33.822 23.978 33.822 ls_scf_init_scf 1 4.0 0.000 0.001 31.073 31.076 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.912 29.971 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.527 27.548 multiply_cannon_multrec 888 9.7 12.613 15.120 21.301 24.334 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.040 22.676 make_m2s 222 7.7 0.006 0.008 17.344 18.556 make_images 222 8.7 1.970 2.292 17.305 18.517 make_images_data 222 9.7 0.004 0.005 9.941 11.016 hybrid_alltoall_any 227 10.6 0.620 2.866 9.565 11.012 dbcsr_mm_accdrv_process 3754 10.4 0.240 0.419 8.193 9.454 mp_sum_l 807 5.4 5.410 9.230 5.410 9.230 dbcsr_mm_accdrv_process_sort 3754 11.4 7.822 9.036 7.822 9.036 multiply_cannon_sync_h2d 888 9.7 6.038 7.346 6.038 7.346 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.191 7.226 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.619 7.224 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.458 6.843 mp_irecv_dv 2335 11.1 2.443 6.792 2.443 6.792 arnoldi_extremal 4 6.8 0.000 0.000 5.094 5.120 arnoldi_normal_ev 4 7.8 0.001 0.005 5.094 5.120 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.941 5.103 build_subspace 16 8.4 0.014 0.020 4.790 4.797 calculate_norms 1584 9.8 4.264 4.661 4.264 4.661 mp_allgather_i34 111 8.7 1.393 3.980 1.393 3.980 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.419 3.748 ls_scf_post 1 4.0 0.002 0.009 3.578 3.586 dbcsr_matrix_vector_mult_local 304 10.0 3.011 3.579 3.013 3.581 ls_scf_store_result 1 5.0 0.000 0.000 3.311 3.402 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.732 2.827 make_images_sizes 222 9.7 0.000 0.000 1.089 2.208 mp_alltoall_i44 222 10.7 1.089 2.208 1.089 2.208 make_images_pack 222 9.7 1.812 2.143 1.815 2.146 dbcsr_sort_data 325 11.1 1.877 2.138 1.877 2.138 dbcsr_data_release 9322 10.9 1.325 2.023 1.325 2.023 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.887000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2170.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_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 3.332948E+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 0.078 0.110 97.729 97.730 qs_energies 1 2.0 0.000 0.000 96.862 96.867 ls_scf 1 3.0 0.000 0.001 95.151 95.156 dbcsr_multiply_generic 111 6.7 0.017 0.022 78.267 78.477 ls_scf_main 1 4.0 0.000 0.000 58.405 58.406 multiply_cannon 111 7.7 0.058 0.115 51.362 56.470 density_matrix_trs4 2 5.0 0.002 0.003 52.421 52.565 multiply_cannon_loop 111 8.7 0.115 0.127 46.290 50.003 ls_scf_init_scf 1 4.0 0.001 0.002 33.497 33.499 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.972 32.056 mp_waitall_1 6369 11.0 22.208 29.595 22.208 29.595 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.477 29.488 multiply_cannon_multrec 1332 9.7 14.256 17.706 22.268 26.282 make_m2s 222 7.7 0.007 0.010 21.085 22.535 make_images 222 8.7 3.142 3.592 21.034 22.486 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 8.818 17.105 make_images_data 222 9.7 0.004 0.004 11.737 13.508 hybrid_alltoall_any 227 10.6 0.798 3.790 11.053 12.741 dbcsr_mm_accdrv_process 3641 10.4 0.197 0.400 7.643 9.182 dbcsr_mm_accdrv_process_sort 3641 11.4 7.265 8.767 7.265 8.767 mp_sum_l 807 5.4 4.256 7.589 4.256 7.589 multiply_cannon_sync_h2d 1332 9.7 5.484 6.248 5.484 6.248 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.096 6.048 mp_irecv_dv 