=== 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: 7f090fb11291b829076ec7a97d9530b1aff7c614 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (03.02.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/01 job id: 45513108 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/02 job id: 45513109 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/03 job id: 45513112 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/04 job id: 45513113 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/05 job id: 45513116 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/06 job id: 45513118 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/07 job id: 45513119 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/08 job id: 45513121 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/09 job id: 45513123 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/10 job id: 45513124 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/11 job id: 45513126 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/12 job id: 45513127 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/13 job id: 45513128 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/14 job id: 45513129 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/15 job id: 45513131 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/16 job id: 45513132 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/17 job id: 45513133 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/18 job id: 45513134 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/19 job id: 45513135 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/20 job id: 45513136 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/21 job id: 45513137 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/22 job id: 45513138 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/23 job id: 45513140 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/24 job id: 45513141 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/25 job id: 45513142 --- 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/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/26 job id: 45513143 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/27 job id: 45513144 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.036 134.681 134.681 farming_run 1 2.0 134.118 134.119 134.649 134.652 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.447084E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1103589. MP_Allreduce 491 2254389. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.029 116.076 116.077 qs_energies 1 2.0 0.000 0.000 115.797 115.799 mp2_main 1 3.0 0.006 0.008 113.744 113.746 mp2_gpw_main 1 4.0 0.020 0.026 112.752 112.758 mp2_ri_gpw_compute_in 1 5.0 0.173 0.192 93.694 94.238 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.469 56.013 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.172 41.785 47.293 get_2c_integrals 1 6.0 0.000 0.000 37.424 38.033 integrate_v_rspace 273 8.0 0.438 0.453 25.128 30.427 pw_transfer 6555 10.6 0.369 0.385 27.406 27.950 grid_integrate_task_list 273 9.0 20.946 26.724 20.946 26.724 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.043 26.419 fft_wrap_pw1pw2_100 2178 12.4 1.219 1.434 23.573 23.959 compute_2c_integrals 1 7.0 0.002 0.003 19.720 19.721 compute_2c_integrals_loop_lm 1 8.0 0.016 0.018 18.937 19.442 mp2_eri_2c_integrate_gpw 1 9.0 2.364 2.426 18.921 19.424 rpa_ri_compute_en 1 5.0 0.032 0.042 18.948 19.114 cp_fm_cholesky_decompose 12 8.2 17.640 18.216 17.640 18.216 cholesky_decomp 1 7.0 0.000 0.000 16.547 17.120 fft3d_s 5443 13.4 16.159 16.580 16.181 16.601 ao_to_mo_and_store_B_mult_1 272 7.0 10.855 15.571 10.855 15.571 calculate_wavefunction 272 8.0 5.466 5.613 12.576 13.226 rpa_num_int 1 6.0 0.001 0.008 10.522 10.522 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.472 10.502 calc_mat_Q 8 8.0 0.000 0.000 9.344 9.455 contract_S_to_Q 8 9.0 0.000 0.000 8.767 8.876 calc_potential_gpw 544 9.5 0.005 0.006 8.258 8.725 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.241 8.562 parallel_gemm_fm 14 9.1 0.000 0.000 8.350 8.432 parallel_gemm_fm_cosma 14 10.1 8.350 8.432 8.350 8.432 potential_pw2rs 545 10.0 0.108 0.110 7.704 8.315 collocate_single_gaussian 272 10.0 0.040 0.042 7.493 7.776 create_integ_mat 1 6.0 0.022 0.028 7.745 7.745 array2fm 1 7.0 0.000 0.000 6.689 7.137 pw_scatter_s 2720 13.7 4.404 4.605 4.404 4.605 pw_gather_s 2722 13.2 3.858 4.233 3.858 4.233 array2fm_buffer_send 1 8.0 2.978 3.172 2.978 3.172 pw_poisson_solve 545 10.5 1.102 1.141 2.175 2.411 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.760378, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2719.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_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 0.028 0.036 396.195 396.196 farming_run 1 2.0 395.464 395.468 396.160 396.164 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.228960E+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 743 386399. MP_Allreduce 1941 22272. MP_Sync 37 MP_Alltoall 77 30213422. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.030 208.931 208.932 qs_energies 1 2.0 0.000 0.000 208.638 208.662 scf_env_do_scf 1 3.0 0.000 0.000 105.681 105.681 qs_ks_update_qs_env 5 5.0 0.000 0.000 104.780 104.788 rebuild_ks_matrix 4 6.0 0.000 0.000 104.779 104.787 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.064 104.779 104.786 hfx_ks_matrix 4 8.0 0.001 0.001 104.386 104.391 integrate_four_center 4 9.0 0.144 0.449 104.385 104.390 mp2_main 1 3.0 0.002 0.008 102.669 102.693 mp2_gpw_main 1 4.0 0.042 0.055 101.823 101.846 integrate_four_center_main 4 10.0 0.117 0.579 96.597 98.987 integrate_four_center_bin 261 11.0 96.480 98.977 96.480 98.977 init_scf_loop 1 4.0 0.000 0.000 91.408 91.408 mp2_ri_gpw_compute_in 1 5.0 0.070 0.094 74.914 76.036 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.615 55.727 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.165 42.342 47.452 integrate_v_rspace 95 8.0 0.398 0.566 28.710 33.760 pw_transfer 2240 10.6 0.145 0.167 29.931 30.357 ao_to_mo_and_store_B_mult_1 91 7.0 10.591 29.929 10.591 29.929 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.953 29.420 grid_integrate_task_list 95 9.0 23.982 29.232 23.982 29.232 mp2_ri_gpw_compute_en 1 5.0 0.084 0.110 26.755 28.501 fft_wrap_pw1pw2_100 730 12.4 1.243 1.384 26.681 27.191 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.840 1.902 24.934 24.943 get_2c_integrals 1 6.0 0.000 0.000 20.211 20.233 compute_2c_integrals 1 7.0 0.007 0.009 19.200 19.217 compute_2c_integrals_loop_lm 1 8.0 0.008 0.011 18.786 19.058 mp2_eri_2c_integrate_gpw 1 9.0 1.743 1.843 18.778 19.047 fft3d_s 1823 13.4 18.467 18.988 18.480 19.002 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.271 14.271 calculate_wavefunction 91 8.0 2.023 2.055 9.753 10.020 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.578 8.788 9.223 potential_pw2rs 186 10.0 0.033 0.034 8.670 9.148 local_gemm 172 8.0 8.230 8.645 8.230 8.645 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.000 0.001 8.194 8.548 mp2_ri_gpw_compute_en_comm 22 7.0 0.501 0.514 7.914 8.401 calc_potential_gpw 182 9.5 0.002 0.002 7.851 8.130 collocate_single_gaussian 91 10.0 0.016 0.019 7.851 8.087 mp_sendrecv_dm3 2068 8.0 5.945 6.462 5.945 6.462 mp2_ri_gpw_compute_en_ener 172 7.0 6.347 6.425 6.347 6.425 mp_sync 37 10.5 2.421 5.645 2.421 5.645 pw_gather_s 912 13.2 4.970 5.330 4.970 5.330 pw_scatter_s 910 13.7 3.876 4.245 3.876 4.245 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.839584, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1516.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_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 451.641344E+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 272. MP_Sync 530 MP_Alltoall 2083 588782. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.094 54.790 54.806 qs_mol_dyn_low 1 2.0 0.003 0.003 54.510 54.544 qs_forces 11 3.9 0.002 0.003 54.430 54.431 qs_energies 11 4.9 0.001 0.001 52.861 52.872 scf_env_do_scf 11 5.9 0.000 0.001 46.817 46.817 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.500 44.500 qs_scf_new_mos 108 7.5 0.000 0.001 33.116 33.387 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.115 33.386 dbcsr_multiply_generic 2286 12.5 0.101 0.135 32.270 32.895 ot_scf_mini 108 9.5 0.002 0.002 31.512 31.703 multiply_cannon 2286 13.5 0.190 0.201 25.245 27.020 velocity_verlet 10 3.0 0.001 0.001 26.486 26.487 multiply_cannon_loop 2286 14.5 1.461 1.559 24.318 25.989 ot_mini 108 10.5 0.001 0.001 18.582 18.849 qs_ot_get_derivative 108 11.5 0.001 0.001 15.614 15.826 mp_waitall_1 245248 16.5 8.006 14.258 8.006 14.258 multiply_cannon_metrocomm3 54864 15.5 0.069 0.076 5.351 12.467 multiply_cannon_multrec 54864 15.5 4.343 6.857 7.559 11.001 rebuild_ks_matrix 119 8.3 0.000 0.000 8.714 8.864 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.714 8.864 qs_ot_get_p 119 10.4 0.001 0.001 8.424 8.687 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.703 7.826 multiply_cannon_sync_h2d 54864 15.5 6.033 7.105 6.033 7.105 mp_sum_l 7207 12.9 4.533 6.584 4.533 6.584 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.212 5.664 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.551 5.591 sum_up_and_integrate 119 10.3 0.012 0.014 5.218 5.225 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.105 5.223 integrate_v_rspace 119 11.3 0.002 0.003 5.205 5.214 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.802 4.894 calculate_rho_elec 119 8.7 0.012 0.017 4.801 4.894 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.807 4.808 init_scf_run 11 5.9 0.000 0.001 4.719 4.719 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.718 4.719 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.500 4.500 dbcsr_mm_accdrv_process 76910 16.1 1.112 1.818 3.137 4.468 cp_fm_redistribute_end 50 14.0 2.291 4.462 2.303 4.466 cp_fm_diag_elpa_base 50 14.0 2.154 4.341 2.160 4.349 rs_pw_transfer 974 11.9 0.011 0.012 3.452 3.561 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.923 3.136 apply_single 119 13.6 0.000 0.000 2.923 3.136 calculate_dm_sparse 119 9.5 0.000 0.001 2.825 2.957 density_rs2pw 119 9.7 0.004 0.005 2.820 2.931 multiply_cannon_metrocomm1 54864 15.5 0.053 0.058 1.494 2.732 ot_diis_step 108 11.5 0.006 0.006 2.726 2.727 jit_kernel_multiply 13 15.8 1.965 2.565 1.965 2.565 potential_pw2rs 119 12.3 0.004 0.004 2.475 2.515 calculate_first_density_matrix 1 7.0 0.000 0.000 2.494 2.500 pw_transfer 1439 11.6 0.052 0.060 2.339 2.432 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 2.264 2.359 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.259 2.306 make_m2s 4572 13.5 0.053 0.055 2.224 2.302 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.288 2.290 init_scf_loop 11 6.9 0.000 0.000 2.287 2.288 make_images 4572 14.5 0.133 0.138 2.142 2.219 acc_transpose_blocks 54864 15.5 0.223 0.253 1.702 2.195 wfi_extrapolate 11 7.9 0.001 0.001 2.156 2.156 grid_integrate_task_list 119 12.3 2.046 2.150 2.046 2.150 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.988 2.062 fft3d_ps 1201 14.6 0.378 0.486 1.911 2.004 mp_sum_d 4135 12.0 1.259 1.957 1.259 1.957 fft_wrap_pw1pw2_140 487 13.2 0.180 0.198 1.729 1.825 mp_alltoall_d11v 2130 13.8 1.535 1.660 1.535 1.660 mp_waitany 12084 13.8 1.345 1.509 1.345 1.509 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.396 1.417 grid_collocate_task_list 119 9.7 1.290 1.358 1.290 1.358 prepare_preconditioner 11 7.9 0.000 0.000 1.164 1.197 make_preconditioner 11 8.9 0.000 0.000 1.164 1.197 make_images_sizes 4572 15.5 0.004 0.005 0.915 1.164 mp_alltoall_i44 4572 16.5 0.911 1.160 0.911 1.160 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.700 1.143 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.104 1.136 mp_alltoall_z22v 1201 16.6 1.072 1.134 1.072 1.134 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.806000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.545455, yerr=0.655555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.677952E+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 54 MP_Alltoall 2060 883190. 