3229 10.9 2.073 5.968 2.073 5.968 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.292 5.895 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.482 5.711 arnoldi_extremal 4 6.8 0.000 0.000 5.183 5.199 arnoldi_normal_ev 4 7.8 0.002 0.005 5.183 5.199 build_subspace 16 8.4 0.015 0.021 4.847 4.854 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.526 4.718 mp_allgather_i34 111 8.7 2.203 4.574 2.203 4.574 calculate_norms 2376 9.8 4.195 4.546 4.195 4.546 dbcsr_matrix_vector_mult 304 9.0 0.006 0.017 3.562 3.847 dbcsr_matrix_vector_mult_local 304 10.0 3.174 3.671 3.176 3.673 dbcsr_sort_data 658 11.4 3.042 3.479 3.042 3.479 ls_scf_post 1 4.0 0.000 0.001 3.248 3.253 dbcsr_special_finalize 555 9.7 0.006 0.007 2.776 3.243 dbcsr_merge_single_wm 555 10.7 0.529 0.663 2.767 3.235 ls_scf_store_result 1 5.0 0.000 0.000 2.979 3.052 ls_scf_dm_to_ks 2 5.0 0.001 0.010 2.886 2.928 dbcsr_data_release 10477 10.7 1.590 2.402 1.590 2.402 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.238 2.239 rebuild_ks_matrix 3 7.3 0.000 0.000 2.214 2.216 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.214 2.216 dbcsr_finalize 304 7.8 0.049 0.061 1.809 1.977 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.730000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2724.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_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 4.638441E+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 0.041 0.056 100.004 100.005 qs_energies 1 2.0 0.000 0.000 99.194 99.200 ls_scf 1 3.0 0.000 0.000 97.252 97.256 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.444 78.708 ls_scf_main 1 4.0 0.000 0.000 62.229 62.229 multiply_cannon 111 7.7 0.086 0.137 55.690 60.368 density_matrix_trs4 2 5.0 0.002 0.003 55.239 55.313 multiply_cannon_loop 111 8.7 0.069 0.077 51.227 53.555 mp_waitall_1 5436 11.0 27.116 32.595 27.116 32.595 ls_scf_init_scf 1 4.0 0.000 0.000 31.051 31.056 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.878 29.916 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.689 27.702 multiply_cannon_multrec 444 9.7 13.999 16.420 21.151 23.748 make_m2s 222 7.7 0.004 0.005 17.862 20.304 make_images 222 8.7 3.723 4.403 17.800 20.244 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.653 16.579 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.059 15.348 make_images_data 222 9.7 0.003 0.004 10.205 12.474 hybrid_alltoall_any 227 10.6 0.788 3.760 9.964 12.401 multiply_cannon_sync_h2d 444 9.7 6.584 8.124 6.584 8.124 dbcsr_mm_accdrv_process 3003 10.4 0.163 0.345 6.856 8.009 dbcsr_mm_accdrv_process_sort 3003 11.4 6.542 7.664 6.542 7.664 mp_allgather_i34 111 8.7 2.692 7.039 2.692 7.039 arnoldi_extremal 4 6.8 0.000 0.000 5.813 5.829 arnoldi_normal_ev 4 7.8 0.002 0.005 5.813 5.829 build_subspace 16 8.4 0.015 0.020 5.421 5.433 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.614 4.796 mp_sum_l 807 5.4 2.925 4.600 2.925 4.600 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.144 4.356 dbcsr_matrix_vector_mult_local 304 10.0 3.672 4.151 3.674 4.153 ls_scf_post 1 4.0 0.000 0.000 3.972 3.977 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.611 3.913 mp_irecv_dv 1241 11.2 1.593 3.893 1.593 3.893 calculate_norms 792 9.8 3.534 3.704 3.534 3.704 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.048 3.513 make_images_sizes 222 9.7 0.000 0.000 1.005 3.513 mp_alltoall_i44 222 10.7 1.005 3.513 1.005 3.513 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.370 3.454 ls_scf_store_result 1 5.0 0.000 0.000 3.411 3.453 dbcsr_finalize 304 7.8 0.062 0.078 2.207 2.331 dbcsr_merge_all 275 8.9 0.473 0.537 2.057 2.166 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.005000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3640.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/523d85c309ba47d334e69c8546862257e289ce38_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 8.760713E+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 0.076 0.094 107.976 107.976 qs_energies 1 2.0 0.000 0.000 106.405 106.418 ls_scf 1 3.0 0.000 0.000 103.398 103.412 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.775 77.883 ls_scf_main 1 4.0 0.000 0.000 65.530 65.531 density_matrix_trs4 2 5.0 0.002 0.003 56.687 56.739 multiply_cannon 111 7.7 0.099 0.159 49.918 51.776 multiply_cannon_loop 111 8.7 0.067 0.069 46.428 47.409 ls_scf_init_scf 1 4.0 0.000 0.000 34.199 34.200 ls_scf_init_matrix_S 1 5.0 0.000 0.001 32.881 32.890 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.049 30.057 mp_waitall_1 4527 11.1 22.074 25.974 22.074 25.974 make_m2s 222 7.7 0.005 0.005 23.870 24.884 make_images 222 8.7 4.589 4.977 23.764 24.775 multiply_cannon_multrec 444 9.7 17.866 18.539 22.553 23.101 hybrid_alltoall_any 227 10.6 1.660 3.631 12.910 15.608 make_images_data 222 9.7 0.003 0.003 13.118 15.580 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.377 10.876 multiply_cannon_sync_h2d 444 9.7 8.847 8.894 8.847 8.894 arnoldi_extremal 4 6.8 0.000 0.000 7.282 7.294 arnoldi_normal_ev 4 7.8 0.004 0.012 7.282 7.294 build_subspace 16 8.4 0.026 0.037 6.736 6.747 dbcsr_matrix_vector_mult 304 9.0 0.010 0.027 5.397 5.544 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.042 5.303 dbcsr_matrix_vector_mult_local 304 10.0 4.977 5.285 4.980 5.288 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.784 4.874 dbcsr_mm_accdrv_process 1814 10.4 0.234 0.307 4.512 4.646 dbcsr_mm_accdrv_process_sort 1814 11.4 4.210 4.348 4.210 4.348 ls_scf_post 1 4.0 0.000 0.000 3.670 3.683 mp_allgather_i34 111 8.7 1.143 3.581 1.143 3.581 make_images_sizes 222 9.7 0.000 0.000 1.421 3.446 mp_alltoall_i44 222 10.7 1.421 3.446 1.421 3.446 ls_scf_store_result 1 5.0 0.000 0.000 3.423 3.432 calculate_norms 792 9.8 3.241 3.280 3.241 3.280 dbcsr_finalize 304 7.8 0.082 0.089 3.087 3.130 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.976 2.976 dbcsr_merge_all 275 8.9 0.890 0.918 2.872 2.914 dbcsr_complete_redistribute 5 7.6 1.441 1.471 2.780 2.900 dbcsr_data_release 12724 10.6 2.332 2.858 2.332 2.858 matrix_ls_to_qs 2 6.0 0.000 0.000 2.419 2.545 dbcsr_sort_data 325 11.1 2.441 2.500 2.441 2.500 mp_sum_l 807 5.4 1.650 2.435 1.650 2.435 dbcsr_new_transposed 4 7.5 0.244 0.256 2.278 2.291 dbcsr_frobenius_norm 74 6.6 2.054 2.129 2.188 2.231 dbcsr_add_d 103 6.2 0.000 0.000 2.131 2.202 dbcsr_add_anytype 103 7.2 0.859 0.890 2.130 2.201 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.189 2.191 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.976000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6860.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 523d85c309ba47d334e69c8546862257e289ce38 Summary: empty Status: OK