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 0.011 0.026 39.243 39.244 qs_mol_dyn_low 1 2.0 0.003 0.003 39.075 39.082 qs_forces 11 3.9 0.002 0.003 38.935 38.936 qs_energies 11 4.9 0.001 0.001 37.218 37.221 scf_env_do_scf 11 5.9 0.000 0.001 32.027 32.028 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.543 29.544 dbcsr_multiply_generic 2286 12.5 0.124 0.126 21.442 21.856 qs_scf_new_mos 108 7.5 0.001 0.001 20.209 20.455 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.209 20.455 ot_scf_mini 108 9.5 0.002 0.003 19.306 19.481 velocity_verlet 10 3.0 0.001 0.001 18.728 18.731 multiply_cannon 2286 13.5 0.207 0.219 16.356 17.868 multiply_cannon_loop 2286 14.5 0.892 0.967 15.236 16.796 ot_mini 108 10.5 0.001 0.001 11.859 12.097 mp_waitall_1 200699 16.5 5.830 11.175 5.830 11.175 multiply_cannon_metrocomm3 27432 15.5 0.067 0.071 4.232 9.748 qs_ot_get_derivative 108 11.5 0.001 0.001 9.384 9.560 multiply_cannon_multrec 27432 15.5 1.959 4.549 5.841 8.680 rebuild_ks_matrix 119 8.3 0.000 0.000 7.335 7.490 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.334 7.489 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.467 6.607 dbcsr_mm_accdrv_process 47894 16.0 2.970 5.347 3.813 5.641 qs_ot_get_p 119 10.4 0.001 0.001 4.533 4.761 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.656 4.502 sum_up_and_integrate 119 10.3 0.024 0.027 4.325 4.329 integrate_v_rspace 119 11.3 0.002 0.003 4.301 4.308 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.055 4.139 apply_single 119 13.6 0.000 0.000 3.055 4.139 mp_sum_l 7207 12.9 2.082 4.052 2.082 4.052 init_scf_run 11 5.9 0.000 0.001 3.984 3.984 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.984 3.984 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.834 3.881 calculate_rho_elec 119 8.7 0.021 0.024 3.833 3.881 rs_pw_transfer 974 11.9 0.010 0.011 2.685 3.090 qs_ot_p2m_diag 50 11.0 0.009 0.012 3.064 3.084 multiply_cannon_sync_h2d 27432 15.5 2.179 2.921 2.179 2.921 make_m2s 4572 13.5 0.052 0.054 2.528 2.762 make_images 4572 14.5 0.200 0.237 2.440 2.674 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.632 2.632 density_rs2pw 119 9.7 0.004 0.004 2.163 2.588 init_scf_loop 11 6.9 0.000 0.000 2.457 2.457 ot_diis_step 108 11.5 0.011 0.011 2.419 2.420 calculate_first_density_matrix 1 7.0 0.000 0.000 2.349 2.350 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.300 2.301 cp_fm_redistribute_end 50 14.0 1.166 2.271 1.170 2.274 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.159 2.251 cp_fm_diag_elpa_base 50 14.0 1.070 2.168 1.099 2.212 calculate_dm_sparse 119 9.5 0.000 0.001 2.057 2.131 potential_pw2rs 119 12.3 0.006 0.006 2.042 2.053 pw_transfer 1439 11.6 0.065 0.069 2.008 2.042 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.994 1.997 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.915 1.952 grid_integrate_task_list 119 12.3 1.834 1.937 1.834 1.937 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.728 1.772 jit_kernel_multiply 9 16.1 0.790 1.755 0.790 1.755 make_images_data 4572 15.5 0.044 0.050 1.166 1.582 wfi_extrapolate 11 7.9 0.001 0.001 1.582 1.582 prepare_preconditioner 11 7.9 0.000 0.000 1.556 1.582 make_preconditioner 11 8.9 0.000 0.000 1.556 1.582 acc_transpose_blocks 27432 15.5 0.108 0.112 1.214 1.531 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.450 1.510 fft_wrap_pw1pw2_140 487 13.2 0.203 0.212 1.468 1.504 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.017 1.494 fft3d_ps 1201 14.6 0.521 0.575 1.465 1.492 mp_alltoall_d11v 2130 13.8 1.256 1.461 1.256 1.461 grid_collocate_task_list 119 9.7 1.226 1.360 1.226 1.360 mp_allgather_i34 2286 14.5 0.564 1.336 0.564 1.336 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.265 1.275 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.194 1.240 mp_sum_d 4135 12.0 0.577 1.022 0.577 1.022 rs_pw_transfer_RS2PW_140 130 11.5 0.138 0.145 0.536 0.947 mp_waitany 5720 13.7 0.517 0.940 0.517 0.940 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.925 0.939 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.935 0.935 acc_transpose_blocks_kernels 27432 16.5 0.181 0.269 0.685 0.922 rs_pw_transfer_PW2RS_50 119 14.3 0.590 0.609 0.787 0.832 mp_alltoall_z22v 1201 16.6 0.723 0.799 0.723 0.799 make_images_sizes 4572 15.5 0.005 0.005 0.549 0.795 mp_alltoall_i44 4572 16.5 0.544 0.790 0.544 0.790 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.244000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.636364, yerr=1.366663 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_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 523.976704E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63497. MP_Allreduce 10075 307. MP_Sync 54 MP_Alltoall 1821 1607820. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.053 33.484 33.495 qs_mol_dyn_low 1 2.0 0.003 0.003 33.156 33.164 qs_forces 11 3.9 0.002 0.003 33.083 33.083 qs_energies 11 4.9 0.001 0.001 31.487 31.490 scf_env_do_scf 11 5.9 0.000 0.001 26.401 26.401 scf_env_do_scf_inner_loop 108 6.5 0.003 0.023 23.808 23.809 dbcsr_multiply_generic 2286 12.5 0.094 0.095 17.173 17.300 velocity_verlet 10 3.0 0.001 0.001 15.470 15.472 qs_scf_new_mos 108 7.5 0.001 0.001 15.327 15.357 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.326 15.357 multiply_cannon 2286 13.5 0.196 0.202 13.675 14.620 ot_scf_mini 108 9.5 0.002 0.002 14.570 14.584 multiply_cannon_loop 2286 14.5 0.630 0.659 12.836 13.837 ot_mini 108 10.5 0.001 0.001 9.023 9.036 qs_ot_get_derivative 108 11.5 0.001 0.001 7.488 7.506 multiply_cannon_multrec 18288 15.5 1.926 2.880 7.127 7.500 rebuild_ks_matrix 119 8.3 0.000 0.000 6.497 6.519 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.497 6.519 dbcsr_mm_accdrv_process 38222 16.0 4.360 5.882 5.116 5.942 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.737 5.756 sum_up_and_integrate 119 10.3 0.031 0.032 4.079 4.084 integrate_v_rspace 119 11.3 0.003 0.003 4.047 4.057 mp_waitall_1 158411 16.6 2.949 3.979 2.949 3.979 init_scf_run 11 5.9 0.000 0.001 3.934 3.934 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.934 3.934 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.622 3.640 calculate_rho_elec 119 8.7 0.031 0.032 3.622 3.639 qs_ot_get_p 119 10.4 0.001 0.001 3.333 3.366 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.709 3.357 rs_pw_transfer 974 11.9 0.009 0.010 2.494 2.742 calculate_first_density_matrix 1 7.0 0.000 0.000 2.716 2.717 init_scf_loop 11 6.9 0.000 0.000 2.577 2.578 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.084 2.387 apply_single 119 13.6 0.000 0.000 2.084 2.386 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.533 2.378 jit_kernel_multiply 11 15.8 0.704 2.377 0.704 2.377 density_rs2pw 119 9.7 0.004 0.004 2.122 2.374 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.222 2.229 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.041 2.043 make_m2s 4572 13.5 0.044 0.045 1.906 2.039 calculate_dm_sparse 119 9.5 0.000 0.001 1.946 1.966 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.958 1.959 make_images 4572 14.5 0.189 0.200 1.822 1.955 pw_transfer 1439 11.6 0.065 0.069 1.936 1.945 grid_integrate_task_list 119 12.3 1.805 1.917 1.805 1.917 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.906 1.914 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.843 1.854 potential_pw2rs 119 12.3 0.007 0.008 1.828 1.835 prepare_preconditioner 11 7.9 0.000 0.000 1.780 1.782 make_preconditioner 11 8.9 0.000 0.000 1.780 1.782 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.633 1.718 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.707 1.716 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.707 1.715 cp_fm_diag_elpa_base 50 14.0 1.683 1.695 1.705 1.713 mp_sum_l 7207 12.9 1.220 1.598 1.220 1.598 multiply_cannon_sync_h2d 18288 15.5 1.369 1.523 1.369 1.523 ot_diis_step 108 11.5 0.011 0.011 1.504 1.504 fft_wrap_pw1pw2_140 487 13.2 0.255 0.261 1.466 1.480 grid_collocate_task_list 119 9.7 1.211 1.356 1.211 1.356 acc_transpose_blocks 18288 15.5 0.075 0.076 1.324 1.345 fft3d_ps 1201 14.6 0.530 0.546 1.327 1.338 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.284 1.290 wfi_extrapolate 11 7.9 0.001 0.001 1.171 1.171 make_images_data 4572 15.5 0.044 0.048 0.884 1.058 hybrid_alltoall_any 4725 16.4 0.055 0.111 0.752 0.958 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.933 0.934 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.889 0.924 acc_transpose_blocks_kernels 18288 16.5 0.209 0.218 0.886 0.898 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.830 0.833 mp_alltoall_d11v 2130 13.8 0.706 0.808 0.706 0.808 mp_waitany 9880 13.7 0.531 0.790 0.531 0.790 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.122 0.522 0.766 cp_fm_cholesky_invert 11 10.9 0.759 0.762 0.759 0.762 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.660 0.744 mp_alltoall_z22v 1201 16.6 0.652 0.724 0.652 0.724 mp_allgather_i34 2286 14.5 0.301 0.688 0.301 0.688 jit_kernel_transpose 5 15.6 0.676 0.680 0.676 0.680 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.495000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=498.181818, yerr=2.442733 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_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 557.236224E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63496. MP_Allreduce 10074 349. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.045 39.228 39.237 qs_mol_dyn_low 1 2.0 0.003 0.003 38.901 38.913 qs_forces 11 3.9 0.002 0.003 38.826 38.827 qs_energies 11 4.9 0.001 0.001 37.020 37.026 scf_env_do_scf 11 5.9 0.000 0.001 31.398 31.399 scf_env_do_scf_inner_loop 108 6.5 0.004 0.009 27.650 27.650 dbcsr_multiply_generic 2286 12.5 0.098 0.102 19.407 19.514 velocity_verlet 10 3.0 0.001 0.001 19.477 19.478 qs_scf_new_mos 108 7.5 0.001 0.001 17.778 17.842 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.778 17.842 ot_scf_mini 108 9.5 0.002 0.003 16.790 16.846 multiply_cannon 2286 13.5 0.224 0.263 15.009 15.551 multiply_cannon_loop 2286 14.5 0.935 0.964 13.953 14.357 ot_mini 108 10.5 0.001 0.001 9.930 10.002 multiply_cannon_multrec 27432 15.5 2.365 3.063 9.028 9.378 qs_ot_get_derivative 108 11.5 0.001 0.001 8.088 8.143 dbcsr_mm_accdrv_process 47916 15.9 5.718 7.541 6.568 7.796 rebuild_ks_matrix 119 8.3 0.000 0.000 7.309 7.359 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.308 7.358 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.504 6.547 sum_up_and_integrate 119 10.3 0.035 0.038 4.454 4.464 integrate_v_rspace 119 11.3 0.003 0.003 4.419 4.429 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.385 4.423 calculate_rho_elec 119 8.7 0.040 0.046 4.384 4.423 init_scf_run 11 5.9 0.000 0.002 4.253 4.254 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.253 4.253 qs_ot_get_p 119 10.4 0.001 0.001 4.129 4.207 init_scf_loop 11 6.9 0.000 0.000 3.723 3.724 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.862 3.340 rs_pw_transfer 974 11.9 0.009 0.010 3.065 3.251 mp_waitall_1 137007 16.6 2.583 3.146 2.583 3.146 density_rs2pw 119 9.7 0.004 0.004 2.745 2.924 make_m2s 4572 13.5 0.054 0.056 2.773 2.892 prepare_preconditioner 11 7.9 0.000 0.000 2.807 2.816 make_preconditioner 11 8.9 0.000 0.000 2.807 2.816 make_images 4572 14.5 0.271 0.333 2.665 2.784 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.401 2.733 calculate_first_density_matrix 1 7.0 0.000 0.000 2.692 2.694 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.622 2.632 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.134 2.610 apply_single 119 13.6 0.000 0.000 2.133 2.610 pw_transfer 1439 11.6 0.065 0.071 2.378 2.415 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.286 2.328 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.270 2.270 calculate_dm_sparse 119 9.5 0.000 0.000 2.182 2.231 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.176 2.206 jit_kernel_multiply 10 15.9 0.789 2.202 0.789 2.202 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.183 2.185 potential_pw2rs 119 12.3 0.009 0.010 2.160 2.168 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.975 1.989 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.926 1.935 cp_fm_diag_elpa_base 50 14.0 1.886 1.906 1.924 1.933 grid_integrate_task_list 119 12.3 1.817 1.910 1.817 1.910 fft_wrap_pw1pw2_140 487 13.2 0.289 0.301 1.763 1.806 ot_diis_step 108 11.5 0.012 0.013 1.802 1.802 fft3d_ps 1201 14.6 0.559 0.607 1.722 1.751 mp_sum_l 7207 12.9 1.209 1.708 1.209 1.708 wfi_extrapolate 11 7.9 0.001 0.001 1.503 1.503 acc_transpose_blocks 27432 15.5 0.112 0.114 1.469 1.486 make_images_data 4572 15.5 0.045 0.048 1.135 1.356 grid_collocate_task_list 119 9.7 1.221 1.346 1.221 1.346 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.316 1.324 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.754 1.204 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.185 1.200 cp_fm_upper_to_full 72 14.2 0.838 1.196 0.838 1.196 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.940 1.175 dbcsr_complete_redistribute 329 12.2 0.120 0.151 0.872 1.143 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.086 1.087 mp_alltoall_z22v 1201 16.6 1.048 1.079 1.048 1.079 mp_alltoall_d11v 2130 13.8 0.942 1.074 0.942 1.074 multiply_cannon_sync_h2d 27432 15.5 0.981 1.040 0.981 1.040 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.897 0.902 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.621 0.887 cp_fm_cholesky_invert 11 10.9 0.871 0.875 0.871 0.875 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.873 acc_transpose_blocks_kernels 27432 16.5 0.268 0.275 0.844 0.855 make_images_sizes 4572 15.5 0.005 0.005 0.620 0.844 mp_alltoall_i44 4572 16.5 0.615 0.839 0.615 0.839 mp_sendrecv_dv 8211 12.7 0.806 0.819 0.806 0.819 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=39.237000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.090909, yerr=3.449757 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_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 598.511616E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63495. MP_Allreduce 10074 348. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.029 32.046 32.047 qs_mol_dyn_low 1 2.0 0.003 0.003 31.843 31.849 qs_forces 11 3.9 0.002 0.002 31.783 31.786 qs_energies 11 4.9 0.001 0.001 30.001 30.006 scf_env_do_scf 11 5.9 0.000 0.001 24.411 24.412 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 21.301 21.302 velocity_verlet 10 3.0 0.001 0.001 16.517 16.519 dbcsr_multiply_generic 2286 12.5 0.092 0.096 13.036 13.128 qs_scf_new_mos 108 7.5 0.001 0.001 12.262 12.287 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.262 12.287 ot_scf_mini 108 9.5 0.002 0.002 11.533 11.557 multiply_cannon 2286 13.5 0.230 0.235 9.906 10.589 multiply_cannon_loop 2286 14.5 0.329 0.343 8.882 9.082 rebuild_ks_matrix 119 8.3 0.000 0.000 6.641 6.675 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.640 6.675 ot_mini 108 10.5 0.001 0.001 6.358 6.389 multiply_cannon_multrec 9144 15.5 1.604 1.885 5.752 5.994 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.947 5.977 qs_ot_get_derivative 108 11.5 0.001 0.001 4.980 5.003 sum_up_and_integrate 119 10.3 0.038 0.041 4.214 4.220 integrate_v_rspace 119 11.3 0.003 0.003 4.176 4.183 dbcsr_mm_accdrv_process 12550 15.8 3.106 4.094 4.046 4.134 init_scf_run 11 5.9 0.000 0.002 4.092 4.092 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.092 4.092 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.046 4.055 calculate_rho_elec 119 8.7 0.060 0.061 4.046 4.054 qs_ot_get_p 119 10.4 0.001 0.001 3.276 3.311 init_scf_loop 11 6.9 0.000 0.000 3.083 3.083 mp_waitall_1 115863 16.7 2.282 2.755 2.282 2.755 calculate_first_density_matrix 1 7.0 0.000 0.000 2.733 2.734 rs_pw_transfer 974 11.9 0.008 0.008 2.429 2.579 density_rs2pw 119 9.7 0.004 0.004 2.329 2.461 make_m2s 4572 13.5 0.034 0.035 2.173 2.396 pw_transfer 1439 11.6 0.066 0.070 2.326 2.344 make_images 4572 14.5 0.268 0.300 2.084 2.306 prepare_preconditioner 11 7.9 0.000 0.000 2.260 2.265 make_preconditioner 11 8.9 0.000 0.000 2.260 2.265 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.232 2.251 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.159 2.161 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.109 2.157 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.946 1.947 grid_integrate_task_list 119 12.3 1.851 1.911 1.851 1.911 potential_pw2rs 119 12.3 0.010 0.011 1.878 1.889 calculate_dm_sparse 119 9.5 0.000 0.000 1.785 1.808 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.756 1.776 fft_wrap_pw1pw2_140 487 13.2 0.366 0.375 1.739 1.759 jit_kernel_multiply 10 15.7 0.903 1.744 0.903 1.744 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.618 1.625 cp_fm_diag_elpa_base 50 14.0 1.588 1.605 1.616 1.623 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.587 1.589 fft3d_ps 1201 14.6 0.563 0.576 1.562 1.578 grid_collocate_task_list 119 9.7 1.282 1.413 1.282 1.413 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.390 1.400 make_images_data 4572 15.5 0.039 0.042 1.017 1.362 ot_diis_step 108 11.5 0.013 0.013 1.351 1.352 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.952 1.344 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.328 1.339 wfi_extrapolate 11 7.9 0.001 0.001 1.302 1.302 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.248 1.250 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.219 1.250 apply_single 119 13.6 0.000 0.000 1.219 1.249 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.160 1.166 cp_fm_cholesky_invert 11 10.9 1.158 1.161 1.158 1.161 mp_alltoall_d11v 2130 13.8 0.959 1.079 0.959 1.079 parallel_gemm_fm 81 9.0 0.000 0.000 1.064 1.067 parallel_gemm_fm_cosma 81 10.0 1.064 1.067 1.064 1.067 make_basis_sm 11 9.8 0.000 0.000 0.957 0.958 acc_transpose_blocks 9144 15.5 0.037 0.038 0.909 0.925 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.921 mp_alltoall_z22v 1201 16.6 0.873 0.911 0.873 0.911 mp_allgather_i34 2286 14.5 0.355 0.861 0.355 0.861 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.819 0.823 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.795 0.806 multiply_cannon_metrocomm1 9144 15.5 0.021 0.022 0.590 0.790 multiply_cannon_sync_h2d 9144 15.5 0.708 0.784 0.708 0.784 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.701 0.756 make_images_sizes 4572 15.5 0.005 0.005 0.496 0.719 mp_alltoall_i44 4572 16.5 0.491 0.714 0.491 0.714 dbcsr_complete_redistribute 329 12.2 0.161 0.170 0.649 0.689 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 0.670 0.686 mp_sum_l 7207 12.9 0.529 0.671 0.529 0.671 cp_fm_cholesky_decompose 22 10.9 0.649 0.654 0.649 0.654 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=32.047000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=568.181818, yerr=2.917601 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_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 761.327616E+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 0.019 0.061 46.224 46.236 qs_mol_dyn_low 1 2.0 0.003 0.003 46.001 46.004 qs_forces 11 3.9 0.002 0.003 45.630 45.632 qs_energies 11 4.9 0.001 0.001 43.553 43.557 scf_env_do_scf 11 5.9 0.001 0.001 37.495 37.496 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 28.546 28.547 velocity_verlet 10 3.0 0.001 0.001 26.474 26.515 dbcsr_multiply_generic 2286 12.5 0.100 0.102 18.972 19.159 qs_scf_new_mos 108 7.5 0.001 0.001 17.699 17.787 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.698 17.787 ot_scf_mini 108 9.5 0.002 0.002 16.507 16.596 multiply_cannon 2286 13.5 0.304 0.317 14.191 15.377 multiply_cannon_loop 2286 14.5 0.343 0.349 12.714 13.708 ot_mini 108 10.5 0.001 0.001 9.786 9.914 init_scf_loop 11 6.9 0.000 0.000 8.918 8.920 multiply_cannon_multrec 9144 15.5 3.375 4.963 8.535 8.634 rebuild_ks_matrix 119 8.3 0.000 0.000 7.975 8.114 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.974 8.113 prepare_preconditioner 11 7.9 0.000 0.000 7.764 7.778 make_preconditioner 11 8.9 0.000 0.000 7.764 7.778 qs_ot_get_derivative 108 11.5 0.001 0.001 7.605 7.696 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.304 7.643 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.209 7.334 dbcsr_mm_accdrv_process 12550 15.8 4.047 5.620 5.034 6.359 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.869 4.932 calculate_rho_elec 119 8.7 0.118 0.121 4.868 4.932 mp_waitall_1 94719 16.7 3.394 4.587 3.394 4.587 cp_fm_upper_to_full 72 14.2 3.191 4.582 3.191 4.582 sum_up_and_integrate 119 10.3 0.064 0.066 4.501 4.505 integrate_v_rspace 119 11.3 0.003 0.003 4.436 4.442 qs_ot_get_p 119 10.4 0.001 0.001 3.802 3.982 init_scf_run 11 5.9 0.000 0.001 3.977 3.977 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.976 3.977 make_m2s 4572 13.5 0.038 0.038 2.925 3.242 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.707 3.159 pw_transfer 1439 11.6 0.069 0.070 3.118 3.122 make_images 4572 14.5 0.353 0.381 2.804 3.120 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.019 3.024 dbcsr_complete_redistribute 329 12.2 0.284 0.291 2.101 2.941 density_rs2pw 119 9.7 0.004 0.004 2.632 2.676 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.676 2.649 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.358 2.570 apply_single 119 13.6 0.000 0.000 2.358 2.570 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.725 2.552 fft_wrap_pw1pw2_140 487 13.2 0.619 0.622 2.462 2.469 calculate_dm_sparse 119 9.5 0.000 0.000 2.366 2.427 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.325 2.328 calculate_first_density_matrix 1 7.0 0.000 0.000 2.249 2.314 mp_alltoall_i22 627 13.8 1.438 2.283 1.438 2.283 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.455 2.276 mp_sum_l 7207 12.9 1.329 2.168 1.329 2.168 cp_fm_cholesky_invert 11 10.9 2.164 2.168 2.164 2.168 ot_diis_step 108 11.5 0.014 0.014 2.134 2.135 grid_integrate_task_list 119 12.3 2.096 2.120 2.096 2.120 fft3d_ps 1201 14.6 0.598 0.614 2.033 2.035 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.021 2.022 make_images_data 4572 15.5 0.043 0.046 1.530 1.991 rs_pw_transfer 974 11.9 0.009 0.009 1.950 1.982 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.505 1.967 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.919 1.964 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.896 1.896 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.790 1.791 potential_pw2rs 119 12.3 0.014 0.015 1.773 1.777 jit_kernel_multiply 10 15.7 0.960 1.757 0.960 1.757 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.673 1.673 cp_fm_diag_elpa_base 50 14.0 1.520 1.575 1.671 1.671 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.601 1.646 wfi_extrapolate 11 7.9 0.001 0.001 1.592 1.592 grid_collocate_task_list 119 9.7 1.518 1.537 1.518 1.537 mp_alltoall_d11v 2130 13.8 1.403 1.505 1.403 1.505 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.331 1.391 mp_alltoall_z22v 1201 16.6 1.298 1.320 1.298 1.320 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.275 1.310 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.125 1.141 mp_allgather_i34 2286 14.5 0.458 1.069 0.458 1.069 multiply_cannon_sync_h2d 9144 15.5 1.048 1.051 1.048 1.051 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.993 1.034 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.978 0.993 qs_create_task_list 11 7.9 0.001 0.001 0.946 0.958 generate_qs_task_list 11 8.9 0.368 0.387 0.946 0.957 acc_transpose_blocks 9144 15.5 0.038 0.038 0.938 0.945 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=46.236000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=711.727273, yerr=14.397543 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_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 501.125120E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66218. MP_Allreduce 9696 492. MP_Sync 52 MP_Alltoall 1938 1952776. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.031 86.347 86.347 qs_mol_dyn_low 1 2.0 0.003 0.003 86.078 86.084 qs_forces 11 3.9 0.003 0.004 86.006 86.008 qs_energies 11 4.9 0.001 0.001 83.047 83.064 scf_env_do_scf 11 5.9 0.000 0.001 73.765 73.767 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 67.343 67.344 dbcsr_multiply_generic 2055 12.4 0.105 0.109 50.822 51.174 qs_scf_new_mos 99 7.5 0.000 0.001 48.445 48.595 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.444 48.595 ot_scf_mini 99 9.5 0.002 0.002 46.080 46.198 velocity_verlet 10 3.0 0.001 0.001 44.530 44.531 multiply_cannon 2055 13.4 0.185 0.190 42.230 43.065 multiply_cannon_loop 2055 14.4 1.487 1.514 41.080 41.917 ot_mini 99 10.5 0.001 0.001 26.317 26.429 qs_ot_get_derivative 99 11.5 0.001 0.001 19.548 19.668 multiply_cannon_multrec 49320 15.4 12.545 13.430 17.478 18.540 rebuild_ks_matrix 110 8.3 0.000 0.000 15.454 15.585 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 15.453 15.584 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.585 13.710 mp_waitall_1 220248 16.4 10.896 11.734 10.896 11.734 qs_ot_get_p 110 10.4 0.001 0.001 11.303 11.455 multiply_cannon_sync_h2d 49320 15.4 10.400 11.158 10.400 11.158 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.362 7.892 qs_ot_p2m_diag 48 11.0 0.012 0.019 7.801 7.816 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.233 7.815 apply_single 110 13.6 0.000 0.001 7.233 7.814 sum_up_and_integrate 110 10.3 0.036 0.043 7.778 7.791 integrate_v_rspace 110 11.3 0.003 0.004 7.742 7.763 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.465 7.601 calculate_rho_elec 110 8.6 0.021 0.026 7.465 7.601 init_scf_run 11 5.9 0.000 0.002 7.129 7.130 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.129 7.129 multiply_cannon_metrocomm3 49320 15.4 0.079 0.083 6.086 7.091 cp_dbcsr_syevd 48 12.0 0.002 0.003 6.957 6.957 ot_diis_step 99 11.5 0.006 0.006 6.596 6.597 init_scf_loop 11 6.9 0.000 0.000 6.387 6.388 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.233 6.260 cp_fm_diag_elpa_base 48 14.0 6.212 6.242 6.231 6.259 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.484 5.550 dbcsr_mm_accdrv_process 87628 16.1 1.851 1.961 4.806 5.118 rs_pw_transfer 902 11.9 0.011 0.013 4.549 5.079 density_rs2pw 110 9.6 0.004 0.005 4.293 4.867 mp_sum_l 6514 12.8 3.659 4.651 3.659 4.651 make_m2s 4110 13.4 0.060 0.066 4.477 4.574 make_images 4110 14.4 0.178 0.192 4.381 4.480 wfi_extrapolate 11 7.9 0.001 0.001 4.235 4.235 prepare_preconditioner 11 7.9 0.000 0.000 4.137 4.160 make_preconditioner 11 8.9 0.000 0.000 4.137 4.160 pw_transfer 1331 11.6 0.056 0.063 4.022 4.136 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.933 4.050 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.913 3.972 calculate_dm_sparse 110 9.5 0.001 0.001 3.856 3.937 multiply_cannon_metrocomm1 49320 15.4 0.060 0.063 2.411 3.929 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.660 3.664 grid_integrate_task_list 110 12.3 3.227 3.406 3.227 3.406 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.277 3.334 potential_pw2rs 110 12.3 0.006 0.007 3.291 3.333 fft_wrap_pw1pw2_140 451 13.1 0.448 0.493 3.135 3.250 fft3d_ps 1111 14.6 0.793 0.889 3.132 3.234 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.151 3.193 calculate_first_density_matrix 1 7.0 0.000 0.000 2.794 2.798 jit_kernel_multiply 13 15.9 2.684 2.727 2.684 2.727 mp_alltoall_d11v 2046 13.8 2.196 2.712 2.196 2.712 cp_fm_cholesky_invert 11 10.9 2.459 2.464 2.459 2.464 mp_waitany 14300 13.8 1.871 2.418 1.871 2.418 grid_collocate_task_list 110 9.6 2.085 2.300 2.085 2.300 acc_transpose_blocks 49320 15.4 0.204 0.216 2.112 2.213 mp_alltoall_z22v 1111 16.6 1.952 2.142 1.952 2.142 make_images_data 4110 15.4 0.042 0.045 1.963 2.115 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.999 2.016 hybrid_alltoall_any 4261 16.3 0.082 0.481 1.689 1.978 mp_sum_d 3889 11.9 1.319 1.859 1.319 1.859 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.781 1.807 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=86.347000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.272727, yerr=2.377581 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 588.816384E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66437. MP_Allreduce 9695 570. MP_Sync 52 MP_Alltoall 1717 2567777. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.061 0.087 74.238 74.243 qs_mol_dyn_low 1 2.0 0.003 0.004 73.671 73.843 qs_forces 11 3.9 0.003 0.004 73.585 73.588 qs_energies 11 4.9 0.017 0.076 70.078 70.098 scf_env_do_scf 11 5.9 0.001 0.002 60.949 60.952 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 52.289 52.291 dbcsr_multiply_generic 2055 12.4 0.116 0.119 38.787 38.977 velocity_verlet 10 3.0 0.001 0.001 38.535 38.536 qs_scf_new_mos 99 7.5 0.001 0.001 34.966 35.093 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.965 35.093 ot_scf_mini 99 9.5 0.003 0.004 33.274 33.393 multiply_cannon 2055 13.4 0.220 0.242 31.452 32.949 multiply_cannon_loop 2055 14.4 0.921 0.952 30.014 30.871 ot_mini 99 10.5 0.001 0.001 19.073 19.200 multiply_cannon_multrec 24660 15.4 7.619 9.107 13.908 15.519 rebuild_ks_matrix 110 8.3 0.000 0.000 14.489 14.615 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 14.488 14.615 qs_ot_get_derivative 99 11.5 0.001 0.001 13.157 13.276 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.692 12.797 mp_waitall_1 176588 16.5 8.386 11.058 8.386 11.058 init_scf_loop 11 6.9 0.001 0.003 8.612 8.615 multiply_cannon_metrocomm3 24660 15.4 0.067 0.069 5.352 8.320 multiply_cannon_sync_h2d 24660 15.4 6.994 7.950 6.994 7.950 qs_ot_get_p 110 10.4 0.001 0.001 7.452 7.596 sum_up_and_integrate 110 10.3 0.052 0.059 7.281 7.306 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.612 7.267 apply_single 110 13.6 0.000 0.001 6.612 7.267 integrate_v_rspace 110 11.3 0.003 0.003 7.229 7.255 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.765 6.774 calculate_rho_elec 110 8.6 0.040 0.048 6.765 6.774 init_scf_run 11 5.9 0.000 0.001 6.543 6.543 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.542 6.543 dbcsr_mm_accdrv_process 52282 16.1 4.758 5.688 6.131 6.489 prepare_preconditioner 11 7.9 0.000 0.000 6.463 6.487 make_preconditioner 11 8.9 0.000 0.001 6.463 6.487 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.019 6.197 ot_diis_step 99 11.5 0.010 0.011 5.856 5.856 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.804 5.497 qs_ot_p2m_diag 48 11.0 0.028 0.044 5.287 5.310 make_m2s 4110 13.4 0.057 0.060 4.675 5.171 make_images 4110 14.4 0.398 0.442 4.567 5.058 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.818 4.818 density_rs2pw 110 9.6 0.005 0.006 3.896 4.372 pw_transfer 1331 11.6 0.067 0.075 4.168 4.343 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.060 4.238 rs_pw_transfer 902 11.9 0.012 0.014 3.720 4.224 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.148 4.161 cp_fm_diag_elpa_base 48 14.0 4.094 4.113 4.145 4.158 wfi_extrapolate 11 7.9 0.001 0.001 3.756 3.756 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.432 3.498 fft_wrap_pw1pw2_140 451 13.1 0.521 0.536 3.226 3.406 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.401 3.403 grid_integrate_task_list 110 12.3 3.150 3.357 3.150 3.357 fft3d_ps 1111 14.6 1.113 1.336 3.033 3.223 cp_fm_cholesky_invert 11 10.9 3.210 3.219 3.210 3.219 potential_pw2rs 110 12.3 0.008 0.009 3.057 3.102 make_images_data 4110 15.4 0.046 0.050 2.551 3.096 calculate_dm_sparse 110 9.5 0.001 0.001 3.064 3.090 hybrid_alltoall_any 4261 16.3 0.103 0.448 2.198 3.059 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.998 3.032 mp_sum_l 6514 12.8 2.061 2.737 2.061 2.737 calculate_first_density_matrix 1 7.0 0.001 0.003 2.673 2.678 grid_collocate_task_list 110 9.6 2.049 2.480 2.049 2.480 mp_alltoall_d11v 2046 13.8 1.883 2.245 1.883 2.245 jit_kernel_multiply 11 16.3 1.023 2.189 1.023 2.189 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.054 2.071 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.954 1.971 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.910 1.915 multiply_cannon_metrocomm4 22605 15.4 0.075 0.080 0.791 1.901 mp_allgather_i34 2055 14.4 0.766 1.851 0.766 1.851 mp_waitany 10164 13.8 1.314 1.823 1.314 1.823 mp_alltoall_z22v 1111 16.6 1.623 1.742 1.623 1.742 mp_irecv_dv 57340 16.2 0.665 1.683 0.665 1.683 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.629 1.643 cp_fm_cholesky_decompose 22 10.9 1.579 1.586 1.579 1.586 dbcsr_complete_redistribute 325 12.2 0.238 0.297 1.300 1.571 acc_transpose_blocks 24660 15.4 0.110 0.113 1.550 1.568 rs_pw_transfer_RS2PW_140 121 11.5 0.208 0.230 0.989 1.492 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=74.243000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.636364, yerr=6.799368 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 659.070976E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66428. MP_Allreduce 9694 567. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.074 60.433 60.434 qs_mol_dyn_low 1 2.0 0.003 0.005 60.035 60.042 qs_forces 11 3.9 0.004 0.013 59.955 59.956 qs_energies 11 4.9 0.009 0.030 56.746 56.749 scf_env_do_scf 11 5.9 0.001 0.001 48.654 48.654 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 40.087 40.087 velocity_verlet 10 3.0 0.004 0.015 32.682 32.684 dbcsr_multiply_generic 2055 12.4 0.106 0.111 28.212 28.456 qs_scf_new_mos 99 7.5 0.001 0.001 24.867 24.953 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.867 24.952 ot_scf_mini 99 9.5 0.003 0.003 23.648 23.753 multiply_cannon 2055 13.4 0.211 0.219 21.985 23.148 multiply_cannon_loop 2055 14.4 0.615 0.630 20.817 21.926 ot_mini 99 10.5 0.001 0.001 13.419 13.527 rebuild_ks_matrix 110 8.3 0.000 0.000 12.522 12.637 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 12.522 12.637 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.038 11.142 multiply_cannon_multrec 16440 15.4 3.982 5.343 9.742 10.967 mp_waitall_1 139946 16.5 6.840 10.148 6.840 10.148 qs_ot_get_derivative 99 11.5 0.001 0.001 9.003 9.120 init_scf_loop 11 6.9 0.001 0.005 8.533 8.534 multiply_cannon_metrocomm3 16440 15.4 0.042 0.046 4.119 7.294 prepare_preconditioner 11 7.9 0.000 0.000 6.696 6.713 make_preconditioner 11 8.9 0.000 0.002 6.696 6.713 sum_up_and_integrate 110 10.3 0.060 0.060 6.636 6.650 integrate_v_rspace 110 11.3 0.003 0.004 6.576 6.590 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.006 6.346 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.132 6.141 calculate_rho_elec 110 8.6 0.059 0.059 6.131 6.140 dbcsr_mm_accdrv_process 34862 16.1 4.694 5.298 5.613 5.739 qs_ot_get_p 110 10.4 0.001 0.001 5.442 5.561 init_scf_run 11 5.9 0.000 0.001 5.536 5.537 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.536 5.536 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.927 5.354 apply_single 110 13.6 0.000 0.000 4.927 5.354 make_m2s 4110 13.4 0.049 0.050 4.181 4.552 make_images 4110 14.4 0.390 0.507 4.067 4.438 density_rs2pw 110 9.6 0.005 0.006 3.203 4.385 multiply_cannon_sync_h2d 16440 15.4 3.705 4.372 3.705 4.372 ot_diis_step 99 11.5 0.011 0.011 4.368 4.368 rs_pw_transfer 902 11.9 0.010 0.012 2.797 3.979 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.850 3.853 pw_transfer 1331 11.6 0.066 0.076 3.731 3.739 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.081 3.692 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.623 3.634 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.517 3.517 grid_integrate_task_list 110 12.3 3.152 3.400 3.152 3.400 fft_wrap_pw1pw2_140 451 13.1 0.641 0.653 3.112 3.125 wfi_extrapolate 11 7.9 0.001 0.001 2.992 2.992 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.975 2.986 cp_fm_diag_elpa_base 48 14.0 2.910 2.941 2.974 2.984 make_images_data 4110 15.4 0.043 0.047 2.423 2.913 hybrid_alltoall_any 4261 16.3 0.105 0.373 2.132 2.798 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.791 2.792 cp_fm_cholesky_invert 11 10.9 2.628 2.634 2.628 2.634 calculate_dm_sparse 110 9.5 0.001 0.001 2.492 2.524 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.466 2.509 fft3d_ps 1111 14.6 1.092 1.105 2.441 2.451 grid_collocate_task_list 110 9.6 2.088 2.448 2.088 2.448 calculate_first_density_matrix 1 7.0 0.001 0.003 2.445 2.446 mp_waitany 17072 13.8 1.194 2.419 1.194 2.419 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.311 2.368 multiply_cannon_metrocomm4 14385 15.4 0.046 0.050 0.859 2.359 potential_pw2rs 110 12.3 0.011 0.011 2.344 2.358 mp_irecv_dv 48980 15.7 0.788 2.237 0.788 2.237 mp_alltoall_d11v 2046 13.8 1.826 2.148 1.826 2.148 rs_pw_transfer_RS2PW_140 121 11.5 0.177 0.182 0.908 2.084 mp_sum_l 6514 12.8 1.450 2.041 1.450 2.041 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.030 2.032 dbcsr_complete_redistribute 325 12.2 0.326 0.354 1.462 1.920 cp_fm_upper_to_full 70 14.2 1.389 1.758 1.389 1.758 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.678 1.692 cp_fm_cholesky_decompose 22 10.9 1.573 1.587 1.573 1.587 mp_allgather_i34 2055 14.4 0.493 1.535 0.493 1.535 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.358 1.469 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.456 1.468 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.987 1.435 rs_gather_matrices 110 12.3 0.234 0.258 1.000 1.290 acc_transpose_blocks 16440 15.4 0.072 0.074 1.249 1.269 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.236 1.244 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.434000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.363636, yerr=8.434483 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_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 739.954688E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66426. MP_Allreduce 9694 608. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.032 65.614 65.615 qs_mol_dyn_low 1 2.0 0.003 0.003 65.332 65.341 qs_forces 11 3.9 0.003 0.003 65.261 65.262 qs_energies 11 4.9 0.001 0.001 61.810 61.814 scf_env_do_scf 11 5.9 0.000 0.001 53.546 53.549 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.776 41.776 velocity_verlet 10 3.0 0.001 0.001 37.238 37.240 dbcsr_multiply_generic 2055 12.4 0.113 0.118 29.352 29.546 qs_scf_new_mos 99 7.5 0.001 0.001 26.534 26.660 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.534 26.659 ot_scf_mini 99 9.5 0.003 0.003 24.909 25.027 multiply_cannon 2055 13.4 0.242 0.259 22.349 23.420 multiply_cannon_loop 2055 14.4 0.881 0.902 20.919 21.460 ot_mini 99 10.5 0.001 0.001 13.963 14.106 multiply_cannon_multrec 24660 15.4 4.239 6.937 12.624 13.894 rebuild_ks_matrix 110 8.3 0.000 0.000 12.350 12.476 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.350 12.475 init_scf_loop 11 6.9 0.000 0.000 11.726 11.727 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.908 11.019 prepare_preconditioner 11 7.9 0.000 0.000 9.967 9.986 make_preconditioner 11 8.9 0.000 0.001 9.967 9.986 qs_ot_get_derivative 99 11.5 0.001 0.001 9.838 9.960 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.195 9.657 dbcsr_mm_accdrv_process 52304 16.0 6.798 8.167 8.238 9.156 sum_up_and_integrate 110 10.3 0.068 0.071 6.615 6.630 integrate_v_rspace 110 11.3 0.003 0.003 6.547 6.560 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.194 6.205 calculate_rho_elec 110 8.6 0.078 0.081 6.194 6.205 mp_waitall_1 121746 16.5 4.276 6.112 4.276 6.112 qs_ot_get_p 110 10.4 0.001 0.001 5.921 6.089 make_m2s 4110 13.4 0.060 0.063 5.409 5.779 make_images 4110 14.4 0.577 0.697 5.268 5.635 init_scf_run 11 5.9 0.000 0.001 5.490 5.490 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.490 5.490 cp_fm_upper_to_full 70 14.2 3.369 4.880 3.369 4.880 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.137 4.151 ot_diis_step 99 11.5 0.011 0.012 4.090 4.090 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.001 4.089 apply_single 110 13.6 0.000 0.000 4.001 4.089 pw_transfer 1331 11.6 0.066 0.075 3.837 3.869 dbcsr_complete_redistribute 325 12.2 0.412 0.458 2.657 3.810 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.730 3.765 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.666 3.666 density_rs2pw 110 9.6 0.004 0.005 3.099 3.613 grid_integrate_task_list 110 12.3 3.260 3.448 3.260 3.448 multiply_cannon_sync_h2d 24660 15.4 3.206 3.402 3.206 3.402 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.267 3.315 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.125 3.258 fft_wrap_pw1pw2_140 451 13.1 0.671 0.694 3.178 3.219 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.354 3.126 hybrid_alltoall_any 4261 16.3 0.120 0.456 2.304 3.110 make_images_data 4110 15.4 0.046 0.050 2.741 3.108 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.025 3.097 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.087 3.097 cp_fm_diag_elpa_base 48 14.0 2.935 2.991 3.085 3.095 wfi_extrapolate 11 7.9 0.001 0.001 3.006 3.006 rs_pw_transfer 902 11.9 0.010 0.011 2.454 2.977 calculate_dm_sparse 110 9.5 0.001 0.001 2.911 2.945 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.763 2.879 mp_alltoall_i22 605 13.7 1.652 2.811 1.652 2.811 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.789 2.791 cp_fm_cholesky_invert 11 10.9 2.755 2.763 2.755 2.763 fft3d_ps 1111 14.6 1.085 1.116 2.487 2.504 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.428 2.457 grid_collocate_task_list 110 9.6 2.182 2.455 2.182 2.455 calculate_first_density_matrix 1 7.0 0.000 0.000 2.392 2.396 qs_energies_init_hamiltonians 11 5.9 0.003 0.016 2.270 2.271 potential_pw2rs 110 12.3 0.013 0.013 2.246 2.255 mp_alltoall_d11v 2046 13.8 1.803 2.116 1.803 2.116 jit_kernel_multiply 10 15.7 1.110 1.815 1.110 1.815 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.696 1.734 cp_fm_cholesky_decompose 22 10.9 1.680 1.715 1.680 1.715 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.599 1.703 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.684 1.697 multiply_cannon_metrocomm4 20550 15.4 0.061 0.064 0.840 1.658 mp_irecv_dv 62702 16.1 0.736 1.570 0.736 1.570 mp_sum_l 6514 12.8 0.916 1.569 0.916 1.569 mp_waitany 13376 13.8 1.063 1.548 1.063 1.548 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.528 1.546 mp_allgather_i34 2055 14.4 0.529 1.538 0.529 1.538 acc_transpose_blocks 24660 15.4 0.106 0.109 1.501 1.525 rs_pw_transfer_RS2PW_140 121 11.5 0.167 0.173 0.808 1.328 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.212 1.319 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.615000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=699.727273, yerr=10.863541 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_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 820.031488E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66424. MP_Allreduce 9694 649. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.047 57.823 57.834 qs_mol_dyn_low 1 2.0 0.003 0.003 57.586 57.591 qs_forces 11 3.9 0.010 0.020 57.320 57.323 qs_energies 11 4.9 0.001 0.001 53.584 53.595 scf_env_do_scf 11 5.9 0.000 0.001 45.144 45.144 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 37.075 37.076 velocity_verlet 10 3.0 0.001 0.001 32.211 32.226 dbcsr_multiply_generic 2055 12.4 0.104 0.109 23.431 23.601 qs_scf_new_mos 99 7.5 0.001 0.001 21.616 21.689 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.615 21.688 ot_scf_mini 99 9.5 0.002 0.002 20.345 20.399 multiply_cannon 2055 13.4 0.248 0.259 17.631 18.998 multiply_cannon_loop 2055 14.4 0.320 0.336 16.209 16.647 rebuild_ks_matrix 110 8.3 0.000 0.000 12.165 12.210 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.164 12.210 ot_mini 99 10.5 0.001 0.001 10.851 10.905 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.814 10.856 multiply_cannon_multrec 8220 15.4 3.198 4.833 7.420 8.644 mp_waitall_1 103326 16.6 6.496 8.029 6.496 8.029 init_scf_loop 11 6.9 0.000 0.000 8.017 8.019 qs_ot_get_derivative 99 11.5 0.001 0.001 6.932 6.985 sum_up_and_integrate 110 10.3 0.079 0.080 6.694 6.706 integrate_v_rspace 110 11.3 0.003 0.003 6.615 6.625 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.495 6.513 calculate_rho_elec 110 8.6 0.114 0.115 6.494 6.512 prepare_preconditioner 11 7.9 0.000 0.000 6.308 6.316 make_preconditioner 11 8.9 0.000 0.000 6.308 6.316 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.884 5.962 qs_ot_get_p 110 10.4 0.001 0.001 5.466 5.540 init_scf_run 11 5.9 0.000 0.001 5.289 5.290 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.289 5.289 dbcsr_mm_accdrv_process 17442 15.9 2.820 3.598 4.092 4.973 make_m2s 4110 13.4 0.038 0.039 4.373 4.595 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.188 4.587 make_images 4110 14.4 0.638 0.692 4.243 4.462 pw_transfer 1331 11.6 0.066 0.072 4.111 4.126 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.003 4.022 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.946 3.950 ot_diis_step 99 11.5 0.012 0.012 3.888 3.889 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.779 3.829 apply_single 110 13.6 0.000 0.000 3.779 3.829 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.623 3.624 density_rs2pw 110 9.6 0.004 0.005 3.226 3.599 grid_integrate_task_list 110 12.3 3.346 3.540 3.346 3.540 fft_wrap_pw1pw2_140 451 13.1 0.835 0.847 3.424 3.446 cp_fm_cholesky_invert 11 10.9 3.139 3.143 3.139 3.143 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.049 3.057 cp_fm_diag_elpa_base 48 14.0 2.993 3.016 3.047 3.055 multiply_cannon_sync_h2d 8220 15.4 2.904 3.024 2.904 3.024 make_images_data 4110 15.4 0.038 0.043 2.488 2.850 hybrid_alltoall_any 4261 16.3 0.201 0.866 2.387 2.847 wfi_extrapolate 11 7.9 0.001 0.001 2.821 2.821 rs_pw_transfer 902 11.9 0.010 0.011 2.382 2.798 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.685 2.693 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.596 2.597 fft3d_ps 1111 14.6 1.135 1.147 2.545 2.557 calculate_dm_sparse 110 9.5 0.001 0.001 2.497 2.546 grid_collocate_task_list 110 9.6 2.277 2.537 2.277 2.537 calculate_first_density_matrix 1 7.0 0.000 0.000 2.369 2.370 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.205 2.221 potential_pw2rs 110 12.3 0.016 0.016 2.199 2.205 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.035 2.061 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.777 2.007 mp_alltoall_d11v 2046 13.8 1.746 1.999 1.746 1.999 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.937 1.966 mp_allgather_i34 2055 14.4 0.606 1.734 0.606 1.734 cp_fm_cholesky_decompose 22 10.9 1.715 1.728 1.715 1.728 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.684 1.691 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.519 1.641 dbcsr_complete_redistribute 325 12.2 0.585 0.599 1.526 1.621 mp_waitany 9240 13.8 1.139 1.571 1.139 1.571 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.434 1.446 jit_kernel_multiply 8 15.9 0.965 1.389 0.965 1.389 qs_create_task_list 11 7.9 0.001 0.001 1.223 1.327 generate_qs_task_list 11 8.9 0.375 0.443 1.223 1.326 multiply_cannon_metrocomm1 8220 15.4 0.020 0.021 0.923 1.311 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.478 1.279 mp_irecv_dv 24056 15.7 0.454 1.240 0.454 1.240 rs_pw_transfer_RS2PW_140 121 11.5 0.163 0.169 0.826 1.235 rs_gather_matrices 110 12.3 0.325 0.365 0.979 1.220 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.164 1.198 mp_alltoall_z22v 1111 16.6 1.177 1.198 1.177 1.198 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.834000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=776.454545, yerr=10.439119 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_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 1.330131E+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 0.026 0.100 87.287 87.294 qs_mol_dyn_low 1 2.0 0.003 0.003 86.916 86.922 qs_forces 11 3.9 0.003 0.003 85.945 85.946 qs_energies 11 4.9 0.001 0.001 81.791 81.793 scf_env_do_scf 11 5.9 0.001 0.001 71.780 71.780 velocity_verlet 10 3.0 0.001 0.001 56.071 56.184 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 43.621 43.624 dbcsr_multiply_generic 2055 12.4 0.118 0.122 28.052 28.134 init_scf_loop 11 6.9 0.000 0.000 28.088 28.092 prepare_preconditioner 11 7.9 0.000 0.000 26.072 26.081 make_preconditioner 11 8.9 0.000 0.000 26.072 26.081 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.295 25.557 qs_scf_new_mos 99 7.5 0.001 0.001 25.490 25.526 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.489 25.525 ot_scf_mini 99 9.5 0.002 0.002 23.745 23.766 multiply_cannon 2055 13.4 0.342 0.362 21.205 21.994 multiply_cannon_loop 2055 14.4 0.341 0.345 19.446 19.764 cp_fm_upper_to_full 70 14.2 12.658 18.170 12.658 18.170 rebuild_ks_matrix 110 8.3 0.000 0.001 14.002 14.031 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.001 14.030 ot_mini 99 10.5 0.001 0.001 13.238 13.263 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.681 12.707 dbcsr_complete_redistribute 325 12.2 1.015 1.027 7.324 10.558 multiply_cannon_multrec 8220 15.4 4.374 4.560 9.545 9.672 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.316 9.550 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.762 8.969 qs_ot_get_derivative 99 11.5 0.001 0.001 8.783 8.807 mp_alltoall_i22 605 13.7 5.383 8.603 5.383 8.603 mp_waitall_1 84994 16.7 7.257 8.018 7.257 8.018 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.688 7.723 calculate_rho_elec 110 8.6 0.226 0.226 7.687 7.722 sum_up_and_integrate 110 10.3 0.150 0.153 7.271 7.285 integrate_v_rspace 110 11.3 0.004 0.004 7.121 7.135 init_scf_run 11 5.9 0.000 0.001 5.756 5.757 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.756 5.756 make_m2s 4110 13.4 0.043 0.044 5.181 5.674 make_images 4110 14.4 0.882 0.932 4.993 5.486 qs_ot_get_p 110 10.4 0.001 0.001 5.399 5.429 dbcsr_mm_accdrv_process 11614 15.7 3.289 3.694 5.027 5.327 cp_fm_cholesky_invert 11 10.9 5.118 5.123 5.118 5.123 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.582 5.058 apply_single 110 13.6 0.000 0.000 4.582 5.058 pw_transfer 1331 11.6 0.075 0.076 5.027 5.033 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.911 4.916 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.532 4.874 ot_diis_step 99 11.5 0.015 0.015 4.438 4.438 fft_wrap_pw1pw2_140 451 13.1 1.339 1.342 4.275 4.282 multiply_cannon_sync_h2d 8220 15.4 3.949 3.955 3.949 3.955 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.907 3.915 density_rs2pw 110 9.6 0.005 0.005 3.721 3.745 grid_integrate_task_list 110 12.3 3.660 3.711 3.660 3.711 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.667 3.668 hybrid_alltoall_any 4261 16.3 0.256 0.555 2.844 3.561 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.469 3.469 make_images_data 4110 15.4 0.042 0.045 2.829 3.437 wfi_extrapolate 11 7.9 0.001 0.001 3.282 3.282 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.791 3.238 calculate_dm_sparse 110 9.5 0.001 0.001 3.136 3.159 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.888 2.888 cp_fm_diag_elpa_base 48 14.0 2.350 2.554 2.886 2.887 fft3d_ps 1111 14.6 1.304 1.312 2.838 2.848 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.832 2.833 grid_collocate_task_list 110 9.6 2.633 2.657 2.633 2.657 potential_pw2rs 110 12.3 0.021 0.022 2.389 2.393 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.346 2.356 calculate_first_density_matrix 1 7.0 0.000 0.000 2.354 2.354 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.336 2.350 rs_pw_transfer 902 11.9 0.010 0.011 2.264 2.291 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.191 2.243 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.087 2.179 cp_fm_cholesky_decompose 22 10.9 1.999 2.018 1.999 2.018 mp_alltoall_d11v 2046 13.8 1.932 2.013 1.932 2.013 qs_create_task_list 11 7.9 0.001 0.001 1.888 1.936 generate_qs_task_list 11 8.9 0.732 0.786 1.887 1.935 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.921 1.923 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.738 1.767 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.294000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1198.454545, yerr=53.647611 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 634.138624E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 57694. MP_Allreduce 11017 802. MP_Sync 86 MP_Alltoall 2210 3486929. MP_SendRecv 24130 18752. MP_ISendRecv 24130 18752. MP_Wait 42150 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.034 204.442 204.444 qs_mol_dyn_low 1 2.0 0.004 0.007 204.008 204.021 qs_forces 11 3.9 0.005 0.005 203.848 203.850 qs_energies 11 4.9 0.001 0.002 198.235 198.248 scf_env_do_scf 11 5.9 0.001 0.001 181.686 181.690 scf_env_do_scf_inner_loop 116 6.6 0.004 0.011 161.029 161.031 dbcsr_multiply_generic 2485 12.5 0.177 0.182 122.421 122.993 velocity_verlet 10 3.0 0.001 0.001 121.859 121.860 qs_scf_new_mos 116 7.6 0.001 0.001 121.349 121.594 qs_scf_loop_do_ot 116 8.6 0.001 0.001 121.349 121.593 ot_scf_mini 116 9.6 0.003 0.003 114.887 115.107 multiply_cannon 2485 13.5 0.238 0.246 99.937 101.366 multiply_cannon_loop 2485 14.5 2.012 2.087 97.781 99.096 ot_mini 116 10.6 0.001 0.001 64.505 64.716 multiply_cannon_multrec 59640 15.5 33.241 35.719 41.315 42.991 qs_ot_get_derivative 116 11.6 0.001 0.001 40.027 40.249 rebuild_ks_matrix 127 8.3 0.001 0.001 33.532 33.746 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.017 33.531 33.745 mp_waitall_1 264810 16.5 27.591 30.524 27.591 30.524 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.076 30.275 multiply_cannon_sync_h2d 59640 15.5 27.626 29.601 27.626 29.601 qs_ot_get_p 127 10.4 0.001 0.001 28.732 29.147 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.113 24.636 apply_single 127 13.6 0.001 0.001 24.113 24.636 ot_diis_step 116 11.6 0.008 0.008 24.266 24.267 qs_ot_p2m_diag 82 11.4 0.078 0.091 22.317 22.400 init_scf_loop 11 6.9 0.000 0.000 20.581 20.582 cp_dbcsr_syevd 82 12.4 0.004 0.005 19.991 19.992 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 18.605 18.770 multiply_cannon_metrocomm3 59640 15.5 0.115 0.120 15.228 17.148 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.053 17.085 cp_fm_diag_elpa_base 82 14.4 16.985 17.019 17.049 17.079 prepare_preconditioner 11 7.9 0.000 0.000 15.948 15.987 make_preconditioner 11 8.9 0.000 0.000 15.947 15.987 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.147 15.321 sum_up_and_integrate 127 10.3 0.089 0.107 14.173 14.188 integrate_v_rspace 127 11.3 0.004 0.005 14.084 14.102 make_m2s 4970 13.5 0.102 0.110 13.686 14.033 make_images 4970 14.5 0.399 0.418 13.507 13.865 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.603 13.702 calculate_rho_elec 127 8.7 0.046 0.064 13.602 13.702 init_scf_run 11 5.9 0.000 0.002 12.385 12.385 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.385 12.385 density_rs2pw 127 9.7 0.006 0.007 7.005 10.263 cp_fm_cholesky_invert 11 10.9 9.116 9.124 9.116 9.124 wfi_extrapolate 11 7.9 0.001 0.001 9.122 9.122 rs_pw_transfer 1038 11.9 0.016 0.018 5.742 8.981 mp_sum_l 7804 13.0 7.538 8.720 7.538 8.720 calculate_dm_sparse 127 9.5 0.001 0.001 8.264 8.351 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.230 8.346 multiply_cannon_metrocomm1 59640 15.5 0.087 0.093 5.849 8.251 dbcsr_mm_accdrv_process 123452 16.2 2.999 3.204 7.631 8.247 pw_transfer 1535 11.6 0.074 0.096 7.710 7.878 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.639 7.745 fft_wrap_pw1pw2 1281 12.7 0.010 0.013 7.508 7.676 make_images_data 4970 15.5 0.065 0.071 6.743 7.602 grid_integrate_task_list 127 12.3 7.017 7.491 7.017 7.491 hybrid_alltoall_any 5155 16.4 0.291 2.242 5.915 7.166 fft_wrap_pw1pw2_140 519 13.2 1.264 1.305 6.596 6.784 cp_dbcsr_sm_fm_multiply 37 9.5 0.004 0.019 6.711 6.721 mp_waitany 15900 13.9 2.679 6.034 2.679 6.034 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.801 5.870 grid_collocate_task_list 127 9.7 4.534 5.722 4.534 5.722 fft3d_ps 1281 14.7 2.125 2.757 5.292 5.599 rs_pw_transfer_RS2PW_140 138 11.5 0.278 0.300 2.097 5.354 mp_alltoall_d11v 2401 14.1 4.262 5.189 4.262 5.189 cp_fm_cholesky_decompose 22 10.9 4.624 4.639 4.624 4.639 potential_pw2rs 127 12.3 0.009 0.011 4.607 4.628 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.444000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=599.272727, yerr=7.875057 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 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 826.896384E+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 1463834560144 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58363. MP_Allreduce 10978 967. MP_Sync 87 MP_Alltoall 1969 4769219. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.064 191.932 191.934 qs_mol_dyn_low 1 2.0 0.007 0.033 191.331 191.343 qs_forces 11 3.9 0.008 0.033 191.205 191.214 qs_energies 11 4.9 0.011 0.070 184.382 184.394 scf_env_do_scf 11 5.9 0.001 0.001 167.670 167.680 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 134.470 134.473 velocity_verlet 10 3.0 0.002 0.010 120.408 120.411 dbcsr_multiply_generic 2507 12.6 0.187 0.192 97.158 98.387 qs_scf_new_mos 117 7.6 0.001 0.001 94.234 94.755 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.233 94.754 ot_scf_mini 117 9.6 0.004 0.005 89.432 90.012 multiply_cannon 2507 13.6 0.476 0.525 77.111 80.885 multiply_cannon_loop 2507 14.6 1.249 1.290 74.114 76.875 ot_mini 117 10.6 0.001 0.002 49.751 50.381 mp_waitall_1 214728 16.6 24.718 37.981 24.718 37.981 multiply_cannon_multrec 30084 15.6 22.008 26.517 31.595 36.492 rebuild_ks_matrix 128 8.3 0.001 0.001 33.412 33.988 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 33.411 33.988 init_scf_loop 11 6.9 0.001 0.003 33.111 33.112 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.057 30.586 multiply_cannon_metrocomm3 30084 15.6 0.091 0.098 15.775 28.846 prepare_preconditioner 11 7.9 0.000 0.000 28.687 28.755 make_preconditioner 11 8.9 0.000 0.001 28.687 28.754 qs_ot_get_derivative 117 11.6 0.001 0.002 27.815 28.376 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.379 27.929 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.021 23.069 apply_single 128 13.6 0.001 0.001 22.021 23.069 qs_ot_get_p 128 10.4 0.001 0.001 21.523 22.264 ot_diis_step 117 11.6 0.016 0.024 21.759 21.766 multiply_cannon_sync_h2d 30084 15.6 19.211 21.195 19.211 21.195 qs_ot_p2m_diag 83 11.4 0.187 0.216 16.764 16.804 cp_fm_cholesky_invert 11 10.9 16.721 16.733 16.721 16.733 make_m2s 5014 13.6 0.089 0.095 14.348 15.651 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.612 15.613 make_images 5014 14.6 1.152 1.337 14.140 15.444 sum_up_and_integrate 128 10.3 0.116 0.132 14.842 14.872 integrate_v_rspace 128 11.3 0.004 0.004 14.725 14.760 qs_rho_update_rho_low 128 7.7 0.001 0.002 14.183 14.210 calculate_rho_elec 128 8.7 0.088 0.104 14.183 14.209 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.414 12.443 cp_fm_diag_elpa_base 83 14.4 12.158 12.258 12.408 12.440 init_scf_run 11 5.9 0.000 0.002 11.620 11.621 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.619 11.621 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.992 11.422 multiply_cannon_metrocomm4 27577 15.6 0.098 0.114 3.772 10.870 mp_irecv_dv 69486 16.3 3.574 10.474 3.574 10.474 make_images_data 5014 15.6 0.063 0.071 8.781 10.372 density_rs2pw 128 9.7 0.006 0.007 7.551 9.887 dbcsr_mm_accdrv_process 62242 16.2 4.525 5.149 9.051 9.643 hybrid_alltoall_any 5200 16.5 0.343 1.513 7.471 9.343 pw_transfer 1547 11.6 0.086 0.096 8.906 8.959 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 8.681 8.738 rs_pw_transfer 1046 11.9 0.015 0.025 6.022 8.409 wfi_extrapolate 11 7.9 0.001 0.001 8.388 8.388 fft_wrap_pw1pw2_140 523 13.2 1.341 1.359 7.651 7.723 grid_integrate_task_list 128 12.3 7.174 7.536 7.174 7.536 cp_fm_cholesky_decompose 22 10.9 7.099 7.189 7.099 7.189 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.172 6.925 calculate_dm_sparse 128 9.5 0.001 0.001 6.460 6.626 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 6.171 6.181 mp_sum_l 7870 13.0 3.923 6.032 3.923 6.032 fft3d_ps 1291 14.7 2.813 2.980 5.980 6.022 grid_collocate_task_list 128 9.7 4.700 5.857 4.700 5.857 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.329 5.477 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.338 5.411 mp_waitany 11748 13.9 2.644 5.109 2.644 5.109 potential_pw2rs 128 12.3 0.015 0.017 5.071 5.099 mp_alltoall_d11v 2415 14.1 4.200 5.036 4.200 5.036 mp_allgather_i34 2507 14.6 1.569 4.692 1.569 4.692 rs_pw_transfer_RS2PW_140 139 11.5 0.378 0.485 2.258 4.633 mp_sum_d 4465 12.1 2.650 3.975 2.650 3.975 dbcsr_complete_redistribute 395 12.7 0.762 0.833 3.121 3.967 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.934000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.181818, yerr=2.587318 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 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 11528908111872 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 93.514772E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755942624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 944.377856E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531401248 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57958. MP_Allreduce 11047 1005. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.053 178.258 178.260 qs_mol_dyn_low 1 2.0 0.003 0.004 177.741 177.808 qs_forces 11 3.9 0.004 0.005 177.632 177.636 qs_energies 11 4.9 0.010 0.078 170.986 170.996 scf_env_do_scf 11 5.9 0.001 0.001 154.943 154.943 scf_env_do_scf_inner_loop 117 6.6 0.009 0.013 119.541 119.542 velocity_verlet 10 3.0 0.001 0.001 113.424 113.426 dbcsr_multiply_generic 2507 12.6 0.179 0.184 82.255 83.336 qs_scf_new_mos 117 7.6 0.001 0.001 81.021 81.385 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.020 81.384 ot_scf_mini 117 9.6 0.004 0.004 76.869 77.260 multiply_cannon 2507 13.6 0.500 0.518 62.240 66.662 multiply_cannon_loop 2507 14.6 0.856 0.883 59.212 61.844 ot_mini 117 10.6 0.001 0.001 42.787 43.194 mp_waitall_1 170520 16.6 25.286 35.638 25.286 35.638 init_scf_loop 11 6.9 0.001 0.004 35.298 35.299 rebuild_ks_matrix 128 8.3 0.001 0.001 31.009 31.593 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 31.008 31.593 prepare_preconditioner 11 7.9 0.000 0.000 31.215 31.263 make_preconditioner 11 8.9 0.001 0.013 31.215 31.263 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.816 30.246 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.921 28.457 multiply_cannon_multrec 20056 15.6 13.494 17.092 22.245 25.921 multiply_cannon_metrocomm3 20056 15.6 0.058 0.062 15.427 25.196 qs_ot_get_derivative 117 11.6 0.001 0.002 22.893 23.285 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.053 21.176 apply_single 128 13.6 0.001 0.001 20.053 21.176 qs_ot_get_p 128 10.4 0.001 0.001 19.454 19.942 ot_diis_step 117 11.6 0.018 0.018 19.794 19.794 multiply_cannon_sync_h2d 20056 15.6 14.185 15.796 14.185 15.796 make_m2s 5014 13.6 0.081 0.086 14.926 15.796 make_images 5014 14.6 1.169 1.269 14.694 15.563 qs_ot_p2m_diag 83 11.4 0.265 0.273 15.176 15.188 sum_up_and_integrate 128 10.3 0.134 0.146 14.790 14.817 cp_fm_cholesky_invert 11 10.9 14.679 14.689 14.679 14.689 integrate_v_rspace 128 11.3 0.004 0.004 14.655 14.680 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.436 14.463 calculate_rho_elec 128 8.7 0.130 0.145 14.435 14.462 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.156 14.158 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.930 10.949 cp_fm_diag_elpa_base 83 14.4 10.518 10.671 10.926 10.946 init_scf_run 11 5.9 0.000 0.001 10.918 10.918 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.918 10.918 make_images_data 5014 15.6 0.060 0.068 9.188 10.544 hybrid_alltoall_any 5200 16.5 0.434 1.989 7.874 9.913 multiply_cannon_metrocomm4 17549 15.6 0.063 0.074 3.477 9.311 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.967 9.247 density_rs2pw 128 9.7 0.006 0.007 7.269 9.209 pw_transfer 1547 11.6 0.086 0.104 8.978 9.077 mp_irecv_dv 50230 16.2 3.352 9.059 3.352 9.059 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 8.753 8.858 dbcsr_mm_accdrv_process 41502 16.2 4.479 5.343 8.202 8.311 fft_wrap_pw1pw2_140 523 13.2 1.419 1.443 7.740 7.860 grid_integrate_task_list 128 12.3 7.274 7.786 7.274 7.786 wfi_extrapolate 11 7.9 0.001 0.001 7.441 7.441 cp_fm_cholesky_decompose 22 10.9 7.328 7.340 7.328 7.340 cp_fm_upper_to_full 105 14.8 5.747 7.298 5.747 7.298 rs_pw_transfer 1046 11.9 0.014 0.014 5.342 7.295 dbcsr_complete_redistribute 395 12.7 1.185 1.251 4.708 6.521 calculate_dm_sparse 128 9.5 0.001 0.001 6.158 6.266 fft3d_ps 1291 14.7 2.737 2.957 5.867 5.950 grid_collocate_task_list 128 9.7 4.903 5.754 4.903 5.754 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.516 5.525 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.516 5.330 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.606 5.281 mp_alltoall_d11v 2415 14.1 4.483 5.107 4.483 5.107 potential_pw2rs 128 12.3 0.021 0.023 4.840 4.860 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.549 4.693 mp_allgather_i34 2507 14.6 1.546 4.688 1.546 4.688 mp_sum_l 7870 13.0 3.272 4.619 3.272 4.619 mp_waitany 11748 13.9 2.420 4.451 2.420 4.451 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.377 4.170 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.047 4.081 rs_pw_transfer_RS2PW_140 139 11.5 0.329 0.356 1.958 3.908 mp_alltoall_i22 716 14.1 1.971 3.901 1.971 3.901 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.788 3.789 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.587 3.636 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.260000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=893.181818, yerr=11.352358 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_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 1.161257E+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 3992 58354. MP_Allreduce 10977 1091. MP_Sync 87 MP_Alltoall 1712 12503084. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.050 187.075 187.085 qs_mol_dyn_low 1 2.0 0.003 0.003 186.696 186.710 qs_forces 11 3.9 0.004 0.005 186.571 186.580 qs_energies 11 4.9 0.001 0.001 179.378 179.389 scf_env_do_scf 11 5.9 0.001 0.001 162.783 162.790 velocity_verlet 10 3.0 0.001 0.001 124.402 124.405 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 116.566 116.567 dbcsr_multiply_generic 2507 12.6 0.186 0.193 79.120 79.950 qs_scf_new_mos 117 7.6 0.001 0.001 78.922 79.181 qs_scf_loop_do_ot 117 8.6 0.001 0.001 78.921 79.181 ot_scf_mini 117 9.6 0.003 0.004 74.443 74.727 multiply_cannon 2507 13.6 0.554 0.584 54.607 58.357 multiply_cannon_loop 2507 14.6 1.176 1.206 50.973 52.894 init_scf_loop 11 6.9 0.000 0.000 46.096 46.097 ot_mini 117 10.6 0.001 0.001 42.062 42.330 prepare_preconditioner 11 7.9 0.000 0.000 42.005 42.035 make_preconditioner 11 8.9 0.000 0.000 42.005 42.035 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.761 40.701 multiply_cannon_multrec 30084 15.6 14.217 19.084 26.098 30.844 rebuild_ks_matrix 128 8.3 0.001 0.001 30.043 30.369 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 30.042 30.368 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.057 27.353 mp_waitall_1 147882 16.7 17.308 26.480 17.308 26.480 qs_ot_get_derivative 117 11.6 0.001 0.002 22.642 22.923 make_m2s 5014 13.6 0.096 0.100 20.161 21.219 make_images 5014 14.6 1.933 2.225 19.854 20.910 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.784 19.371 apply_single 128 13.6 0.001 0.001 18.784 19.370 ot_diis_step 117 11.6 0.018 0.021 19.287 19.289 qs_ot_get_p 128 10.4 0.001 0.001 18.650 18.933 cp_fm_upper_to_full 105 14.8 11.357 16.671 11.357 16.671 cp_fm_cholesky_invert 11 10.9 16.126 16.135 16.126 16.135 multiply_cannon_metrocomm3 30084 15.6 0.047 0.049 6.337 15.127 sum_up_and_integrate 128 10.3 0.140 0.152 14.752 14.776 integrate_v_rspace 128 11.3 0.004 0.004 14.612 14.641 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.600 14.625 calculate_rho_elec 128 8.7 0.173 0.189 14.600 14.624 qs_ot_p2m_diag 83 11.4 0.342 0.390 14.550 14.601 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.284 13.285 multiply_cannon_sync_h2d 30084 15.6 11.795 12.733 11.795 12.733 make_images_data 5014 15.6 0.063 0.069 10.947 12.717 dbcsr_complete_redistribute 395 12.7 1.495 1.624 8.762 12.490 dbcsr_mm_accdrv_process 62264 16.2 7.304 8.197 11.456 11.977 hybrid_alltoall_any 5200 16.5 0.523 2.200 9.580 11.897 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.411 11.156 init_scf_run 11 5.9 0.000 0.001 10.761 10.762 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.760 10.762 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.226 10.235 cp_fm_diag_elpa_base 83 14.4 9.276 9.590 10.219 10.229 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.223 9.871 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.381 9.581 mp_alltoall_i22 716 14.1 5.508 9.320 5.508 9.320 pw_transfer 1547 11.6 0.086 0.099 9.193 9.275 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.967 9.057 density_rs2pw 128 9.7 0.006 0.007 7.018 8.310 fft_wrap_pw1pw2_140 523 13.2 1.564 1.597 7.981 8.070 grid_integrate_task_list 128 12.3 7.532 7.984 7.532 7.984 wfi_extrapolate 11 7.9 0.001 0.001 7.673 7.673 cp_fm_cholesky_decompose 22 10.9 7.523 7.619 7.523 7.619 multiply_cannon_metrocomm4 25070 15.6 0.083 0.094 2.731 6.787 mp_irecv_dv 76098 16.2 2.578 6.510 2.578 6.510 calculate_dm_sparse 128 9.5 0.001 0.001 6.143 6.224 rs_pw_transfer 1046 11.9 0.013 0.014 4.651 5.996 fft3d_ps 1291 14.7 2.843 2.919 5.879 5.956 grid_collocate_task_list 128 9.7 5.075 5.754 5.075 5.754 mp_alltoall_d11v 2415 14.1 4.799 5.621 4.799 5.621 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.461 5.522 potential_pw2rs 128 12.3 0.023 0.024 4.513 4.529 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.423 4.492 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.346 4.429 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.420 4.421 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.144 4.209 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.085000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1092.909091, yerr=16.267394 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243766E+12 0.0% 0.0% 100.0% flops max/rank 5.910792E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547424 0.0% 0.0% 100.0% number of processed stacks 1978768 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 1.527833E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 243264 MPI messages size (bytes): total size 1.342058E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.516879E+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 115488 60548972544 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592540656 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9025 51. MP_Alltoall 9734 793691. MP_ISend 40500 2096702. MP_IRecv 40500 2095807. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - 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 0.028 0.066 174.515 174.526 qs_mol_dyn_low 1 2.0 0.003 0.003 173.992 173.999 qs_forces 11 3.9 0.036 0.038 173.552 173.556 qs_energies 11 4.9 0.001 0.001 165.937 165.949 scf_env_do_scf 11 5.9 0.001 0.001 148.610 148.611 velocity_verlet 10 3.0 0.001 0.001 114.731 114.755 scf_env_do_scf_inner_loop 118 6.6 0.009 0.013 113.214 113.215 dbcsr_multiply_generic 2534 12.6 0.181 0.192 74.654 75.161 qs_scf_new_mos 118 7.6 0.001 0.001 74.206 74.302 qs_scf_loop_do_ot 118 8.6 0.001 0.001 74.205 74.301 ot_scf_mini 118 9.6 0.004 0.004 69.757 69.837 multiply_cannon 2534 13.6 0.588 0.616 54.973 58.877 multiply_cannon_loop 2534 14.6 0.449 0.460 50.276 51.131 ot_mini 118 10.6 0.001 0.001 39.782 39.862 init_scf_loop 11 6.9 0.000 0.000 35.250 35.251 mp_waitall_1 127116 16.7 25.955 33.520 25.955 33.520 prepare_preconditioner 11 7.9 0.000 0.000 31.345 31.366 make_preconditioner 11 8.9 0.000 0.000 31.345 31.366 rebuild_ks_matrix 129 8.3 0.001 0.001 30.151 30.218 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.019 30.150 30.217 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.314 29.594 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.410 27.479 multiply_cannon_multrec 10136 15.6 10.466 14.607 18.002 20.768 ot_diis_step 118 11.6 0.020 0.021 20.030 20.031 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.553 19.952 apply_single 129 13.6 0.001 0.001 19.553 19.951 qs_ot_get_derivative 118 11.6 0.002 0.002 19.684 19.781 multiply_cannon_metrocomm3 10136 15.6 0.023 0.025 12.384 19.285 make_m2s 5068 13.6 0.067 0.070 16.149 18.450 make_images 5068 14.6 2.306 2.800 15.841 18.137 cp_fm_cholesky_invert 11 10.9 17.925 17.931 17.925 17.931 qs_ot_get_p 129 10.4 0.001 0.001 17.003 17.088 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.469 15.508 calculate_rho_elec 129 8.7 0.258 0.269 15.468 15.508 sum_up_and_integrate 129 10.3 0.180 0.190 15.163 15.209 integrate_v_rspace 129 11.3 0.004 0.004 14.982 15.035 qs_ot_p2m_diag 84 11.4 0.501 0.507 13.249 13.266 make_images_data 5068 15.6 0.052 0.061 9.811 12.324 hybrid_alltoall_any 5255 16.5 0.822 3.847 9.620 12.217 multiply_cannon_sync_h2d 10136 15.6 11.742 12.204 11.742 12.204 cp_dbcsr_syevd 84 12.4 0.005 0.006 12.116 12.117 init_scf_run 11 5.9 0.000 0.001 10.553 10.553 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.553 10.553 pw_transfer 1559 11.6 0.087 0.096 9.572 9.589 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.346 9.368 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.113 9.123 cp_fm_diag_elpa_base 84 14.4 8.861 8.947 9.106 9.117 fft_wrap_pw1pw2_140 527 13.2 1.934 1.966 8.269 8.296 grid_integrate_task_list 129 12.3 7.798 8.213 7.798 8.213 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 7.919 7.965 cp_fm_cholesky_decompose 22 10.9 7.839 7.951 7.839 7.951 dbcsr_mm_accdrv_process 20954 16.1 2.677 3.544 7.161 7.833 density_rs2pw 129 9.7 0.006 0.006 6.943 7.824 multiply_cannon_metrocomm1 10136 15.6 0.028 0.029 4.485 7.388 wfi_extrapolate 11 7.9 0.001 0.001 7.381 7.381 mp_allgather_i34 2534 14.6 2.883 7.045 2.883 7.045 calculate_dm_sparse 129 9.5 0.001 0.001 6.229 6.322 grid_collocate_task_list 129 9.7 5.436 6.145 5.436 6.145 mp_alltoall_d11v 2429 14.1 5.043 5.863 5.043 5.863 fft3d_ps 1301 14.7 2.791 2.873 5.785 5.827 dbcsr_complete_redistribute 397 12.7 2.107 2.148 5.005 5.395 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.278 5.289 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.211 5.220 rs_pw_transfer 1054 12.0 0.012 0.013 4.178 5.087 potential_pw2rs 129 12.3 0.027 0.028 4.472 4.484 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.170 4.201 multiply_cannon_metrocomm4 7602 15.6 0.024 0.027 1.860 4.059 mp_irecv_dv 29142 15.9 1.824 3.980 1.824 3.980 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.567 3.886 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.708 3.742 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.654 3.665 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.480 3.591 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.244 3.551 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.466 3.498 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.526000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1429.909091, yerr=56.516216 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 11.696234E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 1964048 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 2.957697E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 100280 MPI messages size (bytes): total size 1.136195E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.330227E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45208 35089547264 4194304 < size <= 16777216 44352 379752284160 16777216 < size 10104 721350232272 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 58886. MP_Allreduce 11047 1513. MP_Sync 87 MP_Alltoall 1712 36974159. MP_SendRecv 1792 218624. MP_ISendRecv 1792 218624. MP_Wait 9802 MP_ISend 6408 1080322. MP_IRecv 6408 1080322. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.105 0.142 299.065 299.078 qs_mol_dyn_low 1 2.0 0.003 0.003 298.269 298.276 qs_forces 11 3.9 0.005 0.005 298.140 298.142 qs_energies 11 4.9 0.001 0.001 288.860 288.869 scf_env_do_scf 11 5.9 0.001 0.001 266.155 266.165 velocity_verlet 10 3.0 0.001 0.001 214.603 214.610 scf_env_do_scf_inner_loop 117 6.6 0.019 0.035 141.702 141.703 init_scf_loop 11 6.9 0.000 0.000 124.204 124.206 prepare_preconditioner 11 7.9 0.000 0.000 119.109 119.136 make_preconditioner 11 8.9 0.000 0.000 119.109 119.136 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.530 116.249 qs_scf_new_mos 117 7.6 0.001 0.001 91.085 91.161 qs_scf_loop_do_ot 117 8.6 0.001 0.001 91.085 91.161 ot_scf_mini 117 9.6 0.004 0.004 86.309 86.345 dbcsr_multiply_generic 2507 12.6 0.211 0.223 84.222 84.826 cp_fm_upper_to_full 105 14.8 52.739 75.293 52.739 75.293 multiply_cannon 2507 13.6 0.695 0.736 59.286 60.052 multiply_cannon_loop 2507 14.6 0.472 0.479 55.489 57.112 ot_mini 117 10.6 0.001 0.001 44.752 44.783 dbcsr_complete_redistribute 395 12.7 3.990 4.049 29.158 41.837 copy_fm_to_dbcsr 209 11.7 0.002 0.002 25.798 38.485 rebuild_ks_matrix 128 8.3 0.001 0.001 38.225 38.312 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.018 38.225 38.311 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.533 36.105 qs_ks_update_qs_env 128 7.6 0.001 0.001 35.276 35.359 mp_alltoall_i22 716 14.1 21.344 33.944 21.344 33.944 cp_fm_cholesky_invert 11 10.9 33.640 33.646 33.640 33.646 mp_waitall_1 103674 16.8 28.638 32.283 28.638 32.283 qs_ot_get_p 128 10.4 0.001 0.001 26.031 26.071 qs_ot_get_derivative 117 11.6 0.002 0.002 24.263 24.304 qs_ot_p2m_diag 83 11.4 0.878 0.883 21.935 21.963 make_m2s 5014 13.6 0.077 0.079 20.211 21.365 make_images 5014 14.6 3.716 3.826 19.731 20.889 ot_diis_step 117 11.6 0.022 0.023 20.449 20.449 qs_rho_update_rho_low 128 7.7 0.001 0.001 20.423 20.436 calculate_rho_elec 128 8.7 0.481 0.482 20.422 20.435 cp_dbcsr_syevd 83 12.4 0.006 0.006 20.195 20.196 multiply_cannon_metrocomm3 10028 15.6 0.023 0.024 18.787 20.162 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.659 19.922 apply_single 128 13.6 0.001 0.001 19.659 19.921 sum_up_and_integrate 128 10.3 0.320 0.325 19.584 19.665 integrate_v_rspace 128 11.3 0.004 0.004 19.264 19.344 multiply_cannon_multrec 10028 15.6 10.470 12.198 17.999 18.103 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.987 16.988 cp_fm_diag_elpa_base 83 14.4 12.593 14.176 16.983 16.983 multiply_cannon_sync_h2d 10028 15.6 15.655 15.665 15.655 15.665 hybrid_alltoall_any 5200 16.5 1.298 3.043 11.107 13.073 make_images_data 5014 15.6 0.059 0.064 11.049 12.903 init_scf_run 11 5.9 0.000 0.001 12.769 12.770 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.769 12.769 pw_transfer 1547 11.6 0.093 0.094 11.500 11.509 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 11.265 11.275 fft_wrap_pw1pw2_140 523 13.2 3.075 3.100 9.981 9.991 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.464 9.497 dbcsr_mm_accdrv_process 20762 16.1 3.805 5.784 7.292 9.159 mp_alltoall_d11v 2415 14.1 8.261 9.087 8.261 9.087 wfi_extrapolate 11 7.9 0.001 0.001 8.985 8.985 cp_fm_cholesky_decompose 22 10.9 8.912 8.934 8.912 8.934 grid_integrate_task_list 128 12.3 8.583 8.778 8.583 8.778 density_rs2pw 128 9.7 0.006 0.006 8.015 8.146 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.989 7.989 calculate_dm_sparse 128 9.5 0.001 0.001 7.184 7.239 grid_collocate_task_list 128 9.7 6.364 6.406 6.364 6.406 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.317 6.374 fft3d_ps 1291 14.7 2.847 2.855 6.336 6.351 rs_scatter_matrices 139 9.7 3.636 4.470 6.010 6.235 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.006 6.102 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=299.078000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2636.454545, yerr=161.755912 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_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 1.261826E+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 40443104. 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 0.018 0.030 84.983 84.984 qs_energies 1 2.0 0.000 0.000 84.544 84.557 ls_scf 1 3.0 0.000 0.000 83.644 83.657 dbcsr_multiply_generic 111 6.7 0.014 0.016 72.528 72.701 multiply_cannon 111 7.7 0.017 0.020 55.832 57.279 multiply_cannon_loop 111 8.7 0.207 0.221 52.407 54.034 ls_scf_main 1 4.0 0.000 0.000 52.192 52.192 density_matrix_trs4 2 5.0 0.002 0.003 46.690 46.778 ls_scf_init_scf 1 4.0 0.000 0.000 28.399 28.400 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.316 27.371 mp_waitall_1 11031 10.9 22.459 25.214 22.459 25.214 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.001 25.029 multiply_cannon_multrec 2664 9.7 8.164 8.857 15.457 17.145 multiply_cannon_sync_h2d 2664 9.7 13.557 15.783 13.557 15.783 make_m2s 222 7.7 0.008 0.011 12.974 13.524 make_images 222 8.7 0.099 0.109 12.952 13.504 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.710 13.402 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.494 8.705 make_images_data 222 9.7 0.004 0.005 7.590 8.090 dbcsr_mm_accdrv_process 4760 10.4 0.506 0.611 6.915 7.884 hybrid_alltoall_any 227 10.6 0.215 1.824 6.576 7.841 dbcsr_mm_accdrv_process_sort 4760 11.4 6.209 7.106 6.209 7.106 calculate_norms 4752 9.8 5.507 6.124 5.507 6.124 apply_matrix_preconditioner 6 5.3 0.001 0.003 5.039 5.116 mp_sum_l 807 5.4 3.184 4.623 3.184 4.623 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.386 3.468 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.084 3.369 mp_irecv_dv 6231 10.9 2.066 3.341 2.066 3.341 make_images_sizes 222 9.7 0.000 0.000 0.657 3.311 mp_alltoall_i44 222 10.7 0.656 3.311 0.656 3.311 arnoldi_extremal 4 6.8 0.000 0.000 3.210 3.235 arnoldi_normal_ev 4 7.8 0.001 0.002 3.210 3.234 build_subspace 16 8.4 0.009 0.012 3.105 3.107 ls_scf_post 1 4.0 0.000 0.000 3.054 3.067 ls_scf_store_result 1 5.0 0.000 0.000 2.883 2.916 dbcsr_special_finalize 555 9.7 0.005 0.006 2.376 2.781 dbcsr_merge_single_wm 555 10.7 0.455 0.586 2.367 2.773 make_images_pack 222 9.7 2.205 2.622 2.207 2.623 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.307 2.548 dbcsr_sort_data 658 11.4 2.170 2.496 2.170 2.496 dbcsr_matrix_vector_mult_local 304 10.0 2.065 2.456 2.067 2.458 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.261 2.330 buffer_matrices_ensure_size 222 8.7 1.749 2.085 1.749 2.085 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.765 1.766 rebuild_ks_matrix 3 7.3 0.000 0.000 1.756 1.757 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.756 1.757 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.984000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1133.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_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 2.060489E+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 0.049 0.072 91.651 91.665 qs_energies 1 2.0 0.000 0.000 91.054 91.058 ls_scf 1 3.0 0.000 0.000 89.730 89.733 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.655 76.036 multiply_cannon 111 7.7 0.029 0.045 53.269 57.340 ls_scf_main 1 4.0 0.000 0.000 55.283 55.288 multiply_cannon_loop 111 8.7 0.115 0.122 49.983 53.297 density_matrix_trs4 2 5.0 0.002 0.003 49.580 49.787 mp_waitall_1 9105 10.9 21.436 31.109 21.436 31.109 ls_scf_init_scf 1 4.0 0.000 0.000 30.873 30.873 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.665 29.777 multiply_cannon_multrec 1332 9.7 13.209 16.896 22.422 27.240 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.191 27.209 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.801 21.518 make_m2s 222 7.7 0.006 0.007 15.660 16.403 make_images 222 8.7 1.574 1.966 15.631 16.375 dbcsr_mm_accdrv_process 4041 10.4 0.307 0.442 8.814 10.316 make_images_data 222 9.7 0.004 0.004 9.070 10.122 dbcsr_mm_accdrv_process_sort 4041 11.4 8.404 9.890 8.404 9.890 hybrid_alltoall_any 227 10.6 0.521 2.482 8.483 9.606 mp_sum_l 807 5.4 5.424 8.784 5.424 8.784 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.227 7.693 mp_irecv_dv 3311 11.0 3.207 7.626 3.207 7.626 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.218 7.012 calculate_norms 2376 9.8 5.994 6.690 5.994 6.690 multiply_cannon_sync_h2d 1332 9.7 4.752 6.143 4.752 6.143 apply_matrix_preconditioner 6 5.3 0.000 0.001 5.065 5.254 arnoldi_extremal 4 6.8 0.000 0.000 4.665 4.686 arnoldi_normal_ev 4 7.8 0.001 0.004 4.665 4.686 build_subspace 16 8.4 0.014 0.021 4.413 4.417 ls_scf_post 1 4.0 0.000 0.000 3.575 3.578 ls_scf_store_result 1 5.0 0.000 0.000 3.253 3.400 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.140 3.368 dbcsr_matrix_vector_mult_local 304 10.0 2.739 3.217 2.741 3.219 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.196 2.929 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.577 2.660 mp_allgather_i34 111 8.7 0.984 2.492 0.984 2.492 make_images_pack 222 9.7 2.028 2.412 2.031 2.413 dbcsr_sort_data 436 11.2 1.858 2.104 1.858 2.104 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.879 1.881 rebuild_ks_matrix 3 7.3 0.000 0.000 1.867 1.869 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.867 1.869 dbcsr_data_new 4174 10.1 1.614 1.836 1.614 1.836 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.665000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1735.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_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 2.790298E+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 0.073 0.146 94.481 94.482 qs_energies 1 2.0 0.000 0.000 93.705 93.757 ls_scf 1 3.0 0.002 0.015 92.029 92.083 dbcsr_multiply_generic 111 6.7 0.016 0.016 76.201 76.508 ls_scf_main 1 4.0 0.001 0.013 57.140 57.146 multiply_cannon 111 7.7 0.037 0.073 52.757 56.487 multiply_cannon_loop 111 8.7 0.100 0.104 49.134 53.173 density_matrix_trs4 2 5.0 0.005 0.035 51.109 51.297 mp_waitall_1 7281 11.0 23.621 33.901 23.621 33.901 ls_scf_init_scf 1 4.0 0.000 0.002 31.227 31.229 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.749 29.823 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.257 27.269 multiply_cannon_multrec 888 9.7 12.617 15.399 21.111 24.506 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.880 23.103 make_m2s 222 7.7 0.006 0.006 16.777 18.034 make_images 222 8.7 1.970 2.293 16.739 17.996 make_images_data 222 9.7 0.003 0.004 9.595 10.714 hybrid_alltoall_any 227 10.6 0.621 2.864 9.331 10.480 dbcsr_mm_accdrv_process 3754 10.4 0.247 0.422 8.047 9.316 dbcsr_mm_accdrv_process_sort 3754 11.4 7.676 8.895 7.676 8.895 mp_sum_l 807 5.4 5.163 8.508 5.163 8.508 multiply_cannon_sync_h2d 888 9.7 5.999 7.608 5.999 7.608 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.766 7.076 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.440 7.062 mp_irecv_dv 2335 11.1 2.425 7.020 2.425 7.020 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.878 6.442 arnoldi_extremal 4 6.8 0.000 0.002 5.150 5.165 arnoldi_normal_ev 4 7.8 0.004 0.028 5.150 5.165 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.962 5.151 build_subspace 16 8.4 0.015 0.027 4.844 4.854 calculate_norms 1584 9.8 4.322 4.742 4.322 4.742 mp_allgather_i34 111 8.7 1.455 3.764 1.455 3.764 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.432 3.759 ls_scf_post 1 4.0 0.002 0.015 3.661 3.715 dbcsr_matrix_vector_mult_local 304 10.0 3.017 3.592 3.019 3.594 ls_scf_store_result 1 5.0 0.000 0.000 3.296 3.397 ls_scf_dm_to_ks 2 5.0 0.001 0.011 2.862 3.002 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.114 2.116 make_images_pack 222 9.7 1.824 2.111 1.826 2.114 rebuild_ks_matrix 3 7.3 0.000 0.000 2.096 2.099 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.022 2.096 2.099 make_images_sizes 222 9.7 0.000 0.000 0.867 2.093 mp_alltoall_i44 222 10.7 0.867 2.093 0.867 2.093 dbcsr_sort_data 325 11.1 1.885 2.066 1.885 2.066 dbcsr_data_release 9322 10.9 1.276 1.995 1.276 1.995 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.482000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2185.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_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 3.337318E+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 0.079 0.168 97.267 97.276 qs_energies 1 2.0 0.000 0.000 96.475 96.487 ls_scf 1 3.0 0.001 0.010 94.743 94.755 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.354 78.585 ls_scf_main 1 4.0 0.001 0.010 58.658 58.660 multiply_cannon 111 7.7 0.066 0.175 51.774 55.737 density_matrix_trs4 2 5.0 0.004 0.022 52.557 52.683 multiply_cannon_loop 111 8.7 0.114 0.126 46.796 49.518 ls_scf_init_scf 1 4.0 0.000 0.002 32.867 32.869 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.426 31.481 mp_waitall_1 6369 11.0 23.024 29.921 23.024 29.921 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.961 28.976 multiply_cannon_multrec 1332 9.7 14.139 17.274 21.990 24.791 make_m2s 222 7.7 0.007 0.008 21.142 22.528 make_images 222 8.7 3.140 3.599 21.091 22.480 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.486 16.882 make_images_data 222 9.7 0.004 0.004 11.802 13.367 hybrid_alltoall_any 227 10.6 0.796 3.763 11.177 12.895 dbcsr_mm_accdrv_process 3641 10.4 0.215 0.429 7.491 9.005 dbcsr_mm_accdrv_process_sort 3641 11.4 7.110 8.589 7.110 8.589 mp_sum_l 807 5.4 3.878 6.954 3.878 6.954 multiply_cannon_sync_h2d 1332 9.7 5.545 6.143 5.545 6.143 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.092 6.085 mp_irecv_dv 3229 10.9 2.068 6.005 2.068 6.005 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.968 5.378 arnoldi_extremal 4 6.8 0.000 0.000 5.207 5.219 arnoldi_normal_ev 4 7.8 0.001 0.004 5.207 5.219 build_subspace 16 8.4 0.014 0.021 4.873 4.879 apply_matrix_preconditioner 6 5.3 0.000 0.001 4.539 4.720 calculate_norms 2376 9.8 4.186 4.592 4.186 4.592 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.558 4.509 mp_allgather_i34 111 8.7 2.091 4.246 2.091 4.246 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.563 3.866 dbcsr_matrix_vector_mult_local 304 10.0 3.183 3.687 3.186 3.689 dbcsr_sort_data 658 11.4 3.059 3.411 3.059 3.411 dbcsr_special_finalize 555 9.7 0.006 0.007 2.808 3.231 ls_scf_post 1 4.0 0.002 0.013 3.216 3.228 dbcsr_merge_single_wm 555 10.7 0.532 0.667 2.799 3.222 ls_scf_dm_to_ks 2 5.0 0.001 0.015 2.984 3.036 ls_scf_store_result 1 5.0 0.000 0.000 2.973 3.019 dbcsr_data_release 10477 10.7 1.585 2.395 1.585 2.395 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.033 2.035 rebuild_ks_matrix 3 7.3 0.000 0.000 2.011 2.012 qs_ks_build_kohn_sham_matrix 3 8.3 0.006 0.047 2.011 2.012 dbcsr_finalize 304 7.8 0.049 0.061 1.794 1.988 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.276000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2745.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_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 4.683022E+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 0.038 0.056 99.136 99.137 qs_energies 1 2.0 0.000 0.000 98.378 98.383 ls_scf 1 3.0 0.000 0.000 96.371 96.381 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.900 78.079 ls_scf_main 1 4.0 0.000 0.000 62.035 62.037 multiply_cannon 111 7.7 0.089 0.165 55.301 60.921 density_matrix_trs4 2 5.0 0.002 0.003 54.892 54.993 multiply_cannon_loop 111 8.7 0.069 0.076 50.902 52.558 mp_waitall_1 5436 11.0 26.660 31.852 26.660 31.852 ls_scf_init_scf 1 4.0 0.000 0.000 30.761 30.766 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.480 29.544 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.317 27.331 multiply_cannon_multrec 444 9.7 13.985 16.108 21.004 22.983 make_m2s 222 7.7 0.004 0.005 17.798 20.188 make_images 222 8.7 3.731 4.413 17.737 20.128 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.123 16.847 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.330 14.179 make_images_data 222 9.7 0.003 0.004 10.011 12.341 hybrid_alltoall_any 227 10.6 0.790 3.769 9.874 12.248 dbcsr_mm_accdrv_process 3003 10.4 0.188 0.343 6.731 7.873 dbcsr_mm_accdrv_process_sort 3003 11.4 6.415 7.523 6.415 7.523 multiply_cannon_sync_h2d 444 9.7 6.561 7.503 6.561 7.503 mp_allgather_i34 111 8.7 2.628 6.984 2.628 6.984 arnoldi_extremal 4 6.8 0.000 0.000 5.805 5.820 arnoldi_normal_ev 4 7.8 0.001 0.004 5.805 5.820 build_subspace 16 8.4 0.015 0.020 5.418 5.429 mp_sum_l 807 5.4 2.840 4.834 2.840 4.834 apply_matrix_preconditioner 6 5.3 0.001 0.002 4.591 4.722 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.685 4.564 mp_irecv_dv 1241 11.2 1.665 4.539 1.665 4.539 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.161 4.357 dbcsr_matrix_vector_mult_local 304 10.0 3.686 4.158 3.688 4.160 calculate_norms 792 9.8 3.553 3.663 3.553 3.663 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.532 3.628 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.905 3.612 ls_scf_post 1 4.0 0.000 0.000 3.574 3.579 make_images_sizes 222 9.7 0.000 0.000 1.127 3.447 mp_alltoall_i44 222 10.7 1.127 3.446 1.127 3.446 ls_scf_store_result 1 5.0 0.000 0.000 3.360 3.398 dbcsr_finalize 304 7.8 0.062 0.078 2.204 2.279 dbcsr_merge_all 275 8.9 0.475 0.524 2.052 2.117 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.064 2.065 rebuild_ks_matrix 3 7.3 0.000 0.000 2.031 2.032 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.031 2.032 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 1.991 1.991 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.137000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3679.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_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 8.801190E+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 0.088 0.105 108.370 108.371 qs_energies 1 2.0 0.000 0.000 106.805 106.819 ls_scf 1 3.0 0.000 0.000 103.888 103.901 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.477 77.615 ls_scf_main 1 4.0 0.000 0.000 65.980 65.981 density_matrix_trs4 2 5.0 0.002 0.003 56.775 56.860 multiply_cannon 111 7.7 0.150 0.285 49.815 51.372 multiply_cannon_loop 111 8.7 0.067 0.070 46.392 47.371 ls_scf_init_scf 1 4.0 0.000 0.000 34.224 34.225 ls_scf_init_matrix_S 1 5.0 0.000 0.001 32.682 32.698 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.852 29.859 mp_waitall_1 4527 11.1 22.268 26.258 22.268 26.258 make_m2s 222 7.7 0.005 0.005 23.934 24.869 make_images 222 8.7 4.602 4.987 23.828 24.761 multiply_cannon_multrec 444 9.7 17.830 18.522 22.422 23.106 hybrid_alltoall_any 227 10.6 1.661 3.619 12.923 15.814 make_images_data 222 9.7 0.003 0.003 13.164 15.545 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.519 11.295 multiply_cannon_sync_h2d 444 9.7 8.849 8.892 8.849 8.892 arnoldi_extremal 4 6.8 0.000 0.000 7.435 7.446 arnoldi_normal_ev 4 7.8 0.002 0.009 7.435 7.446 build_subspace 16 8.4 0.026 0.036 6.887 6.899 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.485 5.642 dbcsr_matrix_vector_mult_local 304 10.0 5.018 5.341 5.021 5.343 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.041 5.307 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.148 5.242 dbcsr_mm_accdrv_process 1814 10.4 0.231 0.317 4.423 4.562 dbcsr_mm_accdrv_process_sort 1814 11.4 4.124 4.266 4.124 4.266 ls_scf_post 1 4.0 0.000 0.000 3.683 3.696 make_images_sizes 222 9.7 0.000 0.000 1.429 3.606 mp_alltoall_i44 222 10.7 1.429 3.606 1.429 3.606 ls_scf_store_result 1 5.0 0.000 0.000 3.431 3.442 mp_allgather_i34 111 8.7 1.066 3.349 1.066 3.349 calculate_norms 792 9.8 3.242 3.286 3.242 3.286 dbcsr_finalize 304 7.8 0.082 0.089 3.088 3.167 dbcsr_merge_all 275 8.9 0.891 0.917 2.874 2.948 dbcsr_complete_redistribute 5 7.6 1.434 1.459 2.772 2.890 qs_energies_init_hamiltonians 1 3.0 0.002 0.004 2.887 2.887 dbcsr_data_release 12724 10.6 2.331 2.841 2.331 2.841 matrix_ls_to_qs 2 6.0 0.000 0.000 2.422 2.547 dbcsr_sort_data 325 11.1 2.444 2.499 2.444 2.499 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.463 2.465 rebuild_ks_matrix 3 7.3 0.000 0.000 2.395 2.397 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.395 2.397 dbcsr_new_transposed 4 7.5 0.243 0.253 2.280 2.291 mp_sum_l 807 5.4 1.403 2.237 1.403 2.237 dbcsr_frobenius_norm 74 6.6 2.054 2.138 2.192 2.232 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.205 dbcsr_add_anytype 103 7.2 0.860 0.893 2.133 2.205 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.371000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6863.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7f090fb11291b829076ec7a97d9530b1aff7c614_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 586.076160E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 76559. MP_Allreduce 13152 2094. MP_Sync 1064 MP_Alltoall 2588 751865028. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.069 260.088 260.090 qs_mol_dyn_low 1 2.0 0.003 0.010 258.986 259.003 qs_forces 5 3.8 0.005 0.021 258.899 258.901 qs_energies 5 4.8 0.002 0.038 255.489 255.519 scf_env_do_scf 5 5.8 0.001 0.055 240.753 240.755 scf_env_do_scf_inner_loop 105 6.6 0.004 0.059 209.935 209.936 qs_scf_new_mos 105 7.6 0.000 0.001 163.753 163.969 qs_scf_loop_do_ot 105 8.6 0.001 0.001 163.752 163.969 ot_scf_mini 105 9.6 0.003 0.004 153.711 153.840 dbcsr_multiply_generic 1445 12.2 0.125 0.133 134.573 135.113 multiply_cannon 1445 13.2 0.276 0.289 113.661 115.717 velocity_verlet 4 3.0 0.000 0.001 113.019 113.020 multiply_cannon_loop 1445 14.2 2.353 2.479 111.459 112.780 qs_ot_get_p 112 10.4 0.001 0.001 62.091 62.528 ot_mini 105 10.6 0.001 0.013 61.212 61.355 qs_ot_p2m_diag 40 11.0 0.020 0.031 50.044 50.161 cp_dbcsr_syevd 40 12.0 0.002 0.002 46.591 46.592 multiply_cannon_multrec 69360 15.2 31.725 36.787 39.425 44.859 mp_waitall_1 488190 16.1 36.411 43.953 36.411 43.953 cp_fm_syevd 40 13.0 0.000 0.000 40.396 40.545 qs_ot_get_derivative 55 11.6 0.001 0.002 39.529 39.682 cp_fm_redistribute_end 40 14.0 17.623 35.157 17.631 35.161 cp_fm_syevd_base 40 14.0 17.516 35.053 17.516 35.053 multiply_cannon_metrocomm3 69360 15.2 0.192 0.204 25.719 34.552 multiply_cannon_sync_h2d 69360 15.2 30.591 34.361 30.591 34.361 rebuild_ks_matrix 110 8.4 0.000 0.000 31.715 31.908 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.014 31.714 31.907 init_scf_loop 7 6.6 0.000 0.017 30.772 30.775 qs_ks_update_qs_env 112 7.6 0.001 0.001 28.939 29.112 prepare_preconditioner 7 7.6 0.000 0.000 25.362 25.397 make_preconditioner 7 8.6 0.000 0.006 25.362 25.397 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.039 23.263 apply_single 62 13.6 0.000 0.000 23.039 23.263 qs_rho_update_rho_low 110 7.6 0.001 0.001 21.284 21.625 calculate_rho_elec 110 8.6 0.030 0.033 21.283 21.625 ot_new_cg_direction 55 11.6 0.001 0.052 21.005 21.042 rs_pw_transfer 690 11.5 0.011 0.012 16.148 17.452 make_full_inverse_cholesky 7 9.6 0.000 0.000 16.994 17.068 density_rs2pw 110 9.6 0.006 0.013 14.921 16.272 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.134 15.350 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.133 14.237 mp_sum_l 4684 12.4 12.900 13.972 12.900 13.972 pw_transfer 1645 12.4 0.085 0.110 13.148 13.387 fft_wrap_pw1pw2 1425 13.5 0.013 0.017 13.004 13.248 init_scf_run 5 5.8 0.000 0.004 11.848 11.848 scf_env_initial_rho_setup 5 6.8 0.002 0.004 11.847 11.848 calculate_dm_sparse 110 9.5 0.000 0.001 11.664 11.847 cp_fm_cholesky_invert 7 10.6 11.521 11.532 11.521 11.532 fft_wrap_pw1pw2_240 915 15.0 1.215 1.306 10.914 11.148 qs_vxc_create 110 10.4 0.002 0.007 10.955 11.010 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.795 10.870 check_diag 80 13.5 8.587 8.854 10.693 10.842 sum_up_and_integrate 60 10.3 0.028 0.031 10.676 10.688 integrate_v_rspace 60 11.3 0.002 0.002 10.648 10.660 fft3d_pb 915 16.0 2.390 2.702 8.942 9.274 dbcsr_mm_accdrv_process 154766 15.8 3.913 4.176 7.570 8.505 make_m2s 2890 13.2 0.076 0.083 7.562 8.189 make_images 2890 14.2 0.243 0.261 7.460 8.086 make_full_single_inverse 7 9.6 0.001 0.026 8.033 8.068 xc_rho_set_and_dset_create 110 12.4 0.076 0.096 7.511 7.778 multiply_cannon_metrocomm1 69360 15.2 0.089 0.094 4.565 7.579 xc_vxc_pw_create 60 11.3 0.038 0.049 7.347 7.402 xc_pw_derive 510 13.4 0.006 0.007 6.616 6.680 potential_pw2rs 60 12.3 0.003 0.003 6.610 6.666 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 6.645 6.661 calculate_first_density_matrix 1 7.0 0.000 0.009 6.564 6.585 mp_alltoall_z22v 2340 17.7 6.008 6.460 6.008 6.460 acc_transpose_blocks 69360 15.2 0.349 0.371 5.983 6.330 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.144 6.203 mp_waitany 7680 13.5 4.577 5.948 4.577 5.948 rs_pw_transfer_RS2PW_30 110 11.6 1.579 1.674 5.171 5.499 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=260.090000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=557.800000, yerr=1.600000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 7f090fb11291b829076ec7a97d9530b1aff7c614 Summary: empty Status: